The Political Economy of Welfare Reform in the United States

The Political Economy of Welfare Reform in the United States The Locke Institute General Director Charles K. Rowley, P...

Author: Mary Reintsma

88 downloads 1051 Views 872KB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

The Political Economy of Welfare Reform in the United States

The Locke Institute General Director Charles K. Rowley, Ph.D.

Financial Director Robert S. Elgin, M.A.

Director of Legal Studies Amanda J. Owens, Esq., J.D.

Program Director James T. Bennett, Ph.D.

Editorial Director Arthur Seldon, C.B.E.

Conference Director Marjorie I. Rowley

Founded in 1989, The Locke Institute is an independent, non-partisan, educational and research organization. The Institute is named for John Locke (1632–1704), philosopher and political theorist, who based his theory of society on natural law which required that the ultimate source of political sovereignty was with the individual. Individuals are possessed of inalienable rights variously defined by Locke as ‘life, health, liberty and possession’, or more directly, ‘life, liberty and property’. It is the function of the state to uphold these rights since individuals would not enter into a political society unless they believed that it would protect their lives, liberties and properties. The Locke Institute seeks to engender a greater understanding of the concept of natural rights, its implications for constitutional democracy and for economic organization in modern society. The Institute encourages high-quality research utilizing in particular modern theories of property rights, public choice, law and economics, and the new institutional economics as a basis for a more profound understanding of important and controversial issues in political economy. To this end, it commissions books, monographs and shorter studies involving substantive scholarship written for a wide audience, organizes major conferences on fundamental topics in political economy, and supports independent research. The Institute maintains a publishing relationship with Edward Elgar Publishing. It also publishes its own monograph series. In order to maintain independence, The Lock Institute accepts no government funding. Funding for the Institute is solicited from private foundations, corporations and individuals. In addition, the Institute raises funds from the sale of publications and from conference fees. The Institute is incorporated in the State of Virginia, USA, and enjoys non-profit, tax-exempt status under Section 501(c)3 of the United States Internal Revenue Code. Officers of the Institute are listed above. Please direct all enquiries to the address given below. 5188 Dungannon Road, Fairfax, Virginia 22030, USA Tel. (703) 934 6934 Fax (703) 934 6927 email: [email protected] http://www.TheLockeInstitute.org

The Political Economy of Welfare Reform in the United States Mary Reintsma Department of Economics, Trinity University, Washington, DC, USA

THE LOCKE INSTITUTE

Edward Elgar Cheltenham, UK • Northampton, MA, USA

© Mary Reintsma 2007 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited Glensanda House Montpellier Parade Cheltenham Glos GL50 1UA UK Edward Elgar Publishing, Inc. William Pratt House 9 Dewey Court Northampton Massachusetts 01060 USA

A catalogue record for this book is available from the British Library Library of Congress Cataloguing in Publication Data Reintsma, Mary. The political economy of welfare reform in the United States / Mary Reintsma. p. cm. — (Locke Institute series) Includes bibliographical references and index. 1. Public welfare—United States. 2. United States—Social Policy. 3. Public Welfare—Law and legislation—United States. I. Title. HV95.R435 2007 361.6’80973—dc22 2006023668

ISBN: 978 1 84376 133 4 Printed and bound in Great Britain by MPG Books Ltd, Bodmin, Cornwall.

Contents vi vii viii x xi

List of figures List of tables List of abbreviations Acknowledgments Introduction 1. 2. 3. 4. 5. 6. 7.

8. 9. 10.

Social welfare policy – overview of major issues Comparative social welfare policy – a look at the evolution of welfare policies in Europe Traditional public interest model The public choice perspective Government and its bureaucracy The origins of the new welfare law – a historical overview The genesis of the new welfare law – the Personal Responsibility and Work Opportunity Act of 1996 (PL 104–193) Institutional analysis An econometric analysis of the variables aﬀecting changes in welfare caseloads Conclusions

1 13 35 52 79 93

118 132 168 194 198 211

Bibliography Index

v

Figures 3.1 3.2 3.3 3.4 3.5 3.6 4.1 4.2 5.1 9.1

The Edgeworth box diagram Utility frontier for given total quantities of goods Overall Pareto optimality Partial and overall utility frontiers The public goods problem A positive externality Preference curves The cost of rent seeking The oversupply of a bureau’s output Historical trends in caseloads

vi

37 38 41 41 44 46 54 68 85 170

Tables 8.1 9.1 9.2 9.3 9.4 9.5 9.6 9.7

State historical spending on AFDC and Medicaid programs CEA Models 1 and 2 Rebecca Blank Model Reintsma Basic Model Regression results Regression results Regression results Results from 2SLS Model

vii

166 174 177 184 187 190 191 191

Abbreviations AARP ADA AFDC AFDC-UP AFL-CIO AFSCME AMA APHSA APWA CARE CBN CBO CBPP CC CDF CEA CEEP CHIP CIA DRA EEC EITC EOA ETUC GAO HEW HHS JOBS NAFTA NGA NUL OBRA

American Association for Retired Persons Americans for Democratic Action Aid to Families with Dependent Children Aid to Families with Dependent Children–Unemployed Parent American Federation of Labor and Congress of Industrial Organizations American Federation of State, County and Municipal Employees American Medical Association American Public Human Services Association American Public Welfare Association Charity Aid, Recovery and Empowerment Act Christian Broadcasting Network Congressional Budget Oﬃce Center on Budget and Policy Priorities Christian Coalition Children’s Defense Fund Council of Economic Advisors European Center of Enterprises with Public Participation Children’s Health Initiative Program Central Intelligence Agency Deficit Reduction Act European Economic Community Earned Income Tax Credit Economic Opportunity Act European Trade Union Confederation General Accounting Oﬃce Department of Health, Education and Welfare Department of Health and Human Services Job Opportunities and Basic Skills North American Free Trade Association National Governors Association National Urban League Omnibus Budget Reconciliation Act viii

Abbreviations

OFBCI PAC PRWORA RFP RGA SEA SEM SSI TANF UAC UNICE WIC

Oﬃce of Faith-Based and Community Initiatives Political Action Committees Personal Responsibility and Work Opportunity Reconciliation Act (New Welfare Law) Requests for Proposals Republican Governors Association Single European Act Single European Market Supplemental Security Income Temporary Assistance to Needy Families President’s Council on Urban Aﬀairs Union of Industrial and Employee’s Confederations of Europe Nutritional Program for Women, Infants and Children

ix

Acknowledgments The completion of this book accompanies a debt to a great many people, including family, friends and colleagues. I would especially like to thank Dr Charles Rowley for his unfailing support, and the Locke Institute under whose auspices the book is being published. I would also like to thank Paul Graney for constant encouragement (and occasional arm-twisting) throughout the dissertation process. Jackie Barclift for her generous help with organization and layout, and Hollis Williams without whose moral support and patient and painstaking assistance with references and citations this work might never have seen the light of day. Finally, a great debt is owed to my children whose love and support provide constant encouragement and profound joy.

Dedicated to my parents, Wilfred and Julia

x

Introduction The welfare system in the United States underwent profound changes as a result of the groundbreaking welfare legislation passed in 1996 and known as the New Welfare Law. Oﬃcially titled the Personal Responsibility and Work Opportunities Reconciliation Act (PRWORA), this law both represented a sea-tide of change in ideology throughout the nation, and also served to illustrate the pervasive impact of powerful and eﬀective interest groups whose members or aﬃliates not only inﬂuenced the legislation but in some cases were responsible for formulating the precise language contained therein. This book examines in detail the legislative process which gave rise to PRWORA. It presents two alternative theories to explain this process, the traditional public interest model of government and the public choice model. The public choice perspective takes the rational choice model characteristically applied in the economic ﬁeld and applies this to decision-making in the political arena. Political actors are assumed to behave in a rationally consistent manner that leads to predictable decision making processes and predictable outcomes. In contrast to traditional public sector economics which viewed the government as a tool for achieving allocative or distributive ends, public choice research sees the government and its various constituent members as actors in their own right, with their own rational objectives. Thus, legislators, voters, government oﬃcials, members of the judiciary, special interest groups and other persons engaged in the political process are the subject of public choice analysis, both in regard to their rational choices and the institutions within which these choices are made. The book tests the validity of the alternative models presented by comparing the predicted and actual legislative outcomes. On the basis of a detailed historical analysis of welfare programs and policies in the United States the author explains the two alternative theories and engages in a detailed institutional and statistical analysis to make a convincing argument for the validity of the public choice paradigm. In the light of recent scandals regarding Washington lobbyists, and growing concern regarding the relationships between the various actors in the political process and legislative outcomes which do not seem to be in the public interest, the light cast by this book on these sometimes shadowy proceedings is enormously relevant and particularly timely. xi

1. Social welfare policy – overview of major issues A remarkable fact about welfare policy is that the basic nature of the underlying issues such policy seeks to address has remained unchanged for decades, and even centuries. Problems such as balancing incentives to work with preventing destitution of those incapable of work, of protecting the truly needy, especially children, without providing conditions which lead to the development of an underclass, of providing adequate employment opportunities and compatibility between required skills and available training, and of doing all this in a ﬁscally responsible manner, continue to plague the United States as well as many other nations. Among the major recurring issues in the US are the increase in the number of single parent families, who comprise a signiﬁcant proportion of welfare recipients, the perpetuation of welfare dependency from one generation to the next, and concern about the development of an underclass. Other concerns that occur over time and across national boundaries are the availability of jobs, the ﬁscal implications of government provision of welfare beneﬁts and services, and appropriate policies for immigrants and those with multiple barriers to employment. In addition to these universal issues, speciﬁc issues that have gained prominence in the US as a result of the most recent welfare legislation include the deﬁnition of welfare caseloads and the appropriate role for charitable institutions.

THE CREATION OF AN UNDERCLASS AND THE STRENGTHENING OF WORK INCENTIVES One of the thorniest issues social welfare policy must address is that of appropriate incentives to ensure that those who are able to work do so, without penalizing those who, through no fault of their own, are unable to do so. The rapid growth in welfare caseloads in the US, especially from the late 1980s through 1994, led some to believe that the existing welfare system was creating an underclass of persons who found it more advantageous to rely on government beneﬁts than to seek work. Children growing up in poor, often single-parent households, which were 1

2

The political economy of welfare reform in the United States

dependent on welfare were seen as being more likely to adopt this lifestyle themselves, thus perpetuating the cycle of poverty. As the following pages will show, the solutions proposed have ranged from institutionalizing welfare recipients in workhouses in the nineteenth century, to sanctioning parents by cutting oﬀ beneﬁts and putting children in orphanages in the twentieth century. To address this issue, the US Congress passed a piece of ground-breaking legislation in 1996 which, though oﬃcially titled the Personal Responsibility and Work Opportunity Reconciliation Act, has become popularly known as the ‘New Welfare Law’. This legislation replaced the major existing cash welfare program, Aid to Families with Dependent Children (AFDC) with a new program, Temporary Assistance to Needy Families (TANF), which placed a very strong focus on work incentives, expanding those already in place and introducing much more stringent new ones. These incentives were directed both at the level of the state, including federal bonuses for moving people from welfare to work, and at the level of the individual, by way of time limits and sanctions for those who were able to work and did not do so. Incentives directed at individuals include limits on the length of time a recipient may receive welfare beneﬁts, which is generally two years at any one time, although states may impose a shorter period, and a total of ﬁve years over a lifetime. In addition, recipients are required to engage in work or work related activities such as training programs for a minimum number of hours per week. The incentives for the states include requirements to have a certain percentage of TANF recipients in the workforce or in speciﬁed work activities for a certain number of hours per week (work participation standards), and bonuses for high job participation and retention rates. The increased emphasis on work activities brought about a cultural change in the welfare program. The focus on eligibility and the issuance of welfare checks was replaced by a focus on placing people in jobs. Eligibility workers were turned into caseworkers and a number of states reallocated responsibilities between traditional human service agencies and state welfare agencies. Moreover, some states granted localities more ﬂexibility to implement welfare to work programs. The impact of this institutional and cultural change in welfare programs should not be underestimated. It involved an increase in investment in human capital as welfare caseworkers, or contract agency workers, worked one-on-one with welfare recipients to develop individual development plans, place the individuals in jobs and provide follow-up services to enhance job retention. The controversial nature of some provisions of the TANF legislation was revealed by the failure of the US Congress to reauthorize the legislation.

Social welfare policy – overview of major issues

3

For four years TANF operated under a dozen temporary extensions until ﬁnally on February 8, 2006, President Bush signed the Deﬁcit Reduction Act (DRA) of 2005 [P.L. 109-171S.1932]. Among the most controversial reauthorization proposals were provisions that sought to raise both the number of hours per week that recipients were obliged to work and the work participation standards. Legislation proposed by the House of Representatives (H.R. 240) required a 40 hour work week for all recipients, while a parallel bill in the Senate (S. 667) required single parents to work only 24 hours if they had pre-school age children and 34 hours if their children were of school age. The DRA maintained the current requirement of 20 hours for those with pre-school children and 30 hours for those with older children. Both the House and Senate proposed increasing the participation standards from 50 percent of the caseload to 70 percent, a provision incorporated in the DRA.

SINGLE PARENTS AND A NEW EMPHASIS ON MARRIAGE PROMOTION AND ABSTINENCE EDUCATION A closely related issue which has grown in importance in recent decades is the increase in the proportion of families headed by a single mother on the welfare rolls. In 2003 children living in such families had a poverty rate of 41.7 percent, compared to 8.6 percent for children living in families with two parents (Congressional Research Service (CRS), 2005). While the welfare programs of the 1930s were originally designed to allow needy widows to stay at home and look after their children, by the late 1960s most welfare recipients were divorced, separated or nevermarried mothers. These changes in the demographics of the caseloads, as well as the increases in the overall numbers, coupled with society’s changing views about the participation of women in the workforce, gave rise to a movement to require parents receiving beneﬁts to work. In 1967 Congress required parents with no preschool age children to register for work activities, but states moved slowly to implement this legislation, and there was little impetus as caseloads remained stable from the mid 1970s to the late 1980s. Even the Family Support Act of 1988, which provided states with additional federal funding to promote work participation under the Job Opportunities and Basic Skills (JOBS) program, did not result in substantial increases in work activities. It was only with the rapid increase in caseloads from 1989 to 1994 that states themselves requested waivers from the federal welfare program to

4

The political economy of welfare reform in the United States

introduce even tougher work requirements and time limits, some of which provided the basis of the 1996 federal law. In addition to policies designed to encourage both the custodial and noncustodial parents to work and to provide ﬁnancial support for their children, the new welfare law put unprecedented emphasis on preventive measures such as abstinence education grants to states for educational programs as well as grants to states to promote responsible fatherhood. The Department of Health and Human Services (HHS) has awarded millions of dollars in legislated bonuses to those states receiving the largest reductions in out-of-wedlock births. The states themselves have introduced a wide array of educational programs to reduce out-of-wedlock births and to promote marriage and responsible fatherhood. Responsible fatherhood programs, which include employment, parenting and social skills, can be funded under the TANF block grant, as well as by the Oﬃce of Child Support Enforcement. The DRA authorized up to $50 million per year for responsible fatherhood initiatives. TANF also provided federal matching funds for abstinence education in response to concern about the high rate of teenage pregnancy and nonmarital births. A 1996 study reported that 40 percent of women under age 20 had already had at least one pregnancy and 80 percent of these were unintended (CRS, 2005). Whether the provision of abstinence-only education is more eﬀective than a more comprehensive approach, which would include information on contraceptives and education on the prevention of sexually transmitted diseases, is a highly controversial area. However, funding for abstinence-only education has increased in each of the past ﬁve years. In addition to abstinence education and responsible fatherhood initiatives, marriage promotion has been an integral part of TANF. Marriage promotion activities are similar to those oﬀered by responsible fatherhood programs, but incorporate both parents and involve media campaigns promoting marriage. According to a recent report by the Congressional Research Service, the impact of such programs on reducing child poverty or improving the outcomes for children on an array of developmental issues is not known, and the administration is currently funding research to address this question (Ibid.). The DRA appropriates about $100 million per year for marriage promotion activities. The 1996 welfare law marked the ﬁrst time that there was an explicit focus on promoting marriage and discouraging out-of-wedlock childbearing. However, the appropriate role of government in this arena remains a subject of considerable debate. Some powerful conservative groups have successfully sought to have their views incorporated in the legislation while other very inﬂuential congressmen, such as Senator Max Baucus, the chair

Social welfare policy – overview of major issues

5

of the Senate Finance Committee, has stated that most of his constituents do not believe that government has ‘much business getting into your life’, and he believes that ‘marriage is a personal and private choice’ (Senate Finance Committee, 2002). Other measures directed at addressing problems related to single parents on welfare range from enhanced child support enforcement, to the provision of work support services such as child care.

CHILD SUPPORT Recent years have seen much stricter child support enforcement, facilitated by the introduction of a national registry of persons with delinquent accounts. Employers are required to register all employees and wages are garnished from delinquent parents in order to recoup child support payments. The Child Support Enforcement program has almost doubled collections since welfare reform was passed, and an increasing amount of these funds are transferred to the custodial parents (Ibid., p. 6). These funds have been collected in part due to the implementation of mechanisms such as withholding federal tax refunds, and the Passport Denial Program (which withholds passports from delinquent parents). Welfare recipients are obliged to establish paternity as a requirement of beneﬁts, with some exceptions such as domestic violence cases, and parents who are in arrears can be denied professional and other licenses and have their debts garnished from wages. In the ﬁrst ﬁve years of TANF implementation, the number of paternities established increased by almost 50 percent. To receive TANF, parents must assign child support rights to the state. In some states, the internet has been used to publish the identity and arrears of nonpaying parents (Ibid.). Currently, a portion of collections are held by the state to cover the costs of welfare programs. A higher percentage would be passed through to families under the DRA legislation. Given that child support constitutes on average 17 percent of total income for those families that receive it, and 30 percent of income for families below the poverty line who do not receive TANF, this could make a signiﬁcant contribution to raising recipients out of poverty. In addition to assisting single parents through stricter child support enforcement, there has been increased emphasis on work support services such as child care as the work requirements for welfare mothers have increased. In order to facilitate single parents working, the funds allocated to states for the provision of child care have seen enormous increases.

6

The political economy of welfare reform in the United States

CHILD CARE Concerns that single mothers were becoming an underclass with negative repercussions for their children and society at large were among the factors that led to a strong emphasis on work requirements for recipients of TANF beneﬁts. Under the 1996 welfare reform, parents whose children are over one year old are generally required to work at least 20 hours a week in return for beneﬁts. The increased work requirements led to huge increases in the need for child care. To deal with this issue, the TANF legislation introduced signiﬁcant increases in the level of federal funding for child care. The Child Care and Development Fund, has provided $4.8 billion per year since 1996 for child care programs for low income families. Proposals for TANF reauthorization contained both substantial increases in work requirements and increased funds for child care. However, the degree to which funding should be increased was a source of major controversy with the Senate bill (S. 667) providing for an increase of $6 billion over ﬁve years while the House bill (H.R. 240) provided only $1 billion over the same period. The DRA authorized the latter. Justiﬁcations for increased funding included both the increased work requirements and increased costs due to inﬂation. Recent estimates by the Congressional Budget Oﬃce calculate that the cost of inﬂation will be $4.8 billion over the next ﬁve years, while the costs due to increased work requirements would be $4.1 billion under the House proposals and $.9 billion under the Senate proposals (CRS, 2005).

WELFARE RECIPIENTS WITH MULTIPLE BARRIERS TO EMPLOYMENT A further group of welfare recipients that is especially challenging is that of the population that has severe and often multiple barriers to employment, including mental and physical disability, addictions to drugs or alcohol, or those living in situations of domestic violence. There is some evidence that states have reclassiﬁed some persons with disabilities to make them eligible for the federally funded Supplemental Security Income (SSI), thus removing them from the rolls, and from state expense. Programs and treatments for all of these categories are far more expensive than the provision of the standard cash beneﬁt check, at least in the short term and possibly in the long term. Moreover, as caseloads fall, the proportion of recipients in these categories increases, and it becomes more diﬃcult for states to achieve reductions in caseloads.

Social welfare policy – overview of major issues

7

EMPLOYMENT The issue of employment became particularly critical with the onset of recession in the US economy. When the new welfare law was passed in 1996 the economy was in a period of expansion, and employers were concerned about the shortage of workers. Moreover, to promote the transition from welfare to work and encourage job retention, the 1997 Balanced Budget Act created a $3 billion welfare-to-work grant program for FY1998– FY1999, administered by the Secretary of Labor. As of December 2000, however, $1.6 billion remained unspent, and, at the request of the President, Congress extended the deadline from three years to ﬁve years from the date of award of the funds. At the end of FY2002 almost 20 percent of the funds were still unspent, and in FY2004 Congress rescinded remaining unspent funds. Congress also established a new tax credit and extended an existing one, both of which were available to employers who hired TANF recipients. The demand for workers, in combination with incentives for companies to hire welfare recipients, government funded training programs, and generous work supports contributed to the major declines in caseloads. The downturn in the economy in 2001 and the increase in the employment rate, combined with the fact that those left on the welfare rolls were the least qualiﬁed, limited the gains that could be made in continuing to place welfare recipients in jobs. In fact, between June 2001 and June 2005, the number of welfare recipients in 23 states actually increased, although the national caseload continued to decline. Clearly, continued success in reducing the caseloads, and in lifting people out of poverty and avoiding the development of an underclass depends on appropriate incentives on the one hand, and on the availability of jobs and the eﬀectiveness of relocation and training opportunities on the other.

FISCAL IMPLICATIONS Under AFDC the federal government provided states with virtually unlimited matching funds. The 1996 welfare law replaced matching funds with a ﬁxed block grant of $16.5 billion to states, based on funding levels prior to the passage of the law. The law contains no provisions to adjust for inﬂation or for population increases (although the original block grant formula did make allowance for a higher allocation to states with more rapid population growth or lower than average beneﬁts). With the rapid decline in the caseloads from 1994 to 1999, money from reduced cash welfare payments was diverted to providing increased work support services, such as child

8

The political economy of welfare reform in the United States

care, transportation assistance, and intensive caseworker services. States were required to maintain a level of funding of 75 to 80 percent of their previous outlays and the combined federal and state funding could be used for a broad range of services designed to fulﬁll the major TANF objectives of ending welfare dependence, reducing out-of-wedlock births, assisting families so that children might live in their homes or with relatives, and promoting stable marriages. The cost of these work supports could far exceed the cost of a welfare check, however, and when state revenues began to falter, with the recession of 2001, states found themselves facing budget crises, and faced with very diﬃcult political decisions. As states began to make cutbacks in spending, work support programs such as child care and job training were pared back. The recipients of work support services include many low income working parents, who might or might not also receive a cash beneﬁt, and who rely on services such as subsidized child care to stay in the workforce. The extent of the cuts and the impact on the former or present recipients is not well documented. However, the extension of work requirements under current proposals is likely to increase demand for work support services such as child care. Moreover, the strategy of transferring welfare recipients who reach federal time limits into programs funded only by state money has increased the pressure on state budgets, a process likely to continue for some time as more recipients reach federal time limits and lose eligibility for TANF.

CHARITABLE CHOICE PROVISIONS The 1996 welfare law allows states to administer and provide services under TANF through contracts with charitable, religious or private organizations, and stipulates that the states must allow religious organizations to provide federally funded services on the same basis as any other nongovernmental provider. Moreover, this must be done in a way that neither impairs the religious character of the organization nor diminishes the religious freedom of recipients. Such provisions have come to be known as charitable choice. The origins of charitable choice can be traced to a bill sponsored by Senator Ashcroft (S. 842), the language from which was adopted in a Finance Committee Bill reported in June 1995, which proposed to replace AFDC with a block grant. The Finance Committee Bill included two sentences which provided that religious organizations that participated in the new program were to retain their independence from government and could not deny aid to needy families on the basis of religious belief or refusal to participate in religious activities. The language was subsequently extended

Social welfare policy – overview of major issues

9

to give states an option to administer and provide services through religious organizations and by means of vouchers and allowing religious organizations the right to consider religion in their hiring practices. This language was adopted almost in its entirety in the 1996 welfare law which ﬁrst enacted charitable choice. It applied to both TANF and other income support programs, and was extended to other programs in 1998 and in 2000. In January 2001, President Bush sought to further extend the role of religious organizations through the introduction of a faith-based initiative and he issued executive orders that established an Oﬃce of Faith-Based and Community Initiatives (OFBCI) in the White House and directed ﬁve Cabinet departments, Education, Health and Human Services, Justice, Labor and Housing and Urban Development, to set up similar oﬃces (the Department of Agriculture and the Agency for International Development were added in 2002). The faith-based initiative incorporated both the expansion of existing charitable choice provisions and tax incentives to promote charitable donations. The President’s initiative became a source of major controversy in the Congress, however. The House voted in 2001 to extend charitable choice rules to numerous new programs, as the President urged, but the Senate refused. H.R.7 had aroused controversy, especially over religious discrimination in employment in organizations receiving federal funds, and possible ‘voucherization’ of social services. Since vouchers can be used by the recipient organization for any purposes, including religious activities, this was seen by some as a major breach of the principle of separation of church and state. A bipartisan bill introduced into the Senate (S. 1924, the Charity Aid, Recovery and Empowerment Act – CARE) removed some of the most controversial provisions of the House bill but retained the equal treatment provisions that stipulated religious organizations could not be required to remove religious art, icons or other symbols and could discriminate on the basis of religion in hiring practices. These equal treatment provisions were dropped even before the Finance committee markup, however, and the bill passed by the Senate in April 2003 contained no equal treatment provisions. While the Congress did not pass charitable choice legislation in 2002, a number of grants were made to promote charitable choice, including a $17.5 million grant from the department of Labor in July 2002 ‘to link faith-based and grassroots community organizations’ to the nation’s OneStop Career system under the Workforce Investment Act, and matching grants from HHS of about $25 million to provide technical assistance, start-up and operational costs to faith- and community-based organizations. The HHS grant was from the Compassionate Capital fund, a $30 million fund created under the President’s faith-based initiative.

10

The political economy of welfare reform in the United States

The trend to expand the role of charitable organizations was continued by President Bush in December 2002, when he issued an Executive Order (EO 13729) to extend charitable choice principles to virtually all social service programs for low income people (CRS, 2003). Between January 4 and July 16, 2004, the departments of Agriculture, Education, Health and Human Services, Housing and Urban Development, Justice, Labor and the Veteran’s Administration issued regulations to incorporate charitable choice provisions into their programs. Opposition to charitable choice has brought together a coalition of religious and secular groups who, for diﬀerent reasons, want to maintain separation of church and state – the former to protect their independence and sense of mission and the latter to guard against use of public funds for religious activities such as proselytizing. As part of this controversy, several court suits have been ﬁled that challenge the constitutionality of TANF charitable choice provisions. In July 2000, a suit charged that a job training and placement program funded by the Texas Department of Human Services and operated by the Jobs Partnership of Washington County, was ‘permeated’ by Protestant evangelical Christianity in violation of both the state and federal constitution. The suit was dismissed in February 2001, but only after the state had discontinued the program. Another suit charged that a job placement and support services program for drug addicts in Milwaukee, Wisconsin, violated the state and federal constitutions by giving welfare-to-work funds directly to a ‘perversely sectarian’ organization (Faith Works) and using the funds to indoctrinate clients in the Christian faith. A federal jury on January 8, 2002 ordered Wisconsin to cease this direct funding as unconstitutional, but said the ruling did not deal with the constitutionality of the 1996 charitable choice law, which does not authorize direct funding of religious activities. The controversy over charitable choice reﬂects, in part, a deep-rooted concern on the part of opponents regarding the ideological agenda of the Administration and allies in Congress. The stated purposes of the Charitable Choice Act (H.R. 7, Title II), discussed earlier, are as follows: ● ●

●

To provide assistance to needy individuals in the most eﬀective and eﬃcient manner; To supplement the Nation’s social service capacity by facilitating new and expanded eﬀorts by religious and other community organizations in the administration and distribution of government assistance; To prohibit discrimination against religious organizations on the basis of religion in the administration and distribution of government assistance;

Social welfare policy – overview of major issues ●

●

11

To allow religious organizations to assist in the administration and distribution of assistance without impairing their religious character, and To protect the religious freedom of those in need who are eligible for government aid, including expanding the possibility of their choosing to receive services from a religious organization.

Clearly, the ideology of charitable choice not only is in line with mainstream Republican ideology of small government, but also can potentially greatly increase the role and funding for organizations of the religious right, which are among the strongest supporters of the current administration.

DEFINITION OF CASELOAD The dramatic decline in the national TANF caseload from 1994 to 1999 was touted by many as a sign of the success of the 1996 welfare legislation. This decline has since slowed to a near halt, and following the 2001 recession, caseloads rose in about half the states from 2001 to 2003. Care must be taken in interpreting this data, however. The deﬁnition of what comprises a caseload has become increasingly murky as the services provided to welfare recipients are also provided to the working poor who are not receiving cash beneﬁts. Moreover, states are increasingly placing individuals and families ineligible for the federal TANF program into separate state-funded programs, with the result that these persons no longer appear on the oﬃcial caseload ﬁgures published by HHS. Among these groups are recipients that have exceeded the two-year limit, or the lifetime ﬁve year limit, and groups which are ineligible under federal rules, such as two parent families and legal immigrants who have been in the country for less than ﬁve years. The practice of having a parallel program funded with only state dollars has resulted in a substantial underestimate of the numbers of families receiving cash welfare beneﬁts. General Accounting Office (GAO) has reported that in four states the number of families served by state-funded programs in 2002 was from 10 to 30 percent of the total caseload, and CRS reported that in New York alone, 40,000 families were moved to a state funded program after they reached their ﬁve year lifetime limit for federal funds (GAO, 2002). Given that the oﬃcial national caseload declined by only 107,000 (from 2,170,586 to 2,063,505) from June 2001 to June 2005 it seems quite possible that the actual number of persons receiving cash welfare has increased during that period. This supposition is supported by data from the Food Stamp Program, where the caseload reached a ﬁve year high of 20.5 million persons in December 2002.

12

The political economy of welfare reform in the United States

The accuracy of using caseload data to estimate the number of persons dependent on the government is also limited by the fact that the states have much greater ﬂexibility to provide child care, workforce development and other services to cash recipients and non-recipients alike. The provision of subsidized child care and other services may well be what keeps a family oﬀ the oﬃcial welfare rolls, but that family is, nonetheless, still dependent on federal funding. In a recent report by GAO, it was estimated that at least 46 percent more families than are counted in the reported TANF caseload are receiving services funded, at least in part, with TANF funds (Ibid.). Moreover, the amount of TANF funding that is allocated to cash payments is only about 40 percent of the total funding for the TANF program. Clearly, tracking and analyzing the impact of the 1996 welfare legislation will be far more complicated than merely tracking caseloads of those receiving cash beneﬁt checks.

CONCLUSION There will always be a ﬁne line between providing for the truly needy and preventing abuse of the system by those who could work. And the balance between these two will be reﬂected in policies that are the result of a combination of the prevailing ideological beliefs, the strength of the various interest groups that inﬂuence policies, and the economic conditions prevailing in the nation and the locality. While the nature of the problems addressed by welfare policy has changed little, the solutions have evolved from the ‘whipping and branding’ of ‘sturdy vagabonds’ in the nineteenth century to the withholding of cash beneﬁts from those who do not comply with program rules in the twentieth century. Societal norms, as well as the far larger number of persons who at some time are on the welfare rolls have contributed to the deﬁnition of what is acceptable policy. Within broadly deﬁned limits, however, the speciﬁcs of welfare policies are increasingly deﬁned by those groups who can most eﬀectively inﬂuence the legislative process. As the following chapters will show, this is brought about by the participation of interest groups in the market for legislation, and those groups with the most powerful constituencies, and which can deliver the largest blocks of votes, are most likely to be rewarded with policies favorable to their interests.

2. Comparative social welfare policy – a look at the evolution of welfare policies in Europe Comparative analysis of social policies in diﬀerent nations is fraught with diﬃculties, not least of which is the problem of terminologies and deﬁnitions which vary widely among countries. The very notion of ‘welfare’, for example, is quite diﬀerent in Europe than in the US. In the latter, the term often applies very narrowly to the program that provides cash assistance to needy families with children. The most common European equivalent is ‘social assistance’. In the US the meaning of the term may be extended to cover the food stamp and Medicaid programs which incorporate primarily, but not exclusively, the same target population. It thus distinguishes means tested programs and beneﬁts. In Europe this distinction is less sharp and countries such as Sweden and the UK have a range of social programs such as health care and retirement beneﬁts which may be available to the general population at subsidized cost and regardless of economic need. One consequence of the broader deﬁnition of welfare in many European nations is that a broader segment of the population derives beneﬁts from welfare programs, and thus popular support for such programs is often greater. The philosophical beliefs upon which welfare programs are founded also diﬀer substantially between the two continents, and indeed among the European nations. Such beliefs reﬂect the diﬀerent political and socioeconomic contexts in which the welfare systems were founded and developed. In the US there has been a greater emphasis on individual responsibility, and this has been particularly apparent in the public debate on welfare throughout the 1980s and 1990s, when fears that the welfare system in the US provided disincentives to work and marry, and was leading to the development of an underclass were widely expressed by academics, politicians and in the media. European politicians have traditionally put more emphasis on rights and entitlements. In Britain, for example, there is currently no work requirement for single parents until their children are 16. These philosophical diﬀerences and the related socio-economic and political factors have historically contributed to the development of very 13

14

The political economy of welfare reform in the United States

diverse social policies across nations. What is interesting to note, however, is that despite these historical diﬀerences, in recent decades there has been a tendency to greater convergence in social policies in the US and Europe as nations have recognized a need for reform and have introduced major changes in their welfare programs. Underlying these reforms has been both the ideological concerns regarding the negative behavioral incentives in the programs and the pragmatic concerns of the ﬁscal costs of seemingly ever expanding programs, a factor which became especially acute in the recessionary periods of the 1970s and 1980s. The pace of welfare reform among nations has varied considerably. In Europe, the Nordic countries, Germany and the UK are at the forefront of reform while countries in southern Europe have generally introduced few reforms. The following discussion examines some of the major changes taking place in the UK, Sweden and Germany, explores the impetus behind them, including the inﬂuence of ideology and interest groups, and compares the processes and outcomes of reforms in these countries with those of the US.

SWEDEN With healthy economic growth, low unemployment, one of the highest living standards in the world, and a very expansive state welfare system, many advocates of the welfare state long considered Sweden as a model to be emulated. Since the 1970s, Sweden has spent a higher proportion of its national income on welfare beneﬁts and services than any other capitalist state, and highly redistributive welfare programs have resulted in one of the lowest levels of income inequality when taxes and transfers are included (Cochrane et al., 2001). However, critics claim that this has been achieved at the expense of individual, family and entrepreneurial freedom. Moreover, since the 1990s, the faltering economy, competitive pressures arising from globalization and higher levels of unemployment have led to a reexamination of the economic costs of the philosophy underlying the Swedish welfare state. The development of social welfare policies in Sweden has followed a far steadier path than in the US and many other European nations, perhaps in part reﬂecting the homogenous nature of the society and the absence of that country’s engagement in civil uprisings or warfare. Swedish social policy is also characterized by a long tradition of liberalism (as opposed to conservatism), the origins of which lie deep in the country’s political and socioeconomic history. The introduction of divorce laws in 1920, and the widespread use of birth control contributed to a decline in birth rates that

Comparative social welfare policy

15

gave rise to concern in the interwar years as Sweden’s birth rate fell to the lowest in Europe (Myrdal, 1945).1 According to Norman Ginsburg, the predominant inﬂuence on Swedish welfare policy from the 1930s through the 1960s was ‘concern with the declining population size, linked to fears of the extinction of the nation’ (Cochrane et al., 2001, p. 213). In the 1930s, Sweden’s birth rate was the lowest in Europe and a number of measures were adopted to support poor families, including free maternity care, child tax allowances, housing, marriage loans and guaranteed income for some single mothers, and in 1948, universal child beneﬁt was introduced. At the same time birth control was widely available ‘to ensure every child would be a wanted child’ (Ibid.). In tandem with the liberal social policies, the Social Democratic Party (SAP) adopted Keynesian counter-cyclical economic policies in the 1930s, including public works programs to provide employment and food price subsidies (Ginsburg, 1992). In a move which had far-reaching consequences in subsequent decades, the SAP forged an agreement with capital and labor which guaranteed certain economic freedoms to the former, including protection from nationalization, in return for the support of capital in other policy areas, including welfare policies (Ibid.). This relationship between government, business and unions has strongly inﬂuenced the development of social policies over the latter half of the twentieth century and the beginning of the twenty-ﬁrst. In the 1940s and 1950s the principles of the Social Democratic ideology were reﬂected in the policies of maintaining full employment and providing universal ﬂat rate beneﬁts to ensure a universal minimum income level. The social value placed on limiting inequality was further reﬂected in the trade union wages policy, which strove for minimizing pay diﬀerentials, and relating pay to worker’s performance rather than the proﬁtability of capital (Ginsburg, 1992). A wages policy was adopted by the trade unions in their national centralized bargaining with employers, based on the principle of equal pay for equal work, and aimed at removing wage diﬀerentials based on diﬀerences in proﬁts or wage paying ability (Bjorklund and Holmlund, 1990).2 With more than 80 percent of the labor force organized in trade unions, such bargains played a major role in employment policy. The state, for its part, established the State Labour Market Board in the 1940s and this presided over a range of programs aimed at retraining and redeploying workers. All vacancies and layoﬀs had to be reported to this board. The unemployed were encouraged to migrate and, until being cutback in the late 1980s, relocation grants were oﬀered as inducements. Other measures taken in this period included the replacement in 1948 of the children’s tax credit by universal children’s allowances linked to inﬂation, and the introduction in the 1950s of

16

The political economy of welfare reform in the United States

statutory provision of birth control and rights to maternity leave (Ginsburg, 2001). The ideological impetus for welfare reforms that was initiated in the 1930s gained renewed vigor in the 1960s, but the inﬂuences behind these reforms diﬀered from the earlier period. While the 1930s had seen broadbased measures addressed at the general population, the 1960s saw increasing pressure from speciﬁc interest groups, including feminists, students, and others seeking greater equality of outcome. This ideology is reﬂected in policy statements of the SAP, such as that adopted in 1969 which stated that: an equalization of living conditions is a means for attaining altered human relations, a better social climate . . . Those who are left behind, with insuﬃcient resources to contribute to the common good, represent an obstacle to both eﬃciency and desirable social change . . . In the Social Democratic conception, there is no reason that extreme diﬀerences in endowments, in health, in intellect, or in work capacity should lead to an assignment of standards and life chances that diﬀerentiate some from others (Heclo and Madsen, 1987, pp. 174–5).3

The report highlighted not only issues such as poverty and unemployment, but speciﬁcally addressed occupational and gender wage diﬀerentials as areas of particular concern, and proposed targeting groups which were perceived to be disadvantaged. According to Ginsburg, this ‘equality movement’ had a considerable impact on SAP ideology and policy-making in the 1970s, above all perhaps with the introduction of a sharply progressive income tax in 1971, but also in a great expansion in welfare beneﬁts and services (Ginsburg, 1992, p. 33). Policies implemented at this time included a greatly expanded day care system, extended statutory parental leave (the basic entitlement is 12 months leave at 90 percent of gross earnings funded by social insurance, a period which can be divided between the parents), and tax reforms which encouraged women to take up part time employment (Scott, 1982). In contrast to many other European states, Sweden did not make major cutbacks in investments in the welfare state in response to the recession of the 1970s. The rapid expansion of beneﬁts in the 1960s and 1970s solidiﬁed middle class commitment to the welfare state and rather than cutting back in the 1980s, Sweden consolidated its welfare programs and even expanded its child care programs (Ibid.). The continuing strength of the trade union wing of the labor movement was instrumental in this outcome, but the power of unions was curtailed in the 1980s. Hoping to have more control over private industry and commerce, the unions wanted to establish trade union controlled ‘wage-earner funds’ out of a substantial proﬁts tax (Ginsburg, 1992, p. 33). These funds

Comparative social welfare policy

17

would be invested in major companies, thus theoretically building up leverage for the union shareholders. The SAP was in opposition for the latter part of the 1970s, however, and with the recession of the early 1980s, labor unions were considerably weakened. The legislation that ultimately passed was far weaker than union proposals with the wage-funds being mainly derived out of workers contributions rather than proﬁts taxes (Ibid.). In the 1980s, the movement for equality of outcome became less inﬂuential. A weakening economy, the struggle of Swedish multinationals to remain globally competitive, and the ﬁscal crisis of the state was accompanied by the increasing inﬂuence in the SAP of politicians favoring free market reforms and the development of the so-called ‘New Right’. The inﬂuence of the New Right was reﬂected in the 1980s and early 1990s by cuts in public expenditure, and major tax reforms, and Sweden faced a major economic crisis in the early 1990s. Between 1990 and 1993 GDP declined by 5 percent and unemployment quadrupled, remaining at unprecedented levels for the rest of the decade (Ginsburg, 2001). The 1990s saw cutbacks in public expenditures, increased charges for medical and other public services, and restrictions on pensions and unemployment beneﬁts, and when the economy recovered in the late 1990s cuts were made in both income and corporate taxes. When ﬁscal pressures and an increasingly conservative society gave impetus to the movement to cut social assistance, the reforms in Sweden were considerably weaker than in the US, reﬂecting the greater strength of interest groups such as unions in Sweden, as well as philosophical diﬀerences such as the greater emphasis on economic individualism and free market ideology in the US. The diﬀerences between the US and Sweden in social policies and in ideology is reﬂected in the characteristics of the social assistance caseload. An increase in the number of immigrants, as well as higher unemployment contributed to this rise in the caseload, and, according to Ginsburg (2001) the costs of social beneﬁts increased fourfold, resulting in cutbacks in the level of beneﬁts, which are determined at the local level in Sweden. While most claimants in Sweden in the early years were elderly or poor families, by the 1980s, single, young and often childless persons made up a large proportion of recipients (Ginsburg, 2001). In the US, the welfare caseload is predominantly composed of single mothers and their children. In contrast to the US where over 90 percent of adult recipients are women, in Sweden the numbers of men and women are virtually equal. Public support for single parents and their children has been comparatively generous in Sweden and the proportion of single-parent families in Sweden is higher than most other European Union countries, although less than in the US. However, in Sweden, almost 90 percent of these parents are employed. Care

18

The political economy of welfare reform in the United States

must be taken when looking at unemployment ﬁgures in Sweden, however. Continuous growth of the public sector employment contributed to an average unemployment rate of about 2 percent for most of the post-war period, rising to around 10 percent in the 1990s, but students looking for work and people participating in labor market programs are not included in this ﬁgure, which would otherwise be about 4 percent higher (Ginsburg, 2001). Women’s assimilation into the labour force has been aided by a combination of publicly-funded day care, statutory rights to paid parental leave, and active labour market policies such as training schemes and generous paid leave policies. On any given day, 20 percent of women are on paid leave of some kind, reaching 30 percent in the public sector (Ginsburg, 1992). By the late 1990s over 80 percent of adult Swedish women below retirement age were economically active in paid employment or training, far higher than in most other European states (Cochrane et al., 2001). The high levels of participation by women in the labor force as well as their eligibility in their own right to pensions, sickness, unemployment and other work related beneﬁts means that fewer are dependent on social assistance than would otherwise be the case. The number of single parent families which rose to about 20 percent in the 1980s declined to about 15 percent in the early 1990s, a ﬁgure below that of both the US and the UK (Ginsburg, 2001). The level of poverty among single mother households in Sweden was less than 4 percent in the 1990s, compared with 60 percent in the US, a ﬁgure attributable to a combination of a slightly higher employment level, much higher wages, and a greater proportion of income coming from beneﬁts and paternal maintenance in Sweden (Ibid.). The low level of poverty among female headed households is all the more striking when one considers that by the late 1980s Sweden had one of the highest birthrates among the OECD countries. It is interesting to note that single mothers are not perceived as a problem group in the same way that they are in the US. As Hobson and Takahashi (1997) note, ‘. . . a discourse on the decline of the traditional family has little resonance in Sweden in contrast to the situation in Britain and the US’ (Ginsburg, 2001, p. 215). In fact, single mothers have always been expected to support themselves and their children and to this end preschool childcare services have been available in Sweden on a scale only beginning to be approached in the US as welfare reform requires single parents to join the workforce. The ability of single mothers to work is also greatly facilitated by generous leave allowances, including 50 days pregnancy leave, 60 days leave with beneﬁts per child per year to care for a sick child, and 330 days leave at 80 percent of beneﬁts up to the child’s eighth birthday (Ibid.).

Comparative social welfare policy

19

A further notable diﬀerence is that a signiﬁcant amount of the growth of the caseload in Sweden in recent decades has been attributed to an increase in the number of refugees and immigrants (Ginsburg, 1992). In the US restrictions on eligibility limit the growth of this category of claimants, although such policies have been highly controversial and eligibility rules have changed several times in the past decade. The size and composition of the caseload in Sweden is also aﬀected by labor market policies such as training schemes, public works and overseas relief projects, and programs for the disabled provided by the Labour Market Board, all of which combine to reduce the oﬃcial unemployment rate and the social assistance rolls. One result of these programs is that only 8 percent of the unemployed in Sweden were unemployed for over a year while the average in the European community is over 50 percent (Ginsburg, 1992). Over 80 percent of the labor force is covered by unemployment insurance beneﬁt (UIB) which pays beneﬁts for up to 12 months, and despite a cut in the early 1990s, UIB still covers about 80 percent of previous income (Ibid.). Since 1996 UIB can be withdrawn for 60 days if a claimant refuses a job or training oﬀer (Ginsburg, 2001). UIB is ﬁnanced by a combination of government, employer and trade union contributions and is administered by voluntary societies under the control of the trade unions (Ibid.). Only Denmark and Luxemburg have more generous beneﬁts for the insured unemployed, and the UIB societies are seen by the unions as an important means of recruiting and keeping members (Ginsburg, 1992). In Sweden, the strength of labor unions, the ideology of equality and the inﬂuence of liberal academics such as Gunnar Myrdal, contributed to an expansionary welfare state until the 1980s. Although ﬁscal pressures and a move to a more conservative public opinion and government led to cutbacks in the 1980s and 1990s, the inﬂuence of interest groups and ideology resulted in these cutbacks being less severe than in the US and there remains a relatively strong infrastructure of beneﬁts. Moreover, while US policies have incorporated a range of sanctions aimed at reducing caseloads, Sweden has tended to provide more work support beneﬁts to encourage employment of single parents.

GREAT BRITAIN The British welfare system, like that of Sweden is characterized by an extensive network of beneﬁts and services, from housing assistance and universal health services to maternal and child welfare beneﬁts, in addition to cash allowances. While some beneﬁts, such as free school meals, are

20

The political economy of welfare reform in the United States

directed only to the poor, a signiﬁcant number, including health services, are provided regardless of income level. The beneﬁts of a wide range of social services provided by the government have thus been widely dispersed over the population. Until recently, the popular support generated by this system, in conjunction with support from trade unions, limited reforms of the British welfare system. In the past two decades, however, the forces for change gained increasing momentum and Britain has become one of the leaders of welfare reform in Europe. The term ‘welfare state’ entered the English Dictionary in 1955 and was deﬁned as ‘a polity so organized that every member of the community is assured of his due maintenance and the most advantageous conditions possible for all’ (Lund, 2002, p. 107). The perspective that this deﬁnition reﬂects is enshrined in the famous report written by Sir William Beveridge in 1942, and named after the author. The Beveridge Report made three assumptions: the state would promote full employment, provide children’s allowances, and deliver a comprehensive national health service. Beveridge advocated social insurance for basic needs, national assistance for special cases, and voluntary insurance for additions to the basic provision (Beveridge, 1942). In the 1940s, the Labour Government implemented most of Beveridge’s suggestions. Among the major legislation was the Education Act (1944) which introduced school milk and meals depending on family income, the Family Allowances Act (1945) which introduced child allowances, and the National Assistance Act (1948) which set up a National Assistance Board to aid persons ‘without resources to meet their needs’ (Lund, 2002, p. 115). Assistance included both a cash beneﬁt and rent, adding up to more than the national insurance payment. When the Conservatives came to power in the 1950s, they recognized the popularity of the welfare state and did not attempt to change it by legislation, but rather chipped away at parts of it and relied on economic growth to foster private provision (Ibid.). In fact, the Conservative’s Party’s 1947 Industrial Charter had promised policies to generate full employment in return for cooperation from the unions and throughout the Conservative’s term of oﬃce in the 1950s and early 1960s, unemployment was never more than 2.6 percent and fell to a low of 1.2 percent in 1955 (Ibid.). The Conservative Government followed a gradualist reform policy of reducing taxes and introducing a wide range of tax allowances including allowances for retirement provision, life assurance, marriage, children and mortgages. The return of Labour in 1964 was followed by an increase in family allowances and an increase in beneﬁts for pensioners and the unemployed under pension age (Ibid.). These were paid for by, inter alia, increasing

Comparative social welfare policy

21

income tax rates, reducing child tax credits from those paying income tax, and introducing new taxes, including capital gains (Ibid.). Some of the gains in family allowances were lost in the early 1970s under Conservative rule, but a new beneﬁt was introduced, the family income supplement for families where the head of household was working, but with a low income (Ibid.). And when Labour returned to power in 1974, it promised big increases in social spending, and higher taxes, reﬂecting the philosophy propounded in the Labour Manifesto of 1974, which noted that ‘taxation must be used to achieve a major redistribution of both wealth and income’ (Labour Party, 1974).4 A number of egalitarian social security measures were introduced in the succeeding years, including the replacement of child tax allowances with child beneﬁts. Recessionary forces in the economy put a sharp brake on Labour’s expansionist aims for the welfare state, however. By 1976, inﬂation was at 24 percent, fuelled by OPEC price hikes, unemployment was high and the value of the pound was falling fast. The Government had to agree to cuts in public spending in return for an IMF loan. While concern about unemployment led to several new measures, including the Job Creation Programme, Work Experience Programme and Youth Opportunities Programme, the Labour government was required to resort to overall cutbacks in social programs, and engaged in direct intervention in the labor market in the form of prices and incomes controls (Lund, 2002). Subsidies to employers to hire and retain workers were introduced in the mid 1970s, rescinded in the 1980s under the Conservative government, and brought back under New Labour’s ‘New Deal’ which subsidized employers speciﬁcally to hire the long term unemployed (Ibid.). The economic crisis of the 1970s and the failure of Keynesian policies to ensure economic stability led to increased concern about public spending and the deleterious eﬀect on the economy of a high level of taxation and government interference in the workings of the market. Thus, conservatives like Keith Joseph, a leader of the new right in the Conservative party, speaking in 1977, noted that ‘unﬂinching control of the money supply though essential is not enough’ and advocated control of social spending and reform of social policy (Lund, 2002, p. 163). Although Joseph fell from grace after remarks about the threat to the ‘human stock’ of children born to ‘mothers least ﬁtted to bring children into the world and bring them up’, his convictions were also held by many others, including Margaret Thatcher, who came to power in 1979 and proceeded to make broad reaching reforms to the welfare system throughout the 1980s (Ibid.). These reforms were carried on by John Major who succeeded her in 1990 and held power until the landslide victory of Tony Blair’s Labour party in May 1997. The ideology of the new right held that the welfare state, rather than being

22

The political economy of welfare reform in the United States

an economic and social investment, was, in fact, socially and economically destructive (Clarke et al., 2001). It was felt that there was a need to break up the monopoly of the state over the provision of welfare services and increase the role of the private and voluntary sectors (Ibid.). The analogy with the evolution of the welfare policy in the US is clear. One area of particular concern to Prime Minister Thatcher as well as to her counterparts in the US, was the issue of child support. As in the US and Sweden, throughout the 1970s and 1980s the number of single parents in the UK had been increasing dramatically and Margaret Thatcher, inﬂuenced by the writings of Charles Murray in the US, turned her attention to reforming the system of child support and requiring the non-custodial parent to support the child (Lund, 2002). A similar reform was under way in the US. Other measures proposed by the Thatcher government included increasing the role of the private sector in providing social services, tax cuts, the replacement of government grants with loans, and reducing the power of the trade unions. The desire to increase the role of the private sector was illustrated in a 1985 White Paper which recommended, inter alia, shifting responsibility for maternity allowance from the state to the employer, replacing grants for special circumstances by loans, and changing the name of Family Income Supplement to Family Credit and paying it through the wage packet rather than through the government (Ibid.). Most of these proposals were implemented in the 1986 Social Security Act. Perhaps one of the most fundamental changes was the shift away from universal beneﬁts to means tested beneﬁts. This targeting of beneﬁts serves to not only reduce spending levels, but also reduce the constituency for the programs aﬀected. When the Labour government returned to power in 1997, its policies had undergone radical revision from those propounded in earlier election campaigns. After suﬀering defeat in these campaigns, Labour had been moving steadily in favor of market processes, and when Tony Blair was elected leader in 1994 he moved quickly to reinvent his party which became known as New Labour. The changing ideology of labour was reﬂected in a 1994 report produced by the Commission on Social Justice, appointed by former labour leader John Smith, which emphasized a ‘third way’ and the development of an ‘Investor’s Britain’, distinct from the ‘Deregulators Britain’ of free market Conservative and the ‘Leveller’s Britain’ of the ‘Old Left’ (Ibid., p. 188). As one commentator noted, ‘New Labour elevated pragmatism into a political creed. Its guiding principle in assessing a course of action was potential for improvement, regardless of the ideological origins of the program’ (Ibid., p. 189). In a major break from the past, New Labour repudiated equality of outcome and set as a policy goal ‘equal opportunity for all and special privileges for none’ (Mandelson and Liddle, 1996).5

Comparative social welfare policy

23

The new approach had much in common with what was happening on the other side of the Atlantic. Welfare to work was promoted in Britain as in the US. For Blair, as for the Christian Conservatives who inﬂuenced the American Welfare Reform, there was a civic duty to work, and this was a fundamental part of Christianity (Lund, 2002). While the US programs focused largely on single mothers, however, initially the British welfare to work program, labeled the ‘New Deal’, focused on young people aged 18 to 25 (Ibid., p. 193). As in the US, the program included job readiness activities, training for a limited period, work experience and subsidized employment. Stringent work requirements for single mothers were introduced as a condition of receiving beneﬁts (Ibid.). A childcare tax credit was introduced and an increase in the number of childcare places was implemented. The National Minimum Wage Act of 1998 introduced a single national minimum wage and the Family Credit, paid from the Beneﬁts Agency, was replaced in 1999 by a Working Families Tax Credit, paid by the Inland Revenue (Ibid.). One aspect of the British system which distinguished it from the system in the US was the establishment in 1997 of a Social Exclusion Oﬃce in the Prime Minister’s Oﬃce (Blair, 1998).6 Blair described social exclusion as covering people who did not have the means, material or otherwise, to participate in social, economic, political, and cultural life. While the principle emphasis of Prime Minister Tony Blair’s attempt to end social exclusion was on the obligation to work, other elements of the strategy included targeted aid to depressed areas, and a focus on health and education services on the most needy.

GERMANY The history of social welfare in Germany is quite distinct from that of the UK, the US and Sweden in part because of Germany’s unique political history, particularly the impact of the Second World War and of reuniﬁcation. The West German welfare system was based on a written constitution, known as the Basic Law, adopted in 1949 and inﬂuenced by the occupying allies, particularly the US. The constitution emphasized individual liberty, strong free-market principles, and limiting the power of the Central State (Bund) while giving considerable autonomy to the federal states (Lander) and local government. A conservative deﬁnition of social rights and obligations provided the foundation for West German social welfare policy. Thus, for example, according to one authority, the concept of the welfare state ‘has a rather pejorative connotation implying a paternalistic dependence undermining individual freedom and

24

The political economy of welfare reform in the United States

initiative, not dissimilar to the meaning of “welfare” in the US’ (Ginsburg, 1992, p. 68). Under the provisions of the Treaty of Monetary, Economic and Social Union, and the formal Act of Reuniﬁcation, the principles of the West German constitution, including social welfare legislation, were extended in 1990 to the former German Democratic Republic (GDR, or East Germany). For most of the post-war period, the dominant political force in West Germany was the conservative Christian Democratic Party (CDU) and their partners the Christian Social Union (CSU). Social welfare policies underwent relatively few major changes prior to the 1990s. A period from 1969 to 1982, when the government was led by a coalition of the Social Democratic Party (SPD) and the Free Democrat Party (FDP), saw some initial expansions of welfare programs, but these were followed by retrenchment after the stagﬂation of the mid 1970s (Ibid.). In general, there is more emphasis on the role of private associations such as employers and unions, as well as individuals and families, to provide economic and social security than is the case in the other countries discussed above. Thus, for example, there is no minimum wage, or explicit wage or price policies, and little discussion of equalizing welfare outcomes, or of a universal safety net (Ginsburg, 1992). The German welfare system is based on a social insurance program which is ﬁnanced by contributions from both employers and employees and which is compulsory for almost all employees working 18 hours or more per week. Contributions are proportional to income, and beneﬁts for sickness, unemployment, disability and old age are paid out of this system according to the contributions made, and thus are also strictly proportional to former income (Poole, 2000). The wage replacement rate varies from about 40 percent of prior income for blue collar workers to about 60 to 70 percent for white collar workers (Ibid.). The program is administered by private corporations composed of unions and employers. This organizational structure results in a system that is decentralized and though regulated by the states, is not under the direct control of either federal or regional bureaucracies. The major means-tested cash beneﬁt program in Germany is Social Assistance. There is also a children’s allowance, which has been means tested since 1983 and which increases for successive children. Social Assistance is available to those who are not insured or whose insurance cover, due to low contributions, would provide a lower level of beneﬁts than the social insurance system. Although beneﬁt levels and eligibility rules are prescribed by federal law, the local Social Welfare Oﬃces have freedom to award beneﬁts according to individual circumstances, and rights of appeal

Comparative social welfare policy

25

are circumscribed. Since Social Assistance is funded primarily by local authorities, the incentives for ﬁscal stringency are clear. One consequence of this is that Germany is considered to have a low take-up rate, which has been explained by Liebfried as the result of the ‘ﬁltration of welfare rights by administrative and social devices’ such as lengthy application forms, strong liable relative provisions and provisions for prosecution for false statements (Liebfried, 1979, p. 76). Moreover, there are few organized advocates for those who feel aggrieved; Germany does not have a strong welfare rights movement comparable to that in the US or the UK, for example. As in the other countries discussed, the Social Assistance caseload has seen large increases since the 1970s, and about half of all recipients are now single heads of households. However, the proportion of single parents in Germany is lower than that in Sweden, the UK or the US, in part due to the tradition in West Germany of strong ideological support of women’s role as homemaker, and a lack of child care and kindergarten facilities throughout the country (Ginsburg, 1992). Moreover, universal child beneﬁt, which increases with additional children, maternity payments, child rearing beneﬁts and child tax allowances all support the role of women as mothers and homemakers, a role reinforced by the absence of any tax relief on child care costs for parents. Ginsburg has suggested that public policy on day care in West Germany ‘suggests that mothers of preschool children should stay at home as their full-time, unpaid caregivers, reﬂecting the male breadwinner . . . (role) . . . Parents who deviate from the conventional model are largely at the mercy of unsubsidized and unregulated forms of private care’ (Ibid., p. 89). In sharp contrast, to the societal norms in West Germany, the GDR had a very liberal social attitude in matters such as divorce and abortion: women were required to work, and welfare and other social services were contingent on such work. The dismantling of the East German economy after reuniﬁcation, which was accompanied by a rise in unemployment to over 15 percent, had a huge impact on women in the East, who were disproportionately employed in social welfare and other sectors most seriously aﬀected (Poole, 2000). Prior to uniﬁcation, 91 percent of women in the GDR were employed, compared to only 55 percent in the Federal Republic. As is the case with social insurance, the principle of subsidiary, or devolution of administrative responsibility, applies to the provision of welfare services (although not to the provision of cash beneﬁts). Thus, the provision of welfare services is ﬁrst and foremost the responsibility of individuals and their families. Only when they have shown themselves unable to provide such services will community, private sector and voluntary

26

The political economy of welfare reform in the United States

sector organizations provide assistance. This has resulted in a large voluntary sector, both religious and secular, that obtains some funding from state grants and insurance funds. When services are not available from such organizations, the individual or family may apply to the state and, ultimately, to the central government. Thus a very hierarchical system has developed for the provision of welfare services. It should be noted, however, that although beneﬁts can only be claimed from the regional or central government when lower levels have been exhausted, they are nevertheless a guaranteed right, and such right is written into the constitution. To the extent that the welfare roles are linked to employment opportunities, Germany’s unique history and the government’s employment policies, particularly relating to the recruitment and repatriation of foreign workers, have inﬂuenced the number of eligible welfare recipients. Unemployment in post-war Germany has followed a pattern quite distinct from that of other western European nations and the US. Whereas Britain and America, for example, had full employment in the 1950s, post-war Germany started out the decade with an unemployment rate of over 10 percent, partly due to the inﬂux in the post-war years, of 14 million refugees, mostly ethnic Germans from Eastern Europe (Ginsburg, 1992). The labour situation was reversed as the economy grew in the 1960s and early 1970s, when a strong demand for labour led to the recruitment of women and migrant workers from Southern Europe. The Federal Labour Oﬃce established recruitment agencies in Italy, Spain, Greece, Morocco, Turkey, Yugoslavia and Poland, although legislation passed in 1969 stipulated that German citizens should always be given preference over foreigners for job vacancies. However, this process ended after the oil shocks of the 1970s when the federal government oﬀered ﬁnancial inducements to these guest workers to return home and active recruitment was replaced by tough immigration controls (Ibid.). Unemployment fell and remained at around 4 percent after the 1973 recession in part as a result of the departure of over half a million guest workers (Ibid.). The more severe recession of 1981 raised the unemployment rate to 8 to 10 percent, however, a rate which continued into the 1990s. Throughout the 1980s and 1990s, in addition to the oil price shocks, Germany faced increasing competition on world markets as a result of globalization. Moreover, the economic pressures described above were exacerbated by pressure from the European Union, which on the one hand required ﬁscal discipline from member states, and on the other required policy changes, such as harmonization of social policies among the member nations, which were potentially extremely costly. Added to these

Comparative social welfare policy

27

external pressures were increased domestic pressures resulting from high unemployment and demographic changes. A falling birth rate and an aging population led to strains on social insurance programs, such as old age beneﬁts and medical insurance, as well as Social Assistance. The increase in unemployment and in demand for welfare services led to concern about welfare dependence and inadequate work incentives and was accompanied by work incentive measures such as retraining schemes as well as a reduction in social spending as a percentage of GDP which continued until reuniﬁcation in 1990 (Poole, 2000). Thus, when East and West Germany were reunited in 1990 the economy in the West was already under considerable pressure. With reuniﬁcation, the East German economy underwent massive restructuring. This process was accompanied by huge transfers of resources from the West, as unproductive sectors were dismantled and unemployment soared. Workers who had previously had job security found themselves unemployed and without access to the free or subsidized housing, daycare, health care and other beneﬁts which had been available in the East. They were, however, by law entitled to the beneﬁts available in the West, resulting in ﬁscal strain for Lander in the West, who were unwilling to take on the full cost of assimilation, and the federal government ultimately bore a large part of the increased costs. To deal with these increased costs and rising concerns about dependency, a series of reductions in beneﬁt levels and increases in social insurance contributions were implemented in the 1990s. Although large numbers of foreign workers left in response to government ﬁnancial inducements, those that stayed often responded to the tougher immigration controls by bringing their families to settle permanently in Germany, thus increasing the number of dependents and potential welfare beneﬁciaries (Ibid.). Despite high barriers to citizenship, including the prohibition of children born to foreigners obtaining citizenship, foreigners may be entitled to a number of welfare beneﬁts, including unemployment beneﬁt and social assistance. Although there have been pressures for greater entitlements from the ethnic minority communities both through the trade union movement and through their own community political organizations, the high levels of unemployment have also given rise to increased demands from other groups for forced repatriation of foreign workers (Ibid.). Overall, social welfare policy in Germany has been far less inﬂuenced by special interests than in the other countries considered here. The growing inﬂuence of left wing groups such as the feminists and environmentalists which would extend the welfare state is counterbalanced by those right wing groups which would curtail welfare programs and repatriate many foreign workers. Moreover, many of these groups are single issue groups.

28

The political economy of welfare reform in the United States

Collective representation of consumers or potential beneﬁciaries is largely in the hands of the political parties or the unions. The absence of major national confrontation may also be due to the fact that while social policy is federally regulated, it is largely administratively decentralized or privatized. The social security system is based on contributory actuarial principles, similar to private insurance companies, for example, and the Catholic and Protestant churches, regulated by local authorities, provide a large proportion of personal social services, including child care. Moreover, since social insurance contributions are directly linked to beneﬁt levels, the social insurance program is not seen as a redistributive system and is widely supported by the middle classes.

THE IMPACT OF THE EUROPEAN UNION ON EUROPEAN SOCIAL POLICY The European Economic Community (EEC) was founded in 1957 on the basis of free market principles with the objective of greater economic growth for community members. It was believed that competition on an equal and fair basis among the six founding members, France, Italy, Belgium, Luxemburg, the Netherlands and the Federal Republic of Germany, would lead to an optimal distribution of resources among the nations and thus maximum economic growth (Hantrais, 2000). Among the speciﬁc policies introduced to achieve these goals were the replacement of national tariﬀs by a common external tariﬀ and measures to encourage the mobility of labor between nations. Apart from a general commitment to raising standards of living and promoting cooperation in the social ﬁeld (primarily related to employment, including working conditions, labor law and social security), the most signiﬁcant social policy contained in the Treaty creating the EEC was the creation of European Social Funds which, inter alia, helped ﬁnance the training and relocation of workers and aided economically declining areas. A more explicit set of social policies was adopted in 1961 by the Council of Europe in a Social Charter which guaranteed, though with no legally binding provisions, fundamental rights for workers and citizens, including the speciﬁc rights of families, mothers and children to social, legal and economic protection (Ibid.). More than a decade later, in 1974, the European Council of Ministers passed a resolution concerning a social action program, that noted that economic expansion was not to be seen as an end in itself, but rather should lead to an improvement in the quality of life (Ibid.). The speciﬁcs, however, again related primarily to the working environment, including full and better employment, improved working condi-

Comparative social welfare policy

29

tions, and greater involvement of management and labor in economic and social decision-making. Moreover, since the Community did not have legal authority to implement a social policy, the resolution was conﬁned to expressing the political will of signatories, and stated that it did not seek a standard solution to all social problems or attempt to transfer to Community level any responsibilities that could be assumed more eﬀectively at other levels (Ibid.). This principle of the subsidiarity of the Community to that of national governments was to continue to be an integral part of European policy throughout the remainder of the twentieth century. A further impetus was given to the adoption of social policy, although again primarily relating to employment, when Jacques Delors became President of the Commission in 1985. Delors believed in the interrelatedness of economic and social policy and stated that: ‘The European social dimension is what allows competition to ﬂourish between undertakings and individuals on a reasonable and fair basis . . . Any attempt to give new depth to the Common Market which neglected this social dimension would be doomed to failure’ (Delors, 1985). Social policy was a signiﬁcant consideration in negotiations preceding the Single European Act (SEA) of 1986 and the social partners, employers and employee representatives, were involved in the discussions.7 A major new directive in this Act was the opening up of new policy areas to be determined by majority vote rather than unanimous voting. For the ﬁrst time member states could be forced to accept a speciﬁc provision even if it was against their noted objections.8 Three years later, in 1989, the heads of all member states, with the exception of the United Kingdom, adopted the Community Charter of the Fundamental Social Rights of Workers, heralded as the social dimension of the SEA (Hantrais, 2000). Reﬂecting the words of Jacques Delors four years earlier, the preamble to this charter stated that ‘the same importance must be attached to the social aspects as to the economic aspects and . . . therefore, they must be developed in a balanced manner’ (Ibid.). It must be understood, however, that, as the title implies, the rights in question related primarily, if not exclusively to workers rather than to citizens in general, and the focus was on rights relating to employment. Much of the debate surrounding this charter focused on whether social protections such as social security should aim at a community minimum or maximum level of provision. There were fears that some states would seek competitive advantage by requiring or oﬀering very little social protection, thus leading to what came to be termed ‘social dumping’ a concept having something in common with the fears in the US of a race to the bottom, when states were allowed more freedom over their welfare programs. Such fears had proven unfounded in the US as states made few changes in their

30

The political economy of welfare reform in the United States

beneﬁt levels after the introduction of the New Welfare law in 1996. Similar fears had also been aroused when Greece joined the Community in 1981, and Spain and Portugal in 1986, all of which had low levels of social protection for workers. One outcome of the Community Charter in the ﬁeld of social policy was a provision that resulted in an action program for the Single European Market (SEM) that came into operation in 1993. The focus was again employment oriented, with few new measures and the methods to be used for implementing the action program were consultation, advisory committees and social dialogue. A balance was sought in implementing the action program between subsidiarity, accepting the diversity of national systems, cultures and practices, and preserving the competitiveness of undertakings (Ibid.). The negotiations at the Maastricht summit in 1992 provided further evidence of the challenges in forging consensus over social issues. The Agreement on Social Policy was, at the insistence of the UK, removed from the Treaty and a separate Protocol on Social Policy, agreed to by the other 11 member nations was included as an annex. This Protocol identiﬁed a number of speciﬁc objectives, including ‘the promotion of employment, improved living and working conditions, social protection, dialogue between management and labour, and the development of human resources with a view to lasting high employment and the combating of social exclusion’ (Ibid., p. 11). As in previous EU policy documents, the role of the Council regarding social issues was limited to issuing directives, monitoring social conditions and setting minimum standards for gradual implementation. Maastricht also was signiﬁcant in that additional areas of social policy were opened up to majority voting rather than unanimity, thus allowing for advancement in policies that would previously have been vetoed.9 That social policy remained very much on the European agenda was clear when the Commission published a green paper on European social policy in 1994, followed by a white paper later that same year. The latter noted that social policy had a vital role to play in ensuring that the people of Europe beneﬁted from economic well-being, social cohesiveness and high overall quality of life. Although employment was still of paramount importance, the white paper speciﬁcally noted that categories of people not in work should also be taken into account and the ‘fundamental social rights of citizens’ should be established as a constitutional element of the European Union. The reference to citizens rather than workers was a clear shift from the position of the Community Charter on the Fundamental Rights of Workers. The increasing acceptance of incorporating social policy into EU Treaties was further evidenced by the Treaty of Amsterdam, signed in

Comparative social welfare policy

31

October 1997, which, although the primary focus was still employment, incorporated the Agreement on Social Policy into the main body of the document after the United Kingdom withdrew objections. During negotiations for this Treaty, member states acknowledged that much needed to be done at national or business level to stimulate employment, but that there was also scope for international cooperation and for supporting action by the community. This was reﬂected in a separate title on employment being added to the Treaty.10 The Treaty also included a legal basis for action to deal with social exclusion by incentive measures, an initiative that was proposed and pressed by Ireland during Treaty negotiations.11 The Treaty of Amsterdam reﬂected a compromise between Sweden, who had a target of full employment, and states such as Germany, the UK and the Netherlands who focused more on competitiveness and sought only a ‘high’ level of employment. There was consensus on active employment policies, however, and a primary goal of all the members, as in the US at that time, was to reduce public expenditure by moving people from welfare to work (Hantrais, 2000, p. 228). Other developments reﬂected in this Treaty were the extension of the jurisdiction of the European Court of Justice and some strengthening of the position of the Economic and Social Committee. A further strong commitment to addressing social policy was contained in a 1999 statement from the Commission on a strategy for modernizing social protection in the Union. This latter document stated that strong social protection systems are an integral part of the European Social model, which is based on the conviction and evidence that economic and social progress go hand in hand and are mutually reinforcing factors. Social protection provides not only safety nets for those in poverty; it also contributes to ensuring social cohesion by protecting people against a range of social risks. It can facilitate adaptability in the labor market and can thus contribute to improved economic performance. Social protection is a productive factor (Ibid.). A number of economic and political factors contributed to an increased role for social policy in the 1990s. Persistently high unemployment, the pending integration of Central and Eastern European states into the Union, the aging of the population in many of the member states and the macroeconomic requirements on states related to the European Monetary Union, all gave impetus for addressing social issues (Ibid.). Member states throughout the Union had shared goals of trying to reduce unemployment and at the same time contain social spending, although the policies they employed to address these problems diﬀered, with some states, such as the United Kingdom and Belgium reducing the duration for which beneﬁts could be claimed (as in the USA) while others

32

The political economy of welfare reform in the United States

lengthened the required contribution time for eligibility, and countries such as the Netherlands and Sweden, which had been very generous, tightened the conditions required to qualify for beneﬁts (Ibid.). In addition, more targeting was introduced into some programs, with child beneﬁts in several counties, including Spain and Italy, being targeted at the most needy cases. Moreover, the growth of structural funds, which reached almost 36 percent of the EU budget in the late 1990s, (second to the Common Agricultural Policy funds at 46 percent) allowed for compensating member states for making concessions and facilitating agreements on otherwise divisive economic and social issues. The overall impact of these should not be overestimated, however, since the total EU budget was equivalent to only 0.46 percent of the Union’s total GNP in 1999 (Ibid.).

CONCLUSIONS When the European Economic Community was founded in 1957, the member states were experiencing a period of economic expansion. Unemployment was not seen as a major problem, the proportion of retirees to workers was relatively small and immigrants contributed to a growing, relatively young workforce. Neither was poverty seen as a major issue, though non-contributory social assistance schemes were available for those who fell through the net. In the ensuing decades, however, in Europe as in the US, the energy crises and the end of the baby boom brought economic and demographic changes and the nature of the problem changed. Whereas in the 1970s most of the poor in Europe were older people, by the 1990s a growing number of working age people were falling into poverty and the number of long-term unemployed, single parents and young people who had never been employed was increasing (Ibid.). Moreover, while the founding states generally had welfare systems based on what has come to be known as the continental model, where social beneﬁts are linked to earnings contribution, those states joining the Community in 1972, Ireland, Denmark and the UK had a diﬀerent welfare tradition or ideology, one based on universal citizenship rights. Separate and apart from national programs, the European Commission introduced measures related not just to employment, but also to address the issue of poverty in member states. From the mid 1970s, the community initiated action programs, which primarily involved researching and monitoring poverty related issues and proposing policies to address them. A second strategy, which was primarily aimed at workers, was the development of the European Social Fund, intended to reduce disparities in development

Comparative social welfare policy

33

among diﬀerent regions and to reduce the ‘backwardness of the least favored regions’ (Ibid., p. 168). The gradual movement from a worker focus to a citizen focus for social action was expressed in a recommendation by the Council in 1992 that recognized ‘the basic right of a person to suﬃcient resources and social assistance to live in a manner compatible with human dignity’ (Ibid., p. 172). The Council further recommended that the amount of resources available to a household should be ‘suﬃcient to cover essential needs with regard to respect for human dignity, taking account of living standards and price levels in the Member State concerned, for diﬀerent types and sizes of households’ (Ibid.). While the rhetoric was becoming more expansive, however, several nations balked at implementing speciﬁc action programs. Thus an action program to deal with poverty, proposed in 1994, was opposed by Germany both on the grounds that the Commission did not have legal authority to implement such a program, and also on the basis of cost. When the Commission went ahead and started to implement the program it was taken to the European Court of Justice by the UK, which succeeded in the case in 1996 (Ibid.). That same year, while funding for the Community Action Programme was blocked, agreement was reached on the allocation of European support funds directed at new employment initiatives, and this trend towards a greater focus on employment, which paralleled the trend in the US, was reiterated in employment guidelines given by the Council in 1998, which advocated active employability measures rather than passive support measures. While policy regarding the European labor market has been the subject of the most extensive discussion, directives and regulation within the Community, the actual impact of such labor market policies is quite limited. Such policies focus primarily on anti-discrimination measures, health and safety issues and the mobility of labor, including portability of rights such as unemployment beneﬁts and social security. While the European Court of Justice has ruled on anti-discrimination cases and had some impact in this area, other areas, such as those relating to migration have remained largely unaﬀected. One reason for this may be the small number of citizens that would be aﬀected, only 1.4 percent of EU citizens, according to Hantrais (2000). Moreover, there is evidence that, in part due to security concerns, there is growing pressure to control and even reverse the ﬂow of migrants among European community countries as well as from outside of the Community. Social policy requirements set by the EU mean a reduced political decision-making role for national bodies. Moreover, implementation of policies must be carried out by institutions which have no role in their formulation and which may not be supportive of the policies. An example is

34

The political economy of welfare reform in the United States

the employee protection measures which were met with heavy criticism and resistance from employers’ organizations in Austria. European level interest groups, on the other hand, such as the Union of Industrial and Employee’s Confederations of Europe (UNICE), the European Centre of Enterprises with Public Participation (CEEP), and the European Trade Union Confederation (ETUC), are strengthened by their direct access to EU policy-making institutions. A further shift in the role and inﬂuence of the diﬀerent interest groups involved in social policy results from the structure of the Union. Membership is linked to an increase in the importance of government, the main addressee of, and participant in, EU activities. In the social policy ﬁeld, countries which previously had a tripartite approach with collaboration between government, employers and employee associations, may see a shift in favor of government. Since employers’ associations generally have policies more in line with the administration, it is likely that employees’ associations will experience the greatest loss of inﬂuence.

NOTES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

Quoted in Ginsburg (1992), p. 48. Quoted in Ginsburg (1992), p. 40. Quoted in Ginsburg (1992), p. 32. Cited in Craig (1975), p. 427 and Lund (2002), p. 150. Quoted in Lund (2002), p. 192. Quoted in Lund (2002), p. 192. Http://www.pitt.edu/~heinisch/eu_integ 2.html, p. 4. Ibid. Http://www.pitt.edu/~heinisch/eu_integ 2.html. Http://www.ireland.com/special/treaty/treaty/white2.htm, p. 6. Ibid., p. 7.

3.

Traditional public interest model

The appropriate role for the government to play in the economic aﬀairs of a nation has been the subject of debate and contention for centuries. There has never been consensus on either the extent of government involvement in the economy or the speciﬁc areas in which it is appropriate for the government to intervene. In the seventeenth and eighteenth centuries, for instance, many economists, particularly in France, believed that the government had a major role to play in controlling international trade. In an era marked by competition among the European powers to establish colonies, political interests and commercial interests coincided in the promotion of trade and industry. The public interest, however, may not have coincided with either. Partly as a reaction to this mercantilist view, Adam Smith wrote the Wealth of Nations (1776) which advocated a much more limited role for government, and which argued that individuals, in pursuing their own self interest, would be led as if ‘by an invisible hand’ to serve the public interest. In this view the proﬁt motive and competitive forces will lead to production, at the lowest prices, of those goods that are most in demand. Thus Smith writes that man: . . . intends only his own gain, and he is in this, as in many other cases, led by an invisible hand to promote an end which was no part of his intention. Nor is it always the worse for the society that it was no part of it. By pursuing his own interest he frequently promotes that of the society more eﬀectually than when he really intends to promote it. (1776, p. 129)

Thus, given the appropriate assumptions, discussed in more detail below, unfettered market forces will tend to lead to an eﬃcient allocation of resources in the economy. The competitive market solution promoted by Smith, in the absence of policy-relevant externalities, can be shown to represent not only an eﬃcient allocation of resources but also an optimal outcome in the sense deﬁned by Vilfredo Pareto. Writing in the 1890s and early 1900s, Pareto set forth criteria upon which the theoretical foundations of welfare maximizing economic policies are largely (and some would say unjustiﬁably) based.

35

36

The political economy of welfare reform in the United States

PARETO OPTIMALITY The Pareto criterion states that an outcome is Pareto optimal if no one person can be made better oﬀ without making at least one other person worse oﬀ. If a particular change in the economy would result in one person being better oﬀ and no-one worse oﬀ social welfare would have improved and the change would result in a Pareto superior position. Theoretically, there exists a frontier of Pareto optimal states of the economy. The degree to which these states are a useful guide for policy is debatable. Moreover, even holding the position that the economy ought to be at a Pareto-optimal position entails a value judgment that lies outside the realm of positive economics. What economic science can contribute, however, is an analysis of the conditions which must hold in order for the economy to be at a point on the Pareto frontier. The following section outlines these conditions.1

PURE EXCHANGE Consider an economy of two individuals and two commodities, x and y. The amount of x consumed by the ith person is denoted xi. The total amounts of x and y are ﬁxed, for simplicity, so that x1 x2 x, and y1 y2 y. In order to calculate the conditions for the Pareto-optimal allocation of x and y between individual 1 and 2, we can formulate the problem mathematically as follows. Maximize: U 2(x2, y2) Subject to: U1 (x1, y1 ) U10 x1 x2 x, y1 y2 y This can be solved using a Lagrangian: L U 2 (x2, y2 ) (U10 U1 (x1, y1 )) x (x x1 x2 ) y (y y1 y2 ) Diﬀerentiating we get: Lx2 U 2x x 0 Ly2 U y2 y 0 Lx1 U1x x 0

37

Traditional public interest model

Ly1 U1y y 0 Combining equations we get: Ux1 x Ux2 Uy1 y Uy2 which is the tangency condition that the consumer’s indiﬀerence curves have the same slope, i.e., the marginal rate of substitution of x for y is the same for both consumers, a condition which must hold if the gains from trade are to be exhausted. This condition can be illustrated by means of the Edgeworth box diagram in Figure 3.1, which measures the total amount of good x available in the economy on the horizontal axis and the total amount of good y on the vertical axis. The origin for consumer 1 is O1, and for consumer 2 it is O2. At an initial allocation A, consumer 1 has quantity x1 of good x and quantity y1 of good y, while consumer 2 has x2 of good x and y2 of good y. Gains from trade can be achieved by a move from A to any point on the contract curve between B and C. The points B and C represent points of tangency between the indiﬀerence curves at which the marginal rates of substitution are equal. At such points, which comprise the contract curve, no Pareto superior trade is available. For any point oﬀ the curve it can be shown that a Pareto superior move is possible. The contract curve thus represents the locus of Pareto optimal distributions of goods among consumers. The utility levels for each consumer represented by the diﬀerent points along the contract curve can be used to derive the Pareto frontier (see Figure 3.2). For any given amount of the goods x and y there exists a set of points for which neither can gain without the other losing. The end points on this frontier represent allocations in which one consumer has all of both goods. O2 C y2

B

A y1 x1

O1

Figure 3.1

The Edgeworth box diagram

x2

38

The political economy of welfare reform in the United States

U2

U2 = f (U1, x, y)

O Figure 3.2

U

U1

Utility frontier for given total quantities of goods

PRODUCTION In order for consumers to be on the Pareto frontier in consumption, the goods must be produced eﬃciently. If production was at a point inside the production possibilities frontier consumers could both gain by a move to a point on the frontier. Hence, in order to deﬁne the Pareto frontier for consumers in the case where x and y are produced (rather than ﬁxed, as in the preceding analysis) it is necessary to deﬁne the production possibilities frontier. This problem can be stated mathematically as follows: Maximize: y f(Ly, Ky ) Subject to: g(Lx, Kx ) x Lx Ly L

Kx Ky K

where f and g are the production functions of y and x respectively; Ly and Ky represent the amounts of labor and capital, respectively, used in the production of good y; and production of x is ﬁxed. The Lagrangian for this problem is: L f(Ly, Ky ) (x g(Lx, Kx )) L (L Lx Ly ) K (K Kx Ky )

Traditional public interest model

39

Diﬀerentiating yields the following conditions: fl l 0 fk k 0 gl l 0 gk k 0 which gives the familiar tangency condition: fl l gl fk k gk i.e., the ratio of marginal products must be equal for all gains from trade to be exhausted in production, and this ratio must be equal to the ratio of the marginal costs. The set of eﬃcient production plans that this condition implies can be represented on the production possibilities frontier, and for a production plan to be Pareto optimal it must in fact be on this frontier.

THE EQUIMARGINAL PRINCIPLE We have so far derived two conditions for optimality: consumers must be on their contract curve for any production level (x, y), and production must be on the production possibilities frontier. One more tangency condition is required: for each consumer, the marginal rates of substitution between x and y must equal the marginal cost of producing x in terms of y, with the quantities of x and y produced being determined by the production possibilities frontier. Thus for overall Pareto optimality (production and consumption), the marginal valuation of each commodity must be the same for all individuals, and that common marginal evaluation must equal the marginal cost of producing that good. This third condition may be derived in a manner analogous to the ﬁrst, with the exception that x and y are now determined by the production possibilities curve rather than being ﬁxed, i.e., the amount of each output is a function of the production of the other output and available supplies of inputs. The problem then is: Maximize: U 2 (x2, y2 )

40

The political economy of welfare reform in the United States

Subject to: U1 (x1, y1 ) U10 x1 x2 x y1 y2 y y y*(x, L, K) The last three constraints, which deﬁne the production possibilities curve can be combined and written in implicit form, h(x, y). The Lagrangian can then be formulated: L U2 (x2, y2 ) 1 (U10 U1 (x1, y1 )) h(x, y) which gives the ﬁrst order conditions: U 2x hx 0 U 2y hy 0 1U1x hx 0 1U1y hy 0 Combining and simplifying we have: U1x U x2 hx U1y U 2y hy where hx/hy is the slope of the production possibilities frontier, and where this slope is equal to that of the consumers indiﬀerence curves. This situation is shown geometrically in Figure 3.3. The production possibilities frontier for given resource endowments is represented by the curve PP. The slope of this frontier equals the marginal cost of producing x in terms of y. At any point, such as A, which represents a certain quantity of x and y that is produced, an Edgeworth box diagram may be constructed. The points inside the box represent allocations of x and y to the two consumers, and OA represents the contract curve. At some point or points on this curve, say A, the consumers marginal rates of substitution will equal the marginal cost of x or marginal rate of transformation, shown as the slope of a line tangent to A. This is an overall Pareto eﬃcient point since the marginal rates of substitution for the consumers are equal and they equal the marginal cost of production.

41

Traditional public interest model

y P

A A´ O Figure 3.3

P

x

Overall Pareto optimality

U1 U U´ U´´ U´´´

O Figure 3.4

U´ U´´

U´´´ U

U2

Partial and overall utility frontiers

Clearly, an Edgeworth box diagram can be drawn for each point along the production possibilities frontier, giving a new frontier of Pareto optimal points for this particular production level of x and y. Moreover, multiple frontiers can be derived for alternative production levels, as shown in Figure 3.4. The envelope curve for all these frontiers UU represents the maximum utility any consumer can achieve for a given level of the other consumer’s utility. It comprises a complete set of Pareto optimal production and allocation levels of goods x and y. The choice of which point is best for a society is nowhere addressed in this analysis. While Paretian welfare economics is widely employed in the economic theory of public policy, it has been criticized on a number of grounds, not least of which is the value assumptions on which it is based (Rowley and Peacock, 1972). The primary Paretian criteria for evaluating social welfare is that the economy is at a Pareto optimum if no reallocation of resources is possible which improves the welfare of at least one individual and leaves

42

The political economy of welfare reform in the United States

no-one worse oﬀ. Thus, interpersonal welfare comparisons cannot be made and intensity of preferences is not considered. A policy which substantially increases the resources of all but the richest member of a society, and leaves him only slightly worse oﬀ, would not be considered a Pareto improvement. A further limitation in Paretian welfare economics is that it holds that social welfare is a function of the welfare of each individual in society, and each individual is the best judge of his own welfare. Thus paternalism of any kind is ruled out, and no individual can impose his preferences on another regardless of prevailing ethics. A major limitation of the Pareto framework is that the Paretian ranking of states is incomplete. For instance, if certain individuals prefer state a to state b, while others prefer b to a, no ranking is deﬁned for these two states; they are Pareto-non-comparable. In situations such as this the status quo dominates and for this reason the Pareto principle has been described as conservative. But the most important objection to the Pareto criteria, according to Paul Samuelson, is ‘the lack of emphasis upon the fact that an optimum point, in his sense, is not a unique point’ (Samuelson, 1947, p. 214). Diﬀerent initial distributions will inﬂuence the optimum points. As Samuelson notes: If transfers of income from one individual to another are arbitrarily imposed, there will be a new optimum point, and there is absolutely no way of deciding whether the new point is better or worse than the old . . . optimum points constitute a manifold inﬁnity of values . . . (which) . . . can be obtained under regimes quite diﬀerent from perfect competition (ibid.).

In order to address some of the limitations described above, a number of extensions to the Pareto principle have been developed. Two of these, the compensation principle and interdependent utility functions are discussed below. The notion of interdependent utility functions recognizes that the utility of one individual may be dependent on that of another. H.M. Hochman and J.D. Rogers (1969) analyzed the case of a Pareto-optimal redistribution of income between two individuals, Mutt who was rich and Jeﬀ who was poor, where the utility of Jeﬀ was an argument in the utility function of Mutt. The authors show that the degree of income redistribution which is Pareto optimal is a function of the initial distribution of income and of Mutt’s marginal rate of substitution between keeping income and transferring income to Jeﬀ. While there is no reason that such transfers could not take place through the private sector for a small economy, the free rider problem could indicate the possibility of gains from collective action. Even here, however, the practical eﬃcacy of transfers is called into question by the ﬁndings of Gordon Tullock (1970) that the vast majority of income

Traditional public interest model

43

transfers in the United States are not from rich to poor but rather within middle income groups. A further attempt to extend the Pareto analysis was contained in the principle of potential compensation. This notion held that a reallocation of resources which resulted in some individuals being better oﬀ and some worse oﬀ, is Pareto superior if the gainers could compensate the losers while themselves remaining better oﬀ. Note that the compensation is not actually paid, for then the Pareto criterion would hold, since no-one was worse oﬀ. The potential compensation criterion, also known as the Kaldor–Hicks criterion since it was derived by these economists, suﬀers from a major weakness, initially noted by Scitovsky (Mishan, 1971). The problem arises because, as Scitovsky demonstrated, a move from allocation A to allocation B may represent a Pareto superior move while at the same time a move in reverse from B to A can be shown to be Pareto superior (Rowley and Peacock, 1972). To deal with this problem, Scitovsky proposed a stricter test which would ﬁrst assess whether the initial move was superior and then whether the reverse move was. Only if a reallocation passed the ﬁrst test and failed the second would the reallocation be considered a welfare improvement. The attempts just described to extend the Pareto criteria are based, as is the criterion itself, on the assumption of perfect markets. In fact, however, instances of market failure are multiple, and these failures are often considered a primary justiﬁcation for government intervention in the economy. Some primary examples are discussed below.

MARKET FAILURE – PUBLIC GOODS One of the most commonly cited reasons for government intervention is the existence of public goods such as national defense, a police force, a ﬁre service, or a system of property rights and the procedures to enforce them. Such goods will generally be undersupplied or not supplied at all by the free market because of certain characteristics: jointness of supply and nonexcludability (Mueller, 1990). Jointness of supply refers to a good whose production costs are ﬁxed and which may be provided to additional persons at no extra cost. In the case of national defense, for example, it costs no more to defend one million persons from attack than one million and one. Formally, an extra unit of the good can be produced at zero marginal cost. This characteristic gives rise to a prisoner’s dilemma situation, in which private incentives lead to a suboptimal provision of the resource and all parties can gain from a collective decision to provide a higher level of resources.

44

The political economy of welfare reform in the United States

Non-excludability describes a situation where it is very diﬃcult, or impossible, to eﬃciently exclude some persons from consumption of the good once it has been supplied to others. When a national defense system is in place it is impossible to exclude one person from the protection of that system. In the absence of government provision of these goods, incentives would exist for non-cooperative behavior such as free riding and they would tend to be under-provided or not provided at all. To determine the eﬃcient output of a public good requires a comparison of marginal costs and marginal beneﬁts. In contrast to a private good, the marginal beneﬁt is not the beneﬁt one individual places on the good since its provision allows a potentially large number of other persons to beneﬁt. Rather the marginal beneﬁts of all individuals aﬀected must be summed, giving the marginal social beneﬁt (MSB). The public goods problem is illustrated graphically in Figure 3.5. The diagram shows units of the public good on the horizontal axis and cost or beneﬁt per unit on the vertical axis. The demand curves of two consumers are represented by DA and DB. Since tastes diﬀer, these curves will diﬀer. The vertical sum of these curves is the MSB. The marginal cost curve is drawn on the assumption of constant marginal costs, for simplicity, and is thus a straight line, MC. Thus the optimal level of provision of the public good is at the point of intersection of marginal cost and marginal social beneﬁt, giving optimum output OQ. Because of the free rider problem, however, individual B has no incentive to contribute to the public good. If individual A maximizes his utility by equating his individual marginal cost MC, MB

MSB DA MC DB

O Figure 3.5

R

Q

The public goods problem

Output

Traditional public interest model

45

and beneﬁt, the output level he demands, output OR, will be the actual and suboptimal output.

EXTERNALITIES Externalities are another major reason for government intervention. An externality exists when the production or consumption activities of one individual or enterprise give rise to changes in the utility or costs of a third party. A factory polluting a stream used by others, or cattle straying onto a farmer’s crops and damaging them are typical examples of negative externalities. Since the total social costs of the activity are not reﬂected in the production costs, these entities will engage in ineﬃciently high levels of production. Externalities may be positive as well as negative. A citizen planting a ﬂower garden which is then enjoyed by her neighbors, would constitute an example of a positive externality. In this case the likely result of the divergence between social and private cost is likely to be an undersupply of the good. In either case, however, since there is no price mechanism to coordinate the activities involved in the production of the externality, the outcome may not be Pareto-optimal. An example of a positive externality is illustrated in Figure 3.6. The competitive supply and demand curves are shown as S and D, with S again drawn horizontally, implying a constant cost industry. The demand curve represents eﬀective market demand and the equilibrium output is at the intersection of S and D, output Q. This output does not take into account external beneﬁts, however, which are represented by the marginal external beneﬁt curve, MEB. This curve reﬂects the beneﬁts to individuals who do not participate in this market, but beneﬁt from the participation of others. The marginal social beneﬁt (MSB) curve reﬂects both direct and external beneﬁts, and is derived by vertical summation of the demand and marginal external beneﬁt curves. It is clear from the diagram that while the equilibrium market output is at level Q, the eﬃcient output, which would equate MSB with marginal cost, would be at the higher output level, Q. A solution to the problem of externalities, commonly associated with A.C. Pigou, was for the government to intervene to bring about a Paretooptimal situation by, for example, levying taxes, oﬀering subsidies, or introducing legislation to adjust the levels of the activity which gives rise to the externality (Pigou, 1920). By engaging in such activities, governments can ‘control the play of economic forces in such wise as to promote the economic welfare, and through that, the total welfare of their citizens as a whole’ (Ibid., p. 129).

46

The political economy of welfare reform in the United States

Price

MSB

D

S MEB

Q Figure 3.6

Q´

Quantity

A positive externality

A clear assumption of this policy prescription is that the government must posses all the relevant information to determine the appropriate tax or subsidy, including the responses of the entities concerned. Both the ability of the government to perform this function and the necessity of it doing so have been questioned. One of the major challenges is that oﬀered by Ronald Coase in his seminal article, ‘The Problem of Social Cost’ (1960).

THE COASE THEOREM Ronald Coase argued that a Pareto-optimal outcome of externality situations could, in theory, be worked out by the concerned parties without any government intervention. Traditional analysis of externality problems was usually based on the treatment of Pigou in The Economics of Welfare (1920), and regulations or taxes were the proposed solution to the divergence of social and private cost entailed by the externality. However, Coase claims that these courses of action are ‘inappropriate, in that they lead to results which are not necessarily, or even usually, desirable’ (Coase, 1960, p. 2). Moreover, Coase argued that the achievement of a Pareto-optimal outcome was independent of the initial assignment of property rights. The Coase theorem holds that: ‘In the absence of transactions and bargaining

Traditional public interest model

47

costs, aﬀected parties to an externality will agree on an allocation of resources that is both Pareto optimal and independent of any prior assignment of property rights’ (Mueller, 1989, p. 28). Coase used several examples to illustrate his argument, including that of straying cattle which destroy crops growing on neighboring land. He begins by pointing out the reciprocal nature of these problems, i.e., imposing taxes on the entity creating the externality is just as much a harm to this person as the imposition of the externality is on the entity adversely aﬀected. Given that the cattle rancher is liable for the damage caused by his cattle, he will take the costs into account in his production decisions, and not increase the size of the herd above the level at which the marginal damage cost he must pay equals the returns from the extra meat provided by an extra steer. Hence private and social costs will be brought into alignment. Coase extends the illustration to show that alternative scenarios, including the construction of fences by the cattle raiser or the payment to the farmer to reduce or cease crop production, could be predicted to result in an optimal outcome. Comparing this solution to the Pigovian solution, he notes: A procedure which merely provided for payment for damage to the crop caused by the cattle but which did not allow for the possibility of cultivation being discontinued would result in too small an employment of factors of production in cattle raising and too large an employment of factors in cultivation of the crop (Coase, 1960, p. 6).

Coase also examines the situation when the damaging business is not liable for costs, and shows that the allocation of resources will be the same in this case as it was in the former. In this case, however, the crop producer would make payments to the cattle raiser in order to bring about an optimal outcome. Coase concludes that: It is necessary whether the damaging business is liable or not for damage caused since without the establishment of this initial delimitation of rights there can be no market transactions to transfer and recombine them. But the ultimate result (which maximizes the value of production) is independent of the legal position if the pricing system is assumed to work without cost (1960, p. 8).

The argument up to this point assumed that no costs were involved in carrying out market transactions. Coase admits that this is a very unreal assumption. He notes: In order to carry out a market transaction it is necessary to discover who it is that one wishes to deal with, to inform people that one wishes to deal, and on what terms, to conduct negotiations leading up to a bargain, to draw up the contract, to undertake the inspection needed to make sure that the terms of the contract

48

The political economy of welfare reform in the United States are being observed, and so on. These questions are often extremely costly, suﬃciently costly at any rate to prevent many transactions that would be carried out in a world in which the pricing system worked without cost (1960, p. 150).

An alternative form of organization which could achieve the same result is the ﬁrm, where an administrative decision is substituted for individual bargains. Even here, however, it does not necessarily follow ‘that the administrative costs of organizing a transaction through a ﬁrm are inevitably less than the cost of the market transactions which are superseded’ (Coase, 1960, p. 16). One alternative to the ﬁrm is direct regulation by the government, which Coase suggests, is, in a sense, a super-ﬁrm (Ibid., p. 17). The government solution may also be extremely costly, however, and has additional problems. Thus, Coase notes that: ‘there is no reason to suppose that the . . . regulations made by a fallible administration subject to political pressures and operating without any competitive check, will necessarily always be those which increase the eﬃciency with which the economic system operates’. For this and other reasons, he suggests a feasible alternative is to do nothing since ‘it will no doubt commonly be the case that the gain which would come from regulating . . . will be less than the costs involved in Government regulation’ (Ibid., p. 18). The appropriate policy, according to Coase, will depend on the speciﬁc details of each individual case. A pragmatic approach is called for in which the analyst examines the total social product yielded by alternative social arrangements (Ibid., p. 43).

ALLOCATIVE EFFICIENCY OR REDISTRIBUTION The foregoing analysis has been framed in terms of the ability of the state to promote allocative eﬃciency and move the society from a point inside the production possibilities frontier to a point on it. For some economists this is the only valid role for the state to play. Thus authors such as Knut Wicksell (1896) maintain that government activity is justiﬁed only if it beneﬁts all citizens and he thus promotes the unanimity rule for collective decision-making. An alternative viewpoint holds that, either in theory or in practice, a major role the state plays is that of the redistribution of national wealth. Aranson and Ordeshook (1981), for example, regard virtually all of government activity as being based on redistribution. Thus a bridge across a river beneﬁts not only the citizens who wish to cross the river but also the contractors, engineers and other workers who construct it, and raises the income of the suppliers of resources such as concrete and steel. The

Traditional public interest model

49

businesses near the bridge have increased trade and property values increase. According to Aranson and Ordeshook, the provision of the public good is a side eﬀect of these transfers of income and wealth. While Aranson and Ordeshook present a positive analysis of the role of government, other economists and political scientists stress the normative aspect and suggest that redistribution can be justiﬁed as a means to reduce income inequality and poverty, and provide insurance against destitution. One of the most inﬂuential studies which addresses the redistributive role of government is John Rawls’ A Theory of Justice (1971). This work considers both the outcome and the process of collective choice. The objective is to establish a set of just institutions in which collective decision-making can take place. It is nowhere explicitly or implicitly implied that the outcome of the process will maximize a social welfare function or be Pareto optimal. Rawls develops a set of basic principles to be applied to the structure of society. These principles are ‘to govern the assignment of rights and duties and regulate the distribution of social and economic advantages’ (1971, p. 61). As Mueller has noted, these principles ‘form the foundation of the social contract, and Rawls’ theory is clearly one of the major, modern reconstructions of the contractarian argument’ (Mueller, 1989, p. 409). The theory consists of two parts. In the ﬁrst part the argument in favor of a contractarian approach is elaborated, and the characteristics of the original position at which the contract is drawn up are described. The second part focuses on the actual principles contained in the social contract. Starting from the premise that social positions and individual attributes are distributed in a random way, Rawls holds that while this initial distribution is neither just nor unjust, in itself, it would be unjust for society to accept this random distribution or to adopt institutions that perpetuate or exaggerate the inequities therein. The objective therefore, is to establish a set of just institutions which will mitigate the eﬀects of the initial random distribution. In order for the process to be impartial, individuals must step through a veil of ignorance regarding their own social position and personal attributes, i.e., they must act as if they did not have this information. After passing through the veil of ignorance, individuals are in an original position of total equality insofar as each has the same information about the potential eﬀect of diﬀerent institutions on his own future position. The original position thus establishes a basis of universal equality upon which the social contract is drawn up. The basis for passing through the veil of ignorance is thus a moral one, in the Rawlsian framework. It is founded on the argument that information about attributes, status and other factors is ‘arbitrary from a moral point of view’ (Rawls, 1971, p. 72). In contrast, having once entered the original position, individuals are to act in their own self-interest. It is assumed that since all individuals have access to the same

50

The political economy of welfare reform in the United States

information, they will all arrive at the same conclusions regarding the just principles that should be contained in the social contract. Unanimous agreement on the terms of the social contract is a direct result of equality of individuals in the original position. Rawls maintains that a contract drawn up on the basis of the procedure just described will contain two basic principles of justice, the liberty principle and the diﬀerence principle. The former holds that each person is to have an equal right to the most extensive basic liberty compatible with the same liberty for others. The latter holds that social and economic inequalities are to be arranged so that they are both (i) reasonably expected to be to everyone’s advantage, and (ii) attached to positions and oﬃces open to all (1971, p. 60). The liberty principle always has precedence over the diﬀerence principle, according to Rawls. He explains this as follows: as the conditions of civilization improve, the marginal signiﬁcance for our good of further economic and social advantages diminishes relative to the interests of liberty, which become stronger as the conditions for the exercise of the equal freedoms are more fully realized . . . as the general level of well-being rises (as indicated by the index of primary goods the less favored can expect) only the less urgent wants remain to be satisﬁed by further advances, at least insofar as men’s wants are not largely created by institutions and social forms. At the same time, the obstacles to the exercise of equal liberties decline and a growing insistence upon the right to pursue our spiritual and cultural interests asserts itself (pp. 542–3).

In this view, liberty is in a sense a luxury good, the demand for which increases with increased income levels. Increased demand for liberty is paralleled with decreased urgency for other goods as needs are met through higher living standards. In addition to the lexicographic ordering of the two principles of justice, the diﬀerence principle itself contains such an ordering, which has been the subject of highly contentious debate. The diﬀerence principle contains a lexicographic ordering of welfare in which the welfare of the worst oﬀ member of society always takes precedence over the welfare of any other individual. Welfare here refers not to a subjective notion such as utility, but rather to primary goods, deﬁned as ‘rights and liberties, powers and opportunities, income and wealth’ (Rawls, 1971, p. 62). The process outlined above is extended by Rawls to apply not only to a preconstitutional stage, but also to the constitutional, parliamentary, administrative and judicial stages of the political process. In each stage the veil is lifted somewhat, but knowledge of speciﬁc individual positions is not allowed, thus preserving impartiality. Proponents of Rawls’ social contract theory claim that the principles derived from the original position are more likely to lead to compliance

Traditional public interest model

51

than those of competing theories. Public goods and externality theories, for example, are often plagued by free rider problems. However, H.A. Hart (1973) has demonstrated that this is not necessarily the case. Hart gives the example of a farmer who has a right to exclude trespassers from his land, and a hiker who has a right to free movement. There is nothing in the liberty principle that dictates which right has priority, and thus compliance cannot be assumed. A similar argument and counter-argument applies to the diﬀerence principle. Rawls argues that compliance with his social contract is more likely than compliance with utilitarianism on the grounds that one could not expect the poor to comply with principles that required them to make sacriﬁces for the rich. Under the diﬀerence principle, however, the rich are required to make sacriﬁces for the poor, a requirement with which they may not wish to comply. In short, as Mueller has noted, Rawls’ social contract and his arguments in support seem to be constructed for the purpose of achieving the compliance of only one group, the worst-oﬀ individuals. A further important criticism of Rawls’ theory has been made by Robert Nozick (1974). Nozick points out that: ‘A procedure that founds principles of distributive justice on what rational persons who know nothing about themselves or their histories would agree to guarantees that end-state principles of justice will be taken as fundamental’ (1974, p. 95). Nozick argues that given so little knowledge about the processes of social and economic interaction, individuals are compelled to ignore any principles that would govern such procedures and focus only on ﬁnal outcomes. Nozick goes on to claim that ‘people meeting together behind a veil of ignorance to decide who gets what, knowing nothing about any special entitlements people may have, will treat anything to be distributed as manna from heaven’ (Ibid.). The question of entitlements is of fundamental importance to Nozick and forms an integral part of his theory of distributive justice. His own theory is that: ‘the holdings of a person are just if he is entitled to them by the principles of justice in acquisition and transfer, or by the principle of rectiﬁcation of injustice . . . If each person’s holdings are just then the total set (distribution) of holdings is just’ (Ibid., p. 49). This theory, in sharp contrast to Rawls’ theory, focuses on the processes whereby a particular distribution came about. If holdings were acquired justly, the holder is entitled to them, regardless of the holdings of any other individual.

NOTE 1. This analysis is based on Silberberg, 1990.

4.

The public choice perspective

In the traditional, organic view of the state outlined in the previous chapter, the government is a benevolent actor instituting policies to correct market failure or public good problems or to obtain a Pareto-optimal allocation of resources. As noted in the previous chapter the Pareto criterion focuses on economic eﬃciency, does not address allocation decisions per se, and could lead to an allocation that is not strictly preferred by any single voter. The traditional viewpoint was challenged beginning in the late 1950s by a number of economists who took a very diﬀerent perspective reﬂected in the public choice literature. The public choice perspective takes the rational choice model characteristically applied in the economic ﬁeld and applies this to decision-making in the political arena. Political actors are assumed to behave in a rationally consistent manner that leads to predictable decision-making processes and predictable outcomes. In contrast to traditional public sector economics which viewed the government as a tool for achieving allocative or distributive ends, public choice research sees the government and its various constituent members as actors in their own right, with their own rational objectives. Thus, legislators, voters, bureaucrats, members of the judiciary, special interest groups, and any other persons engaged in the political process are the subject of public choice analysis, both in regard to their rational choices and the institutions within which these choices are made (Black, 1948; Buchanan, 1949; Downs, 1957; Buchanan and Tullock, 1962; Stigler, 1971; Peltzman, 1976 and 1990; Olson, 1965; Becker, 1983). This chapter and the following chapter summarize the major contributions of public choice theory to the legislative process and the enactment of legislation such as that contained in the new welfare law (Public Law 103–184, the Personal Responsibility and Work Opportunity Reconciliation Act 1996). Section I examines the Median Voter Theorem, upon which much subsequent work was based, and discusses theoretical and empirical criticisms of this model. In Section II the discussion is broadened to incorporate a more complete institutional perspective. The behavior of voters, and speciﬁcally the existence of rational ignorance, is shown to provide opportunities for interest groups and bureaucrats to inﬂuence the legislative process and engage in rent seeking activities such as creating and appropriating larger federal budgets for welfare related programs. 52

The public choice perspective

53

Since welfare programs are often considered ‘ideological’, or based on considerations of equity rather than economic eﬃciency, the role of ideology in the legislative process is also examined.

THE MEDIAN VOTER THEOREM: THE BASIC MODEL One of the pioneering works in the ﬁeld of public choice was Duncan Black’s (1958) seminal work on committee voting procedures, which is the foundation for the median voter theorem, and a great deal of subsequent scholarship which applies economic tools of analysis to political institutional processes. In contrast to the earlier focus of economists such as Bergson (1938) and Samuelson (1947) on deriving aggregate social welfare functions that could be maximized for the good of society by a benevolent government, and Arrow’s (1951) proof that such an exercise was futile, Black focused attention on the actual procedures for aggregating preferences via voting rules. For the ﬁrst time a schema was developed within which majority voting did not necessarily lead to paradoxes such as cycling. Black showed that, under speciﬁed conditions, majority rule will lead to equilibrium outcomes that will coincide with the preferences of the median voter. To arrive at this conclusion, however, required a number of assumptions, which have been the subject of a great deal of criticism in the economics literature. Black begins his argument by supposing that ‘a decision is to be determined by a vote of a committee. Proposals are advanced . . . in the form of motions on a particular topic or in favor of one of a number of candidates’ (1948, p. 133). For convenience he conﬁnes the discussion to motions rather than candidates. He then assumes that ‘each member of the committee ranks the motions in a deﬁnite order of preference’ (Ibid.) and that ‘he votes in accordance with his schedule of preferences’ (Ibid., p. 134). Black claims that: ‘while a member’s preference curve may be of any shape whatever, there is reason to expect that, in some important practical problems, the valuations actually carried out will tend to take the form of isolated points on single peaked curves’ (Ibid., p. 135). This assumption becomes central to the ensuing analysis. Other critical assumptions are that decisions are made by majority vote, and that in the voting each motion is put against every other motion. The method of reasoning employed by Black is illustrated in Figure 4.1, which shows the preference curves of ﬁve members of a committee. The highest point of each curve represents the member’s optimum or most preferred outcome. It can be shown that voting on any two points below the

54

The political economy of welfare reform in the United States

Order of preference

O1 Figure 4.1

O2

O3

O4

O5

Point Set Representing Motions

Preference curves

median, O3, will always result in a majority vote for the point closest to the median since this will always lie on a higher part of the preference curve for all curves to the right. Similarly, voting on points above the median will lead to a majority for the point closest to the median since that point will lie at a higher point on the preference curves to the left. Thus, in a simple majority vote, the motion corresponding to the median, O3, will be the one that is adopted by the committee. Black’s median voter theorem was further developed and extended by Anthony Downs in An Economic Theory of Democracy (1957). Stimulated in part by Arrow’s impossibility theorem, and building on the foundations laid by Black, Hotelling (1929), and Smithies (1941), Downs demonstrated that competition among parties to win votes could lead to an equilibrium outcome of the political process analogous to competition among ﬁrms in the market process. The Hotelling model assumed that people were evenly spaced along a continuum of preferences from left to right on the political scale, and led to the conclusion that competition in a two party system would lead to convergence as each party attempted to obtain more votes by moving along the scale towards its ideological opponent. Smithies introduced the notion that complete convergence would be constrained by the fear of losing supporters at the extremes. Downs introduces into this model a variable distribution of voters, relative ideological immobility and peaked political preferences. The Downs model conﬁrms Hotelling’s conclusions, but suggests that convergence depends on a ‘unimodal distribution of voters with low variance and most of its mass clustered around the mode’ (Downs, 1957, p. 140). Where these conditions are absent, several parties can coexist and there is no automatic tendency for convergence at a median voter position. The Downsian model moves beyond the committee decisions which were the focus of Black’s work, and considers the government as an institution

The public choice perspective

55

made up of individuals such as representatives, bureaucrats and voters, each of whom faces his own constraints and objectives in pursuit of his own self-interest. Downs assumes that political parties and voters act rationally in the pursuit of certain speciﬁed goals. The analogy to economic analysis in which consumers and producers are assumed to act rationally in the pursuit of utility and proﬁts is clear. Pursuing the analogy of the market, Downs demonstrates that competition among parties to win votes can have similar beneﬁcial eﬀects in political markets to the eﬀects of competition among ﬁrms in economic markets. A basic tenet of Downs’ model is ‘that the government exists in a democratic society where periodic elections are held, that its primary goal is reelection, and that election is the goal of those parties now out of power’ (Ibid., p. 11). Formulating and implementing policies are simply means to the end of maximizing votes, and the party that receives the most votes is the one that gains control of the government until the next election. Downs assumes that party members are motivated by ‘their personal desire for the income, prestige and power which come from holding oﬃce’ (Ibid., p. 34) and that the electorate consists of rational voters. When applied in the political arena, the assumption of rational self-interest leads to the hypothesis that ‘parties formulate policies in order to win elections, rather than win elections in order to formulate policies’ (Ibid., p. 28). One of the major contributions of Downs’ work is his discussion of ‘rational ignorance’ – the concept that a rational self-interested individual who maximizes utility will not necessarily ﬁnd it in his interest to become rationally informed regarding issues or candidates. In a world of perfect certainty and knowledge, the rational voter would estimate the expected utility from a government controlled by alternate parties and, assuming the diﬀerence was greater than the cost of voting, vote for that party which provided him with the highest level of utility. With imperfect knowledge and uncertainty – which are integral to the Downs model – the voter must incorporate the costs of becoming informed, which may be substantially higher than any expected utility gain. Moreover, the existence of imperfect information and uncertainty ‘leads to attempts at persuasion by men who provide correct but biased information’, according to Downs (Ibid., p. 94) (clearly it can also lead to persuasion by men who provide very incorrect information). Thus, uncertainty opens the door to competition among political parties who wish to inﬂuence the electorate, as well as interest groups (who claim to represent popular will) and favor buyers (who represent themselves) both of whom want to inﬂuence both government and the electorate. The information disseminated by these groups comes at a price, as Downs notes: ‘they get an inﬂuence over policy formation greater than their numerical proportion in the population’ (Ibid., p. 95).

56

The political economy of welfare reform in the United States

The existence of uncertainty in the political process is also important, according to Downs, in that it contributes to the adoption of party ideologies as a means whereby parties can gain the support of various social groups without estimating returns from speciﬁc policies. Similarly, party ideology gives voters an indicator of how close each party is to their own view of a good society, thus shortcutting the necessity to examine the parties’ individual policies on issues. In order for this process to work, party policies have to consistently follow party ideology, occasionally leading to conﬂict with the ultimate goal of maximizing votes. However, as Downs notes, his hypothesis is upheld as long as ‘parties behave most of the time as though election is their primary objective’ (Ibid., p. 113).

CRITIQUE OF THE MEDIAN VOTER MODEL A large body of critical literature has developed around the median voter model, both theoretical and empirical. The theoretical criticisms focus on the implications for the median outcome of dropping some of the assumptions, including unimodal and symmetric preferences and voting by all voters, while the empirical analyzes deal primarily with methodological problems in the empirical studies (Rowley, 1986; Romer and Rosenthal, 1979a). A detailed critique of the theoretical literature is contained in ‘The Relevance of the Median Voter Theorem’, by Charles Rowley (1986). This paper reviews the implications of the spatial theory for generality, realism and predictive power by reference to the existence, uniqueness and stability of the equilibrium generated by the median voter model. Concerning the existence of equilibrium, the author notes that political elections tend to be held several years apart, and where votes are the equilibrating mechanism, once a position no longer represents an equilibrium, many years may pass with no possibility of movement toward a new equilibrium. Rowley concludes that: ‘discontinuity (of elections) poses an insuperable problem of reconciliation for those who assess relevance by the criteria both of generality and of realism’ (Ibid., p. 110). Additional problems with the concept of existence of equilibrium in the model relate to the paradox of voting and the problem of identifying issue dimensions. In the case of the former, Rowley points out that: ‘if the vote mechanism collapses, political equilibrium will not exist and the spatial approach will be rendered empty’ (Ibid., p. 111). Since a majority of the population do in fact vote, this is not so much a problem for realism in the model as it is for generality, since there is no generally accepted explanation for the voting paradox.

The public choice perspective

57

The most intractable problem for the median voter theorem, according to Rowley is identifying issue dimensions. While the Downsian model is generally presented in terms of choices along a unidimensional ideological issue space, Davis et al. (1970) claimed that spatial theory must allow for multi-dimensional issue space if it is to retain descriptive and predictive power. The practical diﬃculties of identifying and measuring issue dimensions are considerable, however. Moreover, as Rowley notes: ‘the ability to ascertain and to order voter preferences along a set of issue dimensions presupposes a very considerable facility in the conceptualization and measurement of attitudes within a common multidimensional issue space’ (1986, p. 112). After undertaking considerable research, Hinich (1978) eventually concluded that voters collapse issue space into a single liberal-conservative issue space, turning to secondary dimensions only when candidates or parties are indistinguishable using the primary categorization. Rowley summarizes the problem by noting that: ‘it is not even clear either that the relevant issue dimensions can be isolated or that the voter loss metric which relates party preference to party issue space position can be identiﬁed’ (1986, p. 114). He concludes that, ‘In such circumstances spatial theory looks extremely suspect, whether judged by criteria of generality, of realism or of predictive power’ (Ibid.). Even where an equilibrium could be posited to exist, there remain problems of uniqueness and stability. Rowley demonstrates that uniqueness is highly sensitive to the assumptions of the Downsian model. When the assumptions of all voters voting is relaxed to allow some voters to abstain through alienation, uniqueness of the equilibrium ‘is rendered doubtful’, while the existence of multi-party competition may give rise to strategic voting which would also confound the median voter equilibrium solution (Ibid., p. 117). The problem of instability where preferences are not single peaked was noted by Downs himself, who pointed out the possibility of cycling among alternative preferred voter positions. Rowley notes that this problem intensiﬁes as the issue dimensions expand, which can result in cycling over the diﬀerent dimensions even though preferences within each dimension may be single peaked. The result will be instability of equilibrium in the multidimensional space (Ibid., p. 119). The numerous theoretical diﬃculties of the median voter model have not deterred its use in a wide range of empirical studies. Some of the limitations of the model are, however, implicitly illustrated by the fact that many of these studies chose issues which avoid the areas of weakness. The problem of unidimensional issue space, for example, is easily circumvented by choosing a single issue election, and single issue candidates, as is the case with school board elections and school funding.

58

The political economy of welfare reform in the United States

Empirical models used to test the median voter theorem generally assume that voters maximize utility subject to a budget constraint that incorporates their tax price for the public good in question. Thus, demand by the median voter for the public good is a function of median income, median tax price and a ‘ “vector of taste parameters” (number of children, Catholic or non-Catholic, etc.)’ (Mueller, 1989, p. 190). This model is generally tested using cross-sectional data on local expenditures for the public good being studied. Randall Holcombe (1989), for example, uses data at the school district level for 257 Michigan school districts, and ﬁnds that the theoretical median voter model which he develops in the paper provides a good explanation of empirical reality. Similarly, Denzau and Grier (1984) claim support based on their analysis of data on New York school districts. Other studies have modeled municipal expenditures in ten American states (Bergstrom and Goodman, 1973), school expenditures in Long Island (Inman, 1978), school expenditures in Connecticut, and a number of studies have tested the model using data from Switzerland (Pommerehne and Frey, 1976; Pommerehne, 1978; Pommerehne and Schneider, 1978). In addition, as Romer and Rosenthal (1979a) note, assumptions leading to median voter dominance have been employed to incorporate political processes in a wide range of economic contexts, including pollution control, income redistribution, minimum wage legislation and union behavior. While most of these studies claimed support for the median voter theorem, an extensive review of the empirical work based on the median voter model undertaken by Romer and Rosenthal (1979a) concluded that the studies examined failed to indicate that actual expenditures corresponded to those desired by the median voter. While admitting that the empirical studies generally show that ‘operational measures’ of median income and median tax price can statistically account for expenditures, the level of desired spending may in fact not be that actually desired by the median voter, but instead a multiple of this level. The Barr and Davis (1966) study is illustrative of this point. In their original study, Barr and Davis assumed identical tastes, and initially identical incomes (although this assumption was later dropped) for voters in a cross-section of Pennsylvania counties. Thus the median voter in one county would diﬀer from his counterpart in another only in eﬀective tax price. As a surrogate for tax price the authors used assessed property values per capita and owner occupied residences per registered voter. The statistical results were as hypothesized with high tax prices leading to low expenditures. As Romer and Rosenthal note, however, the fact that expenditures rise as price falls does not mean that expenditures are at the level preferred by the median voter; they could be 50 percent less or 100 percent

The public choice perspective

59

greater than those desired by the median voter (1979a, p. 148). This is what the authors term the ‘multiple fallacy’. Romer and Rosenthal also point out the possibility of what they describe as the ‘fractile fallacy’ (Ibid.). This exists because there is nothing in the model which provides evidence that the use of the median gives superior results to the use of the mean, the 25th, the 75th or any other percentile. In the same vein, these authors also noted that the median voter model is rarely tested against alternative theoretical or statistical models that might provide alternative and superior bases for the phenomena being analyzed. One of the strongest arguments Romer and Rosenthal present against the use of median voter models is that even if the median voter is pivotal, the outcome of any policy decision may not be that which would be preferred by this voter. That this may be the case was suggested by two studies that examined voting in referenda for budgets proposed by a school board against speciﬁed reversion levels (Barkume, 1976; Rubenfeld, 1977). In this case the median voter does not necessarily have the choice of his preferred expenditure. If political institutions are such that the ideal point is not placed on the ballot, those institutions can thwart the will of the median voter. Romer and Rosenthal conclude that ‘expenditures depend not only on the preferences of voters but also on the structure of political institutions’ (1979a, p. 143). They suggest that the presence of bureaucratic threats can result in expenditures signiﬁcantly in excess of those desired by the median voter, and point to the practice of school boards which can propose alternative budgets, none of which represent median voter preferred spending levels. As supporting evidence for the relevance of institutional factors they note that while the simple median voter model has the best ﬁt for direct democracies, a model that incorporates variables for the complexity of the revenue system and for the time before the next election greatly improves the ﬁt for models of representative democracy (Ibid., p. 160). Complex revenue systems lead to signiﬁcantly higher expenditures, and the more time before an election, the greater the expenditure. Both variables reﬂect information cost problems, and the authors suggest that the town meeting and frequent referenda of direct democracy serve to keep citizens better informed and to allow for a free amendment process. This line of reasoning suggests that governmental agencies or bureaus may seek to maximize budgets, a proposition which is strongly supported in the work of Niskanen (1971), discussed in chapter five. The weaknesses of the median voter theorem outlined above limit the usefulness of this construct in analyzing the political decision-making process and policy outcomes. This lacuna is partially ﬁlled by the new institutional economic approach which attempts to: ‘analyze the behavior of

60

The political economy of welfare reform in the United States

incumbent governments and their opposition not merely at the point of election but throughout expected life both in and out of oﬃce, within the speciﬁc constraints of the constitution they inherit, and subject to all the self-seeking pressures to which they are exposed’ (Rowley, 1986, p. 125). A key component of this area of study is the behavior of voters, upon whom the other agents are dependent for their positions and power.

THE NEW INSTITUTIONAL ECONOMICS: VOTERS, THE PARADOX OF VOTING, AND RATIONAL IGNORANCE The behavior of voters and speciﬁcally the paradox of voting casts doubt on the robustness of models which assume voters are fully informed and rational. As Downs (1957) and later Riker and Ordeshook (1968) have noted, rational behavior will normally lead to non-voting and yet most of the population does vote. The argument is based on a simple calculus of expected costs and beneﬁts, which can be expressed as: R (BP) C where R is the reward or utility a voter receives from the act of voting, B is the diﬀerential beneﬁt the candidate receives from the success of his more preferred candidate over his less preferred one P is the probability that the voter will, by voting, bring about the beneﬁt, B, and C is the cost to the individual of the act of voting. Clearly where there is a large number of eligible voters, the probability of any one being decisive is extremely small. Riker and Ordeshook (1968) suggest that for a national election in the US, P might be 108, for example. Since this would render the term BP in the equation very small, and since the costs of voting, including time spent becoming informed of the voting choices, time spent going to the polls etc., are generally signiﬁcant, the value of R is generally assumed to be negative. Thus making it irrational for the voter to vote. Several explanations have been proﬀered for the apparent paradox of voting. Most of these claim that voters are motivated by a consumption or investment motive. Riker and Ordeshook, for example, stress the former. They claim that voters gain satisfaction or utility from ‘compliance with the ethic of voting’, ‘aﬃrming allegiance to, and eﬃcacy in, the political system’, and ‘aﬃrming a partisan preference’, as well as social satisfaction

The public choice perspective

61

from becoming informed and the act of going to the polling booth (Ibid., p. 28). This argument is analogous to that of Downs (1957) who argued that voters vote in order to preserve the political system. Unlike Riker and Ordeshook, however, Downs felt that their decision was independent of their own short run gains and losses. An alternative explanation for the voting paradox is oﬀered by Barzel and Silberberg (1973), who consider marginal rather than average probabilities in the voting calculus. Estimating an econometric model of voter turnout in gubernatorial elections, these authors ﬁnd that voter turnout increases the greater the probability of a vote being decisive. They conclude that voter turnout can in fact, ‘be explained, in part, on the basis of rational, wealth maximizing, behavior on the part of the electorate’ (Ibid., p. 51). George Stigler pursues a similar line of argument, suggesting that ‘if election outcomes are not all-or-nothing (49 percent is defeat) and instead inﬂuence is a monotonically increasing function of vote share, then the probability that one’s vote will make a diﬀerence is unity, not some inﬁnitesimal fraction’ (1972, p. 104). Noting further that additional votes may give a party additional inﬂuence, Stigler concludes that the investment motive is far more convincing than the consumption motive as an explanation of the voting paradox (Ibid.). In a sense, however, the issue of what motivates the act of voting is moot. The fact is many citizens do vote. Of greater interest and signiﬁcance for the political process is the behavior of voters after arriving at the booth, and speciﬁcally the degree of rationality in their actual vote. Given that they are going to vote, for consumption, investment or whatever other reason, the question arises as to how the calculus is made regarding the costs and beneﬁts of voting on a speciﬁc issue or for a speciﬁc candidate. Clearly, this involves estimating the expected costs of gathering information on candidates or policies and the expected costs or beneﬁts associated with a particular incumbency or implementation of a particular policy. A rational choice model would clearly imply an inverse relationship between information costs and the degree of rationality in a particular vote. In order to make a rational voting decision, a voter needs to obtain and process information regarding the potential costs and beneﬁts that will accrue to him as a result of alternative outcomes of a speciﬁc vote. The more accurate assessment he wishes to make, the more information he will require and the more information processing he will need to undertake, hence the more costly the process will be. Assuming the equimarginal principle applies, a rational voter will invest in such activities up to the point at which the expected return per dollar (or dollar equivalent unit of time) spent in information gathering and processing activities is equal across issues or candidates. Thus, where the potential returns are high, the voter

62

The political economy of welfare reform in the United States

is likely to invest signiﬁcantly more and be far better informed than when he has little at stake. Similarly, if the costs of obtaining information on certain issues are relatively less costly, the voter is likely to obtain more information, since he will equate marginal return per dollar. This line of analysis has several important implications. In the ﬁrst place, it implies a continuum of rational ignorance rather than the binary approach often found in the literature. In the second place it suggests that voters will be better informed about local issues than national ones, since national sources of information are supplemented by local media, local informational agencies, and personal contacts and experiences, thus generally making information on such issues more readily available. Thirdly, it implies that voters will be relatively better informed, and thus vote more rationally, on issues such as inﬂation and unemployment which are well covered by the media and other information sources, thus providing cheaper information. What evidence is available to support the foregoing hypothesis? A number of empirical studies suggest that voters remain ‘rationally ignorant’ of both issues and candidate positions in elections (Mueller, 1989). D.E. Mayhew (1974), for example, carried out a survey in which he found that ‘only about half the electorate, if asked, can supply their House members’ names’ (p. 49). This result has been interpreted as showing that voters are not rational. However, our theory would suggest that voters have no incentive to learn the names of candidates unless those candidates took positions that would result in a signiﬁcant gain or loss for the voter. A notable study by Peltzman (1990) concluded that voters are highly eﬃcient in their use of information on unemployment and inﬂation. In this model voters accurately distinguished permanent from transitory eﬀects, ignored irrelevant information and used all information that was relative. This high degree of eﬃciency or rationality is consistent with our argument that voters will exhibit a high degree of rationality in voting on those issues which directly aﬀect them and for which information is readily available. Inﬂation and unemployment data are readily available through the media and households have direct experience of these issues both as consumers and employees and are directly aﬀected by them. Hence rationality is likely to be far greater on these issues than, say, national defense or transfer programs which do not directly aﬀect an individual voter. The greater the degree of rational ignorance, the greater are the possibilities for lobbyists, interest groups, bureaucracies and others to inﬂuence the political process to their own advantage. From the point of view of these groups, potential opposition costs are minimized when the costs of a particular policy are spread over a large number of voters, and information regarding these costs is diﬃcult or costly for the voter to obtain. Additional

The public choice perspective

63

evidence for the equimarginal rational voting model is thus suggested by the phenomenon of government subsidization of special interest projects at the expense of the majority of voters. Clearly, if the cost of subsidizing the few is spread over many voters, there is little loss to any individual and little incentive to become well informed. Moreover, the existence of ﬁscal illusion renders the information costs high, thus further reducing the incentive for the voter to become well informed (Oates, 1988). The existence of rational ignorance thus plays an enabling role in the development of public policies. To the extent that all the agents involved are rational self-interested individuals, they will seek policies that result in personal gain. For legislators this is assumed to be votes and the support of special interest groups. Since the interest groups generally comprise a minority of voters, whose gains are at the expense of the majority of taxpaying citizens, lack of knowledge or interest on the part of voters is requisite for passage of legislation favoring the minority group. The role of interest groups and the mechanisms whereby policy may be inﬂuenced by them, is further discussed below.

THE ROLE OF INTEREST GROUPS The role of interest groups has been modeled, inter alia, by Mancur Olson (1965) and Gary Becker (1983). Olson’s basic premise is that members of interest groups are rational decision-makers and will only contribute to such a group if the beneﬁt they receive is greater than the cost of contributing. Thus, even if all the members of a group would gain from the achievement of a group objective it does not follow that rational members of the group would act to achieve these ends. In fact it would be inconsistent for rational self-interested individuals to act to achieve their group interests in the case of large groups. This is so because ‘the achievement of any common goal or the satisfaction of any common interest means that a public or collective good has been provided for that group’ (1965, p. 15). A public good has two basic characteristics, non-excludability and jointness in supply. Olson focuses on the former which he deﬁnes as ‘any good such that, if any person Xi in a group X1, . . . Xi, . . . Xn consumes it, it cannot feasibly be withheld from others in that group’ (p. 14). Since any individual in the group can enjoy the group beneﬁts whether or not he has contributed, the rational self-interested individual will not act to achieve group beneﬁts. Moreover, ‘even if the member of a large group were to neglect his own interests entirely, he still would not rationally contribute toward the provision of any public or collective good since his own contribution would not be perceptible’ (Ibid., p. 64).

64

The political economy of welfare reform in the United States

Large groups must, therefore rely on alternative strategies to attract members. A large group can attract participation through the use of ‘an incentive that operates, not indiscriminately, like the collective good, upon the group as a whole, but rather selectively toward the individuals in the group’ (Ibid., p. 51). Selective incentives may be negative, such as the coercion used by some union groups, or positive such as the provision of publications, insurance, social and other activities provided by groups such as the American Association for Retired Persons (AARP). Where membership relies on these incentives, lobbying is, in a sense a by-product. The case of small groups is both qualitatively and quantitatively diﬀerent from that of large groups according to Olson’s theory. Olson shows that some small groups can provide themselves with a collective good without the use of selective incentives or coercion. This will be the case if at least one member of the group ﬁnds that his personal gain from having the collective good exceeds the cost of paying for it. This line of thought leads to the implication that groups with the greatest inequality in potential beneﬁt are most likely to be able to provide the good since the greater the interest of any single member, the greater the likelihood that it will be proﬁtable for him to provide it. A further result of this tendency is what Olson describes as the ‘exploitation of the great by the small’, i.e., those members who have a smaller share in the beneﬁts have incentives to free ride and take advantage of the investments of the member with the larger share (Ibid., p. 35). An additional feature of small groups which facilitates action to obtain collective goods is the potential for social pressure. In groups which are small enough that the members know one another social pressures may lead to participation, even though the individual would otherwise rationally choose not to act in the group interests. Among the most inﬂuential of the groups to aﬀect welfare policies are the AARP, which has successfully lobbied for more programs for the elderly, and the American Medical Association (AMA), which has successfully lobbied against health insurance. Both of these organizations ﬁt quite well into the Olson theory, for while they are both large organizations, they each provide numerous individual beneﬁts to members. The AARP provides several types of insurance and a wide range of reduced price goods and services to its members, in addition to newsletters keeping them informed of proposed legislative changes and how they would be aﬀected. The AMA also provides insurance, research, and legislative and other information to its members. The AMA, however, has been suﬀering a decline in recent years in terms of the percentage of doctors who participate. This phenomenon illustrates a lacuna in Olson’s theory. He does not adequately address the issues of

The public choice perspective

65

growth or decline of groups. Nowhere, for example does he describe how a small group, which he claims is qualitatively diﬀerent from a large one, turns into that large group. Conversely, he does not explain the decline over time in the inﬂuence of groups. One potential cause in the case of the AMA is the competition from other organizations oﬀering similar products to the selective incentive. Presumably, any lobbying group is of necessity at a disadvantage vis-à-vis a private enterprise which does not have to fund lobbying activities out of its resources, and thus could oﬀer the same ‘beneﬁt’ at a lower price (Stigler, 1974). Clearly this is an empirical question, which cannot be fully addressed here. However, the question of competition between interest groups is discussed below. In addition to problems with explaining dynamic processes of groups, Olson’s theory fails to adequately deal with certain classes of large groups such as welfare advocates, and civil rights groups, which neither use coercion nor provide selective goods or services for their members. According to the Olson thesis then, their members are acting irrationally. Yet clearly such groups have ﬂourished at diﬀerent times in the history of social welfare. An alternative theory of interest groups is that of Gary Becker who claims that ‘Individuals belong to particular groups . . . that are assumed to use political inﬂuence to enhance the well-being of their members’ (1983, p. 372). In this model optimization is at the level of the group rather than the individual, and the optimal expenditure by the group on inﬂuencegaining activities will be equal to the returns from such activities at the margin. The results of competition between groups for political inﬂuence will, according to Becker, be an equilibrium structure of taxes, subsidies and other political favors. Becker assumes two homogenous groups, taxpayers (t) and subsidy recipients (s), and a balanced budget in which total taxes (net of deadweight costs) equal total subsidies (also net of costs). The amount raised by all taxes on t can be written as: S ntF(Rt ) where n is the number of members of t, and R is the taxes paid by each member. The function F is the revenue from a tax of R and incorporates the deadweight costs that result from the ‘distorting eﬀects of taxes on hours worked, investments and other taxpayer choices’ (p. 374). Similarly, the subsidy to each member of s is given by: nsG(Rs ) S ntF(Rt )

66

The political economy of welfare reform in the United States

that gives the budget equation, and in which ns is the number of members and Rs is the subsidy to each member. G is the cost of providing Rs and incorporates the deadweight costs from the distorting eﬀects of subsidies on hours worked, investment and other choices by recipients. The amount raised in taxes is determined by an inﬂuence function, and inﬂuence is itself a function of pressure (p) exerted by the two groups as well as other, unspeciﬁed, variables (x): ntF(Rt ) It (ps, pt, x) Likewise the amount available to subsidize s is determined by an inﬂuence function nsG(Rs ) Is (ps, pt, x) The assumption on the budget equation gives the result that the aggregate inﬂuence of the two groups is zero, and so the political game is ‘zerosum in inﬂuence and negative-sum in taxes because of deadweight costs’ (Ibid., p. 376). The deadweight costs that Becker introduces into his model reﬂect the distortions that are caused by taxes and subsidies. These have an asymmetric eﬀect on taxed and subsidized groups, since taxed groups will be adversely aﬀected by the additional expense, and thus stimulated to exert increased pressure, while subsidized groups are less aﬀected since their subsidy is reduced by the deadweight loss. The implications of Becker’s analysis are that politically successful programs are eﬃcient, and that competition among pressure groups favors eﬃcient methods of taxation since both groups are better oﬀ when deadweight costs are reduced. One of the major criticisms of the Becker model relates to its dependence on several quite stringent assumptions. A fundamental assumption is the political budget equation, in which taxes must equal subsidies net of deadweight costs, which implies that the net inﬂuence of pressure groups is zero. It is diﬃcult to reconcile this conclusion with the huge budget deﬁcits throughout most of the 1980s and early 1990s which suggest a rather large imbalance between taxpayers and subsidy recipients. Of course it could be argued that the eﬀorts to balance the budget in the late 1990s are precisely what Becker foretold, and the previous years merely represented movement towards this equilibrium. But it is diﬃcult empirically to conclude that taxpayer clout has undergone such a major transformation in the past few years. A long period of strong growth in the economy and the end of the Cold War would seem to be more likely places to look for causes of a balanced budget.

The public choice perspective

67

Similarly problematic for the Becker thesis is the growth in the overall size of government. If competition among interest groups leads to eﬃciency, and if the taxed and subsidized groups are zero-sum in inﬂuence, then whence the growth in government? Competition should lead to decreases in deadweight costs, as the most eﬃcient groups are most successful in exerting pressure, and thus a decline in the size of the government sector. Becker’s theory nowhere explains from whence comes the pressure for increases in total government spending, as opposed to spending on individual programs, as a result of group inﬂuence. Casual empiricism seems to indicate a positive-sum result to the process. A ﬁnal diﬃculty with this model is the focus on the group rather than the individual. In Olson’s model the individual calculus of cost and beneﬁts is the driving force and constitutes a fundamental reason for groups acting, or failing to act, in the collective group interest. This is ignored by Becker who simply assumes that ‘individuals belong to particular groups’, without articulating the decision-making process (1983, p. 372). Relatedly, Becker attaches very little importance to free riding, claiming that the emphasis on free riding in many discussions of the eﬀectiveness of pressure groups is a ‘little excessive’ because political success is determined by the relative, not absolute, degree of control over free riding (Ibid., p. 380). The relative merits of the alternative theories of interest group behavior for understanding the legislative process vis-à-vis welfare policies is further developed below. However, before dealing with this issue it is necessary to note that the eﬀectiveness of such groups in obtaining their demands from legislators may be enhanced or constrained by the federal or local bureaucracies involved at certain stages in the formulation and implementation of such policies. A rational choice model would clearly suggest that where the objectives of these two groups coincide, a legislative initiative could be expected to be highly eﬀective. Where the groups have conﬂicting objectives, however, compromises could be expected. Further, in analyzing the role of bureaucracies, it is crucial to recognize that they too, can be considered a species of interest group. The role of the bureaucracy is discussed in more detail in the next chapter, which also analyzes the role of the legislative branch and the Presidency. Before turning to these issues, however, this chapter explains the rent-seeking process and the inﬂuence of ideology in the legislative process.

RENT SEEKING The practice of rent seeking was ﬁrst analyzed by Gordon Tullock in ‘The Welfare Costs of Tariﬀs, Monopolies and Theft’ (1967). Tullock’s basic

68

The political economy of welfare reform in the United States

Price

P1 R

L

P2

Q1 Figure 4.2

Q2

Quantity

The cost of rent seeking

insight was that if there are monopoly proﬁts to be gained, a monopolist will invest resources in acquiring these proﬁts, which resources constitute a social cost in addition to the welfare cost of the monopoly. Thus, in Figure 4.2, triangle L represents the eﬃciency loss due to monopoly, and traditionally, rectangle R was considered merely a transfer from consumers to producers. Tullock, however, pointed out that a rational monopolist would be willing to invest resources up to the value of R to obtain this monopoly. To the extent that these resources could have been employed productively elsewhere, such a use constituted a social welfare loss. The rent creating powers of government and the rent seeking eﬀorts of the private sector in the area of government regulation were analyzed by George Stigler (1971). Stigler’s central thesis is that, in contrast to the traditional view which holds that regulation is primarily a method of protecting consumers, regulation is in fact ‘acquired by the industry and is designed and operated for its beneﬁt’ (Ibid., p. 3). Since the state has a unique power to coerce, in the form of taxing and regulating citizens, it can be utilized by industry to increase proﬁtability. Stigler contends that there are four types of policies which an industry may seek: 1) cash subsidies (often not sought because they would have to be divided among an increasing number of competitors), 2) control over entry by new rivals, 3) policies which aﬀect complements and substitutes (support production of the former and suppress the latter), and 4) price controls which will allow price discrimination (Ibid., pp. 3–4). In return for granting these regulations, the politicians receive votes and resources, including campaign contributions, contributed services, and educational campaigns on certain issues. These constitute rent seeking costs for the industry seeking the regulation. Stigler undertakes an econometric analysis of the trucking industry to examine his thesis that regulation primarily beneﬁts the regulated industry

The public choice perspective

69

and ﬁnds strong statistical support. The evidence is less clear in his analysis of occupational licensing. Building on the foundation laid by Stigler, and noting that what is basically at stake in regulatory processes is a wealth transfer, Sam Peltzman (1976) developed a model of wealth transfers incorporating both producers and consumers. In this model control of regulation or taxation rests on direct voting, and the beneﬁciary pays with votes and dollars. The regulator politician seeks to maximize his majority, M, deﬁned by the function: M nf (N n)h where n number of potential voters in the beneﬁciary group f (net) probability that a beneﬁciary will grant support N total number of potential voters h (net) probability that he who is taxed (every non-n) opposes. Since gainers and losers face transactions and information costs, f and h lie between zero and unity, depending on the amount of the group member’s gain or loss. The probability of support, f, is speciﬁed as: f f(g) Where g is per capita net beneﬁt, deﬁned as: g

T K C(n) n

with: T total dollar amount transferred to the beneﬁciary group Kdollars spent by beneﬁciaries in campaign funds, lobbying and so on to mitigate opposition C(n)cost of organizing both direct support of beneﬁciaries and eﬀorts to mitigate opposition. The number of votes in support of the politician depends on n in two oﬀsetting ways: a larger n provides a broader base of support, but it also reduces the net gain per member of the group, and so the probability of their support. Peltzman assumes that the wealth transfer, T, is generated by a tax, t, on the wealth, B, of each member outside the beneﬁted group: T tB(N n), or t

T B(N n)

70

The political economy of welfare reform in the United States

Opposition is generated by the tax rate and mitigated by voter education expenditure per capita (z), so: h h(t, z), z

K (N n)

In this characterization of the political process, the regulator politician must choose the size of the group that will beneﬁt (n), the amount they will ask the group to spend to mitigate opposition (K), and the amount they will transfer to the beneﬁciary group (T). The results of this formalized model may be summarized as follows: 1.

2.

3.

The size of the winning group will be restricted due to the costs of organizing and due to the fact that imperfect information renders it diﬃcult to translate a tax into political opposition, inducing the regulator to tax the many and concentrate favors on the few; If groups organize according to an economic interest (producers v. consumers) political entrepreneurship will lead to members of the losing group being included in the winning group, and The winning group will not obtain the maximum possible gain from the regulation because of the activities of opposition groups. For example, the vote maximizing politician may favor the regulated industry’s producers, but will stop short of setting prices at the rent maximizing level due to the loss in votes from utility maximizing consumers.

While the studies cited above have concentrated primarily on the industry regulation, it is clear that public goods may also give rise to rent seeking activities. As Mueller has noted, ‘The entire federal budget can be viewed as a gigantic rent up for grabs for those who can exert the most political muscle’ (1989, p. 243). Government contracting is one avenue of potential exploitation, and federal relief programs are another. Wallis (1986), looking at data from the 1930s, found that large states used their numerical advantages in the House of Representatives to obtain greater shares of federal relief programs than the Senate was willing to allow them. This phenomenon is considered in more detail in the context of the New Welfare Law in later chapters.

THE INFLUENCE OF IDEOLOGY ‘Laws may be passed because of self-interest or because of ideology’, according to Kau and Rubin (1979). In fact, it seems more plausible that

The public choice perspective

71

many laws are passed as a result of a combination of these two motivating factors. Identifying the role of ideology in the legislative process has proved illusory, however, despite the development of a relatively extensive literature incorporating a substantial number of empirical analyses (Silberman and Durden, 1976; Danielson and Rubin, 1977; Kalt and Zupan, 1984; Dougan and Munger, 1989 and Goﬀ and Grier, 1993). One of the earliest analyses of ideology was the study by Kau and Rubin (1979), which sought to measure the eﬀect of ideology on roll call votes in the Congress. Ideology is measured in this study on the basis of a Congressman’s ratings by the Americans for Democratic Action (ADA). Using 26 votes cast in 1974 and covering a wide range of issues, the authors found that in all cases but two, the ideological variable was signiﬁcant. While certain problems exist with the measurement used [the authors note correctly that ‘economists do not have a well developed theory of the nature of ideology’ (Ibid., p. 384)] and with the possibility of omitted economic variables, the strength of the results requires that the ideology variable be seriously considered in any analysis of the legislative process. The hypothesis that ideology is a determinant of Congressional votes is also supported by the work of Kalt and Zupan (1984). These authors examine voting on strip mining issues and conclude that ‘the evidence thus far suggests the need for some broadening in the economic theory of politics’ (Ibid., p. 298). They note that ideology is ‘unusually diﬃcult to identify’ and is, moreover, often considered empirically unimportant. In spite of this they claim that ‘approaches which conﬁne themselves to a view of political actors as narrowly egocentric maximizers explain and predict legislative outcomes poorly’ (Ibid., p. 279). Other studies discount the inﬂuence of ideology. Goﬀ and Grier (1993) claim that the empirical techniques used in studies such as those of Kalt and Zupan (1984), and Kau and Rubin (1979) are not empirically sound and thus their results are invalid. Peltzman (1982), using a diﬀerent research methodology ﬁnds little support for the hypothesis that ideology is a signiﬁcant determinant of voting in the Senate. Anderson and Tollinson conclude that while ideology played a part in the repeal of the Corn laws ‘there is reason to believe simple economic self-interest actually dominated’ (1985, p. 211). A similar qualiﬁed level of support is found by Nelson and Silberberg who analyze votes in the 97th Congress on defense expenditures and ﬁnd evidence that ideology has greater explanatory power in general as opposed to speciﬁc expenditure bills (1985, p. 23). A diﬀerent approach is taken by Downs who sees ideology as a means for voters to economize on information: ideological reputations of politicians provide low cost measures of likely future voting patterns (Downs, 1957). Along similar lines, Dougan and Munger: conclude that ‘Even in a world

72

The political economy of welfare reform in the United States

of fully rational and non-ideological voters, there appears to be a potential role for ideology to play in mitigating the adverse eﬀects of rational ignorance on legislative behavior’ (1989, p. 139). As the above discussion shows, there is no consensus on the role of ideology. If indeed it is an important explanatory variable, it could be expected that it would play a particularly strong role in an area such as welfare policies, for while some programs, such as food stamps, provide beneﬁts to vested interests, cash transfer payments would appear to have minimal impact on such interests.

RELEVANCE OF EXISTING THEORIES TO THE DEVELOPMENT OF WELFARE PROGRAMS Traditional theories of the public sector, which view government as a benevolent dictator implementing polices to maximize the welfare of society, are fraught with diﬃculties both of a theoretical and empirical nature when confronted by the realities of the existing political process in the United States today. As we saw in the ﬁrst chapter, this normative view of government does not correspond with the positive analysis of how the government carries out its business, and leaves much to be explained. Alternative theories, which generally take a positive rather than a normative approach, are to be found in the public choice literature. Predominant among them are the median voter model and the new institutional economics. The former has contributed substantially to a more complete analysis that incorporates key institutional factors. However, it too has serious theoretical and empirical ﬂaws, as discussed earlier in this chapter. We are left, then, with the so-called new institutional economics, which predicts rational self-interested behavior by individuals and groups who stand to gain through participation in the market for legislation. The remainder of this section examines the relevance of this approach and speciﬁes what predictions it might lead to regarding legislation on welfare policies and programs. A key actor in the market for legislation is the supplier of the product, the vote maximizing legislator. Other things being equal, this individual is likely to favor those individuals and groups which can deliver the most votes, directly through their own votes and those of their members, or indirectly by providing ﬁnancial or organizational support, inﬂuencing the media, and so on. The purchase price of the legislation is then the expected number of votes delivered. Thus, a group such as the elderly, who have a high propensity to vote, would be expected to be able to purchase legislation providing them with government transfers more eﬀectively than would

The public choice perspective

73

an equally numerous, but non-voting group of children. And, in fact, a glance at the trends in cash beneﬁts transferred to each of these groups over the past several decades provides some support for this contention. When looking at transfer programs however, it is important to bear in mind that the ultimate, or ostensible, target recipients of such programs are not the only, or necessarily the primary, beneﬁciaries. In the case of health care programs, for example, such as Medicaid and Medicare, the health care and insurance interests stand to gain substantial pecuniary beneﬁts. Similarly, food stamp programs beneﬁt a chain of interests from agricultural producers to retail grocers. This is also true of direct cash payments to individuals, such as those issued under the new welfare program Temporary Assistance to Needy Families (TANF), which replaced Aid to Families with Dependent Children (AFDC) in 1996. The beneﬁciaries of such programs can be categorized into three groups: (1) special interest groups, (2) philanthropic groups and (3) altruists. The ﬁrst category, special interest groups, contains several sub-groups: (1) providers of services such as health care, child care or transportation to the welfare recipients; (2) program administrators and bureaucrats at the federal, state and local level; and (3) private consulting ﬁrms receiving contracts to evaluate the program. These groups are, in general, the most well organized and ﬁnanced and stand to gain or lose substantially from the legislation. Therefore, a rational choice model would predict that they would be highly active and highly eﬀective in their bid for beneﬁcial legislation. The second group consists of philanthropic entities, such as churches, that have a history of providing welfare beneﬁts which long predates provision of such beneﬁts by the state. These organizations are generally not as well-ﬁnanced or organized (for political lobbying purposes, at least) as the ﬁrst group, and any pecuniary gains from legislation may be less direct and less signiﬁcant in terms of their overall operations. They might thus be expected to be less eﬀective. However, such organizations often have a strong ideological component, which potentially compensates for a weaker pecuniary motive for actively seeking welfare legislation. The ﬁnal group is comprised of those altruists who, based on ideological principles, provide votes and direct ﬁnancial support in order to obtain a policy which they believe in, but which is not of any direct and concrete economic beneﬁt to them. This group may or may not have economic resources to contribute, and may have no pecuniary incentives to demand welfare legislation, but may have strong ideological motivation. In addition to these broad groups of potential beneﬁciaries of welfare legislation, welfare recipients themselves stand to gain cash beneﬁts, while legislators have at stake votes and other forms of support.

74

The political economy of welfare reform in the United States

The primary potential losers from welfare legislation are taxpayers, who are generally not well-organized as a group, and may have little or no incentive to inﬂuence welfare legislation if the expected returns are less than the expected costs of obtaining and processing information on welfare legislation and participating in the political process. Public choice theory would predict that the policy outcome would reﬂect the balance of inﬂuence among potential losers and gainers, and that any redistribution of resources would result in a larger slice of the federal pie going to those most capable of inﬂuencing the legislative process. In this view, the attributes and strategies of the various interest groups are key determinants of policy outcomes. It is necessary, therefore, to evaluate the contribution of existing interest group theories to an explanation of the development of welfare legislation. The theories of Olson (1965) and Becker (1983) have made signiﬁcant contributions to the understanding of the eﬀect of interest group activities in the economy in general. Olson’s theory serves well to relate the rational self-interested behavior to the formation of interest groups, and to show how this behavior aﬀects groups of diﬀerent sizes. His model is supported by empirical evidence of the eﬀectiveness of certain large groups such as the AARP, relative to, for example, groups of consumers or taxpayers. Olson’s model breaks down, however, in the case of one type of group which is particularly important in the study of welfare policies. Large groups such as civil rights and welfare advocates, who have been major lobbying forces for welfare policies, are generally characterized by neither selective incentives nor coercion of membership. Olson addresses the question of such groups, which he terms ideological groups, but dismisses their importance. He notes that there is ‘the logical possibility that groups composed of either altruistic individuals or irrational individuals may sometimes act in their common or group interests’, but, he claims, ‘this logical possibility is usually of no practical importance’ (1965, p. 2). To justify this viewpoint, Olson takes the example of the state. He notes that despite one of the strongest motivating factors, patriotism, the state must rely on compulsory taxation. What Olson ignores is the fact that in times of war, i.e., when there is a cause in which people believe, hundreds of thousands of citizens will oﬀer not only their taxes but their lives. Certainly there will be the free riders, the draft dodgers and deserters, but surely this example indicates the potential strength of ideological motives to aﬀect human behavior on a large scale. While Olson admits the possibility of ideological groups, he cites three reasons for not using them to explain any examples of group action. First, it is not possible to get empirical proof of people’s motivations, and thus the theory is untestable. Second, ‘there will be suﬃcient explanations on

The public choice perspective

75

other grounds for all the group action that will be considered’ (Ibid., p. 61). Third, ‘most organized pressure groups are explicitly working for gains for themselves, not gains for other groups, and in such cases it is hardly plausible to ascribe group action to any moral code’ (Ibid.). None of these arguments provides a convincing reason for ignoring ideological groups. Regarding the ﬁrst, it would surely limit the possibilities for research if only those theories which were empirically testable were to be developed. Moreover, while a scientiﬁc test may not be available, there is certainly a preponderance of evidence which can be used regarding the likely motivations of individuals in ideological groups. The second argument, that there are ‘suﬃcient explanations on other grounds’, is refuted in the case of welfare policies since many were signiﬁcantly inﬂuenced by large welfare advocacy groups. Regarding the third reason, the fact that ‘most’ organized groups are working for themselves hardly justiﬁes ignoring those potentially powerful groups which are working for the gains of others. In sum, we must agree with Olson when he says ‘the theory is not at all suﬃcient where philanthropic lobbies, that is, lobbies that voice concern about some group other than the group that supports the lobby, or religious lobbies, are concerned’ (Ibid., p. 160). I would like to present an alternative hypothesis to Olson’s. Groups such as welfare rights advocates and civil rights activists are inherently diﬀerent from economically deﬁned interest groups, because the reason for their formation is ideological or altruistic. They are not playing the economic maximizing game. They are playing the ideology game and striving to achieve ends other than economic rewards. Ideology or altruism, which seeks the advancement of others – whether increased welfare or aﬃrmative action – replaces the selective incentive or coercion of Olson’s groups. Hence they can rationally participate in large interest group activities and still be acting in their own self-interest and in a rational manner. While the objective of action may be the economic advantage of others (and may lead to ancillary beneﬁt for those involved – such as welfare jobs, or civil rights jobs) the activity itself may be an end for those with ideological motivations. It is precisely because the self-interest of ideological groups does not depend on economic factors that they are successful even though they do not ﬁt into Olson’s general theory. In this sense, they are a special case of large interest groups that Olson has failed to consider. The importance of including ideological or philanthropic groups in any analysis of the legislative process is clearly revealed by examining the activities and eﬀectiveness of groups such as the Settlement House workers (see chapter six). Giving up well-to-do lifestyles, the workers in the Settlement House movement lived in the slums, published books and magazine articles, held meetings and made speeches and fought in legislative halls. Their

76

The political economy of welfare reform in the United States

activities contributed to the founding of the ﬁrst Juvenile Court, and the passage of state child labor laws. They founded, in collaboration with other groups, the Charities Publication Committee, which undertook research and led to the publication of the massive Pittsburgh Survey of social conditions, which in turn inﬂuenced the Progressive Party’s 1912 presidential platform. Today, many ideological groups, including a large number of those aﬃliated with diﬀerent religious denominations, engage in research and disseminating literature on social issues of concern to their members, and have full-time paid lobbyists in Washington, whose primary function is to meet with congressional staﬀers to inﬂuence the legislative process. Clearly a model which ignored these groups could not be expected to accurately reﬂect the legislative process. It is also important to examine the role of other groups, such as federal, state or local oﬃcials, which are not generally considered as interest groups. Several organizations representing these groups have been extremely inﬂuential in aﬀecting policies, and in creating and appropriating public goods. The National Governors Association (NGA), for example, was instrumental in passage of the new welfare law. When President Clinton vetoed the conference bill for being tough on children and weak on work, the NGA responded by proposing language in the bill that would give the states more funds for both children and work programs. These proposals addressed the concerns of the President, and assured the governors even more federal funds than they had previously agreed to! (It is also interesting to remark that the American Federation of State, County and Municipal Employees was the second largest Political Action Committees (PAC) contributor to federal candidates in 1993–1994.1) Although government oﬃcials may not at ﬁrst appear to ﬁt into the interest groups model, it is clear that they do comprise a group with a common interest, such as higher funding for their state in the case of governors. Moreover, the transfer of funds from taxpayers, in general, to state administrators, in particular, constitutes the creation and appropriation of a public good, just as much as a tax loophole for a business entity would constitute such a good. Under the Olson thesis, a group such as the governors can be particularly eﬀective, since they both constitute a relatively small group, and thus have readily apparent gains, and they also have selective incentives, insofar as a governor’s re-election prospects, where re-election is possible, are, presumably, positively correlated with the ﬁnancial situation of his state. We turn now from the actors to the interactions which aﬀect the legislative process. Becker (1983), as mentioned above, does not explain the formation of interest groups, but simply assumes that they will form as a result of market forces and eﬀectively exert pressure to inﬂuence policy decisions.

The public choice perspective

77

All the groups described above ﬁt into an expanded Becker model which would incorporate competition among groups. The relevancy of his model, which focuses on the competition between taxpayer groups and subsidy recipients, is suggested by the increasing numbers of groups which have been formed in protest at what is considered high levels of taxation – groups such as Americans for Tax Reform, Citizens Against Government Waste, Citizens for a Sound Economy and a myriad other groups seeking to protect the rights of taxpayers. Similarly, consumer groups such as Common Cause and Public Citizen have developed to counteract the inﬂuence of business groups. However, Becker does not address a crucial issue in the interaction of interest groups, namely that of competition between diﬀerent subsidy-seeking interest groups. One of the earliest scholars to examine competition among interest groups was Earl Latham, who analyzed the role of groups in governmental policy-making in his book, The Group Basis of Politics (1952). Latham concluded that the ‘legislative vote on any issue tends to represent the composition of strength, i.e., the balance of power, among the contending groups at the moment of voting’ (Ibid., p. 16). This has been a feature of the legislative process for decades and even centuries, but one to which little attention is given. Among the welfare issues which have given rise to high levels of competition were restrictions on child labor, proposed by the Children’s Bureau and fought by business interests, and the Sheppard–Towner Act which provided for federal grants to the states for medical facilities, and which was ﬁercely attacked and eventually defeated by the AMA. More recently, competition between interest groups was evident during and after passage of the new welfare law, where diﬀerent groups lobbied heavily for changes that would be more favorable to them. Among the groups competing for changes were diﬀerent immigrant groups who had been cut oﬀ from beneﬁts, agricultural interests, including grocers who stood to lose from cutting the food stamp program, and representatives of states who sought to gain or lose block grant bonuses depending on the way in which block grant formulae were written. The tendency to ‘up the ante’ as each group tries to outdo its competitor leads not only to high levels of rent-seeking activity, but also gives a competitive advantage to those groups which are supported by highly capitalized interests. For while grass roots lobbying can to some extent rely on a high degree of volunteer labor, the impact of media campaigns and other technological resources used by lobbyists generally require a high level of capital inputs. This last point can be illustrated by cases such as Proposition 9 in California, the Clean Air Initiative. A month before the election, polls showed that 64 percent of the electorate supported the initiative. The poorly funded and relatively disorganized proponents, who spent $250,000

78

The political economy of welfare reform in the United States

were outspent ten to one by the opposition interest group, a coalition of oil, chemical and utility companies. After a massive media campaign in the ﬁnal month before voting, the initiative was defeated two to one. Such instances are clearly extremely common, and serve to highlight the degree to which legislation can be purchased by a well-funded and well-organized interest group. A major eﬀect of competition between interest groups is that the bidding war increases, and, paradoxically, competition leads to higher prices and greater rent seeking, since there can be no increase in the supply of the good, which is generally a unique piece of legislation. The increase in the value of legislation as well as the increased competition from other groups has led to spectacular growth in both membership and expenditures of interest groups in recent years. The AARP, for example, grew from 1 million members in 1967 to 33 million in 1995 (Hrebenar, 1997). The increases in interest group expenditures are particularly evident among business groups. The oil industry, for example, spent millions of dollars on a campaign to defeat a bill to break up the 18 largest oil companies. Senator Birch Bayh (D-Ind.) called this eﬀort ‘the most sophisticated, expensive and elaborate lobby eﬀort I’ve ever seen’ (Ibid., p. 125). Given the expected returns to these investments, the levels of expenditures are hardly surprising. Hrebenar cites the case of an antismoking proposition on the ballot in California in 1978, in which opponents spent $5.5 million to successfully defeat the measure. Ninety-nine percent of the funding came from tobacco companies who had at stake over $1.7 billion in sales annually, and they estimated that if the proposition passed tobacco sales might decline 15 percent. While welfare oriented groups are generally not in the same league as multi-million dollar companies, the increasing privatization of welfare services may well lead to similar trends in the future.

NOTE 1. Federal Elections Commission, PAC Activity in 1994 elections, cited in Trattner (1994), p. 199.

5.

Government and its bureaucracy

In the previous chapter, the primary focus of analysis was abstract voting models and the role of non-governmental agencies and individuals, speciﬁcally voters and special interest groups. This chapter moves to the center of the legislative process, presents a public choice model of this process and analyzes the roles of the institutions of government, including the legislature, the executive branch and the government bureaucracy.

MODELING THE LEGISLATIVE PROCESS: THE PUBLIC CHOICE APPROACH A major contribution of the public choice approach has been the application of the tools used for economic analysis to the political environment, and speciﬁcally the legislative process. This analogy, however, has certain limitations. For example, compared to traditional theory, the public choice model leads to the logical conclusion that not only are the suppliers and demanders of legislative output not clearly deﬁned, but the product of government is itself shown to be misidentiﬁed. The legislation produced by government is often of little worth, in itself. Its value lies in the regulations and programs it authorizes and funds, or more precisely, in the beneﬁts that these confer on aﬀected parties. Thus, for example, food stamp legislation (a ﬁrst stage intermediate product) has no value per se. But the food stamp program (a second stage intermediate product) provides increased proﬁts (ﬁnal product) to grocery and other agricultural interests, as well as increased income to food stamp recipients and increased funds to government agencies implementing the program. The relationships between the producers and consumers of government output are also shown to be far more complicated and ambiguous than traditional theory suggests. It is clear that special interests that beneﬁt from government regulations and tax loopholes as well as government programs, are consumers. If, however, the speciﬁc legislation that beneﬁts these groups is funded by a tax on the general population, these taxpayers in general are clearly not consumers. To the extent that they are unable or unwilling to organize to ﬁght the legislation, due, for example, to rational ignorance or to transactions costs, they may be considered as coerced suppliers who gain 79

80

The political economy of welfare reform in the United States

nothing but are obliged to pay taxes. Thus, the lines between supplier and demander are blurred and voters will simultaneously be suppliers in some legislative markets and demanders in others. Similarly, special interests play a role in both the supply and demand side of the legislative market place, although not simultaneously in any one market. Not only do these groups make their demands known through political contributions, they also engage in the process of producing legislation by, for example, appearing as expert witnesses at hearings and conferring with those staﬀ members responsible for writing the legislation. Thus, they provide inputs to the provision of that legislation which they are demanding, giving rise to a major conﬂict of interest at the heart of the legislative process. This conﬂict of interest is particularly acute in the case of those interest groups that themselves form part of the federal or state bureaucracy. For example, in the case of the New Welfare Law of 1996, representatives of state governments played a very active role in developing the formulas that would be used to calculate federal funds allocated to the states for certain programs contained in the law. Clearly, in this case the states are themselves interest groups, each trying to gain the best deal for its citizens at the expense of the citizens of other states. Similarly, an agency such as the Department of Health and Human Services, charged with implementing the New Welfare Law, is also charged with writing the regulations regarding implementation of the law, a situation which also holds dangers of conﬂict of interest. Finally, we need to consider the role of the legislature itself, including members and their personal and committee staﬀs. The public choice approach sees this institution as primarily a broker of legislative beneﬁts. To the extent that government programs are redistributive in nature, transferring general taxes to speciﬁc interest group beneﬁciaries, either directly or indirectly, the gains received by these groups are rents. In this view, the sale of legislation to the highest political bidder may be an intermediate step, but the ﬁnal result is a brokering of rents to those groups oﬀering the highest political rewards from those groups least able to resist. The view that government is primarily a broker of rents, does not, however, rule out the possibility that in certain instances legislation may be produced which is primarily ideological or altruistic in origin and purpose. Thus, for example, a Member of Congress who holds strong ideological1 beliefs, who is strongly identiﬁed with these positions and who has strong electoral support, may well be able to introduce and vote on legislation that is ideological in nature. The degree to which this is possible will depend, inter alia, on the safety of the member’s seat, the degree to which he is beholden to special interests who may oppose the legislation for campaign funding and support, and the degree to which the ideological position itself

Government and its bureaucracy

81

guarantees voter support. When viewed through the economist’s lens of supply and demand, the legislator in this case is himself a consumer of the ideological legislation.

LEGISLATIVE STABILITY AND CONGRESSIONAL DOMINANCE The application of rational choice theory to the legislative branch of government has produced a body of literature which incorporates analyses of the internal processes of decision-making, the stability of legislative outcomes, and the external relationships of the legislature with other branches of government such as the bureaucracy. Gordon Tullock examined the implications of the median voter theorem on the legislative process in a paper entitled ‘Why So Much Stability?’ (1982). Tullock noted that with complex issues to be decided by a process of majority voting, there is a strong likelihood of cycling over issues and stable outcomes would be highly improbable. However, he pointed out that: ‘If we look at the real world, not only is there no endless cycling, but acts are passed with reasonable dispatch and then remain unchanged for very long periods of time’ (Ibid., p. 189). A major reason for such stability, according to Tullock, is the existence of logrolling and coalition formation. Noting that most government actions have the characteristic of giving a rather intense beneﬁt to a small group, at a small cost to each member of a large group, Tullock claims that under simple majority voting such bills would not be passed. If a coalition of several small groups was to get together and logroll such bills could succeed, he claims, but there are signiﬁcant costs. The process of coalition formation is inhibited by information and transactions costs, in particular the bargaining among members and the use of credible threats of abandoning the coalition in order to gain privileges. Because of these diﬃculties, ‘the end product is apt to be the individually bargained logrolling model, with each congressman making a series of individual bargains with others because bargaining problems of attempting to maintain coalitions for a long period of time are too great’ (Ibid., p. 198). This explicit logrolling process may be replaced by a procedure in which a committee canvasses members to discover which bills would be passed and bundles the selected measures in one bill as a timesaving device. Tullock concludes that whether logrolling is explicit or implicit, it can lead to stable, though generally ineﬃcient, legislative outcomes. The view that logrolling accounts for the stability of legislative outcomes is strongly contested by, among others, Kenneth Shepsle and Barry Weingast

82

The political economy of welfare reform in the United States

who maintain that real-world legislative practices constrain the degree of legislative instability by ‘restricting the domain and the content of legislative exchange’ (1981, p. 504). These authors, in their paper entitled ‘Structure-Induced Equilibrium and Legislative Choice’ (1981), note that logrolling and vote-trading can also lead to endless cycles and policy reversals due, inter alia, to enforcement problems, and they see these practices as part of the problem, not part of the solution. Shepsle and Weingast present several examples of institutional arrangements that, by modifying majority rule, can contribute to legislative stability. One such rule, ﬁrst proposed by Fiorina (1980), would limit amendments to a given set of projects to those which: 1. 2. 3.

strike a project; add a project; substitute one project for another.

Under these rules it can be shown that a stable policy choice exists (Fiorina, 1980). A further example, suggested by McKelvey (1979), is the stability that results when an individual or set of perfectly conspiring individuals have suﬃcient agenda control to ensure that the ﬁnal outcome is at his (or their) ideal point (Shepsle and Weingast, 1981). Other restrictions include Tullock’s (1967) proposal that small changes not be allowed, thus restricting the set of proposals, Shepsle and Weingast’s notion that legislators be restricted to proposing only one dimension of the status quo at a time, the germaneness rule which restricts amendments, and rules requiring that the status quo be voted on last (Shepsle and Weingast, 1981). All these rules contribute to stability by restricting the behavior of legislators and the potential for legislative exchange to upset an equilibrium. The rules governing legislative processes not only contribute to stability of outcomes, but are arguably a source of committee power. This has been demonstrated by Shepsle and Weingast who note committees have enormous power over legislation in their area of responsibility. Alternative explanations of this power have focused on the role of the committees as gatekeeper of new legislation, their specialized knowledge and expertise in their area of jurisdiction and their control over the legislative agenda. According to Shepsle and Weingast (1987) these characteristics of committees serve more to describe committee power than to explain from whence it originates. Moreover, none of them is convincing in and of itself, and each is subject to certain checks and balances. For example, the advantages of a committee’s access to specialized knowledge and information is mitigated by the existence of an extensive congressional staﬀ system, supported by research institutions such as the Congressional Research Service

Government and its bureaucracy

83

and the Congressional Budget Oﬃce, as well as by the experts and lobbyists from the various entities aﬀected by the legislation. Shepsle and Weingast (1987) propose an alternative explanation of committee power, claiming that ‘it resides in the rules governing the sequence of proposing, amending, and especially of vetoing in the legislative process’ (quoted in Mueller, 2003, p. 424). They emphasize particularly the last stage of the legislative process, the conference procedure, in which bicameral diﬀerences are resolved and the committee has ex post veto power over the legislation. This power, the authors maintain, ensures that changes in the status quo adverse to the interests of a decisive committee majority can be denied ﬁnal passage. It is the credible threat of the use of this power, rather than informal reciprocity agreements that makes gate keeping and proposal power eﬀective, according to these authors. Weingast and Marshall (1988) take the institutional analysis of the legislature a step further in a later paper which develops a theory of legislative institutions more directly comparable to the theory of the ﬁrm. In this paper, the authors note that, ‘Like market institutions, legislative institutions reﬂect two key components: the goals and preferences of individuals, here legislators seeking reelection from their constituents, and the transactions costs that are induced by imperfect information, opportunism and other agency problems’ (Ibid., in Mueller, 2003, p. 445). While both kinds of institutions are subject to agency problems, the speciﬁc types of problems and the appropriate solutions will diﬀer. In the market, mechanisms such as vertical integration evolve as solutions to agency problems such as moral hazard and adverse selection, while the legislature requires mechanisms to deal with problems such as enforcement of agreements on vote trading or logrolling. Weingast and Marshall (1988) trace the root of the agency problem to the fact that ‘a huge variety of interests are represented in the legislature, and almost none is represented by a majority’ (in Mueller, 2003, p. 444). Thus, ‘for most interests to gain policy beneﬁts, representatives with diﬀerent constituents must agree to exchange support’ (in Mueller, 2003, p. 444). This diversity of interest creates gains from exchange within the legislature, and such exchange is facilitated through the legislative committee system which ‘provides substantial protection against opportunistic behavior, thereby providing durability to policy bargains’ (in Mueller, 2003, p. 455). To illustrate their point, Weingast and Marshall use the example of one group of legislators who seek dams and bridges for their constituents while another group of legislators seek a regulatory agency for their voters. Where either voting or implementation of the legislation is not simultaneous, the agreement is subject to reneging. However, under the committee system, members who gain most from regulatory policy will tend to sit on

84

The political economy of welfare reform in the United States

the relevant regulatory committee and those who seek dams and bridges will seek positions on committees with jurisdiction in this area. Since each committee has control over the legislative agenda in their area of jurisdiction, they can block any eﬀort to renege and ensure that the agreed-upon policy is implemented. All of the papers discussed above analyze the legislature as a self-contained entity, in isolation from other branches of government. The intricate links between the legislature and the bureaucracy is illuminated in a paper by Terry M. Moe, entitled, ‘An Assessment of the Positive Theory of Congressional Dominance’ (1987). The theory of congressional dominance is part of a strand of rational choice theory that moved away from abstract voting models to an emphasis on the role of institutions in political outcomes. The basic argument of the theory is that Congress controls the bureaucracy. Moe argues that the theory focuses almost exclusively on the Congress, and speciﬁcally certain committees, and ignores the institutional form and dynamics of the bureaucracy. He illustrates in great detail how changes in the leadership of one bureaucracy, the Federal Trade Commission, led to policy changes within that agency, which were neither initiated, nor generally approved of, by the overseeing committee. The means by which Congress could control agencies – budgets, appointments and legislation – are shown to be theoretically feasible, but in practice diﬃcult to implement.

GOVERNMENT BUREAUCRACY In the public choice literature, the role of the government bureaucrat is often considered analogous to the role of the manager of a ﬁrm, and the bureaucrat’s relationship to elected oﬃcials is comparable to a manager’s relationship to stockholders. Both are subject to principle-agent problems, and in both cases performance may not be fully reﬂected in remuneration (Williamson, 1964; Mueller, 1990; Baumol, 1959). One of the earliest and most inﬂuential studies of bureaucracy in a public choice framework was that of William Niskanen (1971). In marked contrast to the traditional view of theorists such as Max Weber (1947), Niskanen did not hold the view that the principle role played by bureaucracies was to serve the public interest by implementing the policies of a benevolent government. Rather, Niskanen applied rational choice analysis, maintaining that bureaucrats pursue their own rational self-interest. Noting that government bureaucrats generally have a monopoly on the provision of a good or service, thus making monitoring diﬃcult, and that salary structures do not contain eﬃciency incentives, Niskanen proposed that bureaucrats have goals other than economic eﬃciency or proﬁt

85

Government and its bureaucracy

maximization. He cited, inter alia, perquisites of oﬃce, power, prestige, public reputation, patronage, output of the bureau, ease of making changes and ease of managing the bureau. All but the last two, he claimed, were positively related to the size of the bureau’s budget, leading to an expansion of output and maximization of the budget, subject to costs being covered. The bureau’s budget is obtained by a process of bargaining with an appropriations committee, which is assumed to be a monopsonist. An important feature of the model is that the non-market nature of the bureau output, and the monopoly structure of the bureau leads to asymmetry of information regarding costs. The funder cannot accurately assess costs, since there is no market for the output. Niskanen suggests that under these conditions the bureau is the dominant bargainer, and the result of its maximizing strategy will be that the budget will be expanded beyond the level where marginal cost equals marginal public beneﬁt. In Figure 5.1, the optimum output would be at Q*, where the marginal cost function, C, intersects the marginal beneﬁt function, B. However, the bureau, according to the Niskanen model, can use its monopoly power to appropriate monopoly rents (see below for a detailed discussion of rent seeking), represented by triangle B, equivalent to the consumer surplus at equilibrium, represented by triangle, A. To the extent that the bureau is successful, the output will approach Q’. The goals proposed by Niskanen have been questioned by other authors, including Weingast and Moran (1983), Williamson (1964), and Gist and Hill (1981). Like Niskanen, Weingast and Moran employ a principle-agent model in their analysis of bureaucracies. However, they maintain that committee members possess suﬃcient rewards and sanctions to create an incentive system for agencies and that, agency mandate nothwithstanding, rewards go to those agencies that pursue policies of interest to the current B,C

C

A

B

B Q* Figure 5.1

Q´

The oversupply of a bureau’s output

Q

86

The political economy of welfare reform in the United States

committee members; those agencies that fail to do so are confronted with sanctions. It is noteworthy that the authors speciﬁcally refer to the policies which the bureaus produce: they do not refer to the economic eﬃciency with which the bureau operates. As Weingast and Moran note, both their approach and that of Niskanen may lead to observationally equivalent behavior. The committee’s apparent lack of oversight, represented by such things as few oversight hearings, perfunctory attention to the ongoing operations of the bureau, and infrequent congressional investigations, could equally well indicate that an eﬀective incentive system obviated the need for such activities. Weingast and Moran claim that this is in fact the case. They note that the threat of ex post sanctions creates ex ante incentives for the bureau to serve a congressional clientele. Thus, the more eﬀective the incentive system, the less often sanctions would be observed. The authors conclude that direct and continuous monitoring of inputs rather than of results is an ineﬃcient mechanism by which a principle constrains the action of his agent (Weingast and Moran, 1983). Thus, for Weingast and Moran, the balance of power lays with the committee, which constrains the activities, though not necessarily the overall budget, of the bureau, while Niskanen holds that bureaus have suﬃcient power and independence to increase both output and total costs beyond that which would be socially optimal. In order to shed more light on the balance of power between bureaucrat and politician, it is important to deﬁne the objectives of each. While committee members have incentives to deliver certain policies or programs to voters and/or interest groups, it is not evident that they have any incentive to monitor the eﬃciency with which such policies and programs are delivered. For such an incentive to exist, the rewards for discovering and rectifying ineﬃciencies would need to exceed the not insigniﬁcant monitoring costs. Such rewards would presumably be in the form of votes or support from interest groups. There is little evidence, however, to suggest that interest groups or voters are suﬃciently interested in the eﬃciency of individual government bureaus (although they may be quite dissatisﬁed with the overall level of eﬃciency in government) and, in fact, ﬁscal illusion and rational ignorance theories would suggest the contrary. Furthermore, in some cases there would seem to be strong reasons for interest groups to actively discourage monitoring by committees. A producer of armaments, for example, lobbying for increased expenditures on his product, is hardly likely to encourage a committee to put pressure on the Defense Department to minimize its costs! In sum, the committee may have policy control over the overall policy agenda of the bureau, as Weingast and Moran suggest, while the bureau has monopoly control over the production processes, as Niskanen has asserted.

Government and its bureaucracy

87

Empirical tests provide conﬂicting evidence on whether the output of bureaus exceeds the optimal output. Romer and Rosenthal (1978, 1982) ﬁnd corroboration that budgets are higher than that which a median voter model would predict in their analysis of Oregon school budgets, and Weingast and Moran (1983) provide evidence that bureaucratic actions are constrained by the congressional oversight committees. In the ﬁnal analysis, it is clear that in order for bureaucrats or other interest groups to gain from transfer programs the funds must be appropriated and allocated within the federal budget. This wealth transfer between political constituents in the political market is analogous to the wealth transfer between consumers and producers described by Peltzman (1976), Stigler (1971) and others in the rent-seeking literature.

THE IRON TRIANGLE The relationship between the relevant Congressional oversight committees, the bureaucracy and special interests has been described by some as an ‘Iron Triangle’. This term conveys not only the strong links between these three components of the political process, but also the mutually supportive and reciprocal nature of the relationship. In the popular press, the term Washington establishment is often used to convey the same idea. One of the earliest writers to explore the nature of the Washington establishment was Morris Fiorina (1977a). In Congress: Keystone of the Washington Establishment, Fiorina summarizes his conclusions thus: There is a Washington establishment. In fact, it is a hydra with each head only marginally concerned with the other’s existence. These establishments are not malevolent, centrally directed conspiracies against the American people. Rather, they are unconsciously evolved and evolving networks of congressmen, bureaucrats, and organized subgroups of the citizenry all seeking to achieve their own goals. Contrary to what is popularly believed, the bureaucrats are not the problem. Congressmen are. The Congress is the key to the Washington establishment. The Congress created the establishment, sustains it, and most likely will continue to sustain and even expand it. But . . . the disturbing aspects of the Washington establishment follow from the uncoordinated operations of the overall system, not from any sinister motivation of those who compose it . . . many of those in the heart of the establishment are genuinely unaware that they are members in good standing (1977a, p. 3).

The continuing operation of this type of establishment, according to Fiorina, ‘has potentially disturbing implications for the future welfare of this country’ (Ibid., p. 2). Fiorina builds his case for the existence of a Washington establishment

88

The political economy of welfare reform in the United States

on the basis of research carried out by David Mayhew and published in an article entitled ‘The Case of the Vanishing Marginals’ (1977b). This article documents the decline of marginal voting districts and the accompanying increase in the number of ‘safe’ political districts represented by career politicians. Mayhew characterizes the modern Congress as ‘an assembly of professional politicians spinning out political careers’ (Fiorina, 1977b, p. 15). He notes that the jobs oﬀer good pay and high prestige and there is no shortage of applicants. Successful pursuit of such a career requires continual reelection. And in fact this is hypothesized to be the major goal of incumbents (which is not to deny that some among them desire reelection for reasons of ideology or altruism as opposed to personal power or wealth). According to Fiorina the marginal Congressman is necessary for the health of the political system since changes in popular sentiments are reﬂected through these members rather than members who have safe seats and are likely to remain in them for the long term, regardless of marginal shifts in voter opinion. Marginal districts build responsiveness into the electoral system, and as such districts disappear, according to Fiorina, ‘we face the possibility of a Congress composed of professional oﬃceholders oblivious to the changing political sentiments of the country’ (Fiorina, 1977b, p. 14). The decline in the number of marginal districts has been attributed to a number of factors, including changes in the socioeconomic homogeneity of the district (either endogenously or through redistricting), changes in the eﬀectiveness of the congressional incumbent, and changes in voter behavior. Fiorina rejects the ﬁrst of these on the basis, inter alia, that socioeconomic change tends to be gradual and takes decades to show up, while the decline of the marginals between the 1950s and 1970s (the period under study) was quite rapid. Similarly, he rejects the argument that redistricting led to the declining marginals, noting that the decline in marginal states among those that were not subject to redistricting was just as great as in those states which were redistricted. This argument is clearly weakened by the fact that there would be no need to redistrict in those states in which there was evidence that other factors were leading to declining marginalism. A combination of increasing homogeneity in some districts and redistricting in others could have signiﬁcantly contributed to the increase in so-called safe districts. This, however, is rejected by Fiorina. Fiorina also dismisses the argument that increased eﬀectiveness of the incumbent politician led to the increase in safe seats. Higher visibility due to being an oﬃce holder, in combination with the increased resources available to oﬃce holders, gives the incumbent a signiﬁcant advantage vis-à-vis any challenger. The basis of Fiorina’s rejection is that if it were correct one

Government and its bureaucracy

89

would ‘expect to see the name recognition of congressional incumbents increasing over the period during which the marginals were vanishing’, which was not the case (1977b, p. 22). Fiorina proposes an alternative explanation of the vanishing marginals, suggesting that voters are the culprits. By changing the principles on which their votes depend they have aﬀected the levels of congressional turnover. This behavioral change argument holds that voters became disillusioned with political parties and stopped using party aﬃliation as a determinant of voting, instead using incumbency, and assuming that if an incumbent made no serious mistakes, they were satisfactory and should therefore be reelected. The implausibility of this argument is noted by Fiorina who points out that it implies voters who are increasingly cynical and alienated by politicians are simultaneously increasingly supportive of their incumbent representatives (Ibid., p. 27). In order to shed more light on the problem, Fiorina undertakes ﬁeld work in two districts which have similar socioeconomic proﬁles, but one of which is a vanishing marginal while the other is still a marginal district. He concludes that: The changing nature of congressional elections in these two districts stems directly from the changing behavior of the congressmen who represented them. Both districts are heterogeneous in a socioeconomic sense and consequently in their basic political allegiances (e.g., as illustrated by registration). So long as (they) are represented by congressmen who function principally as national policymakers . . . reasonably close congressional elections will naturally result. For every voter a congressman pleases by a policy stand he will displease someone else. The consequence is a marginal district. But if we have incumbents who deemphasize controversial policy positions and instead place heavy emphasis on nonpartisan, nonprogrammatic constituency services (for which demand grows as government expands) the resulting blurring of political friends and enemies is suﬃcient to shift the district out of the marginal camp (Ibid., p. 37).

The change in behavior of congressmen is attributable, at least in part, to the ‘growth of an activist federal government’ which has ‘stimulated a change in the mix of congressional activities’ (Ibid., p. 46). The growth of government bureaucracy has provided opportunities for congressmen to serve their constituents by acting as powerful intermediaries. Congressmen have acceded readily to the increased demand for casework services that accompanied the growth of bureaucracy. For while traditional legislative activities generally please some constituents and displease others, few enemies are made by casework activities. Moreover, as a legislator, a congressman is only one of several hundred, and responsibility and credit are less clear than when a direct service is provided. The reward for this service is reelection. And as seniority increases, so goes the argument, so does

90

The political economy of welfare reform in the United States

casework eﬀectiveness and thus the probability of continued reelection. The increasing role of pork barrel activities can be explained along similar lines. As congressmen divert their resources from lawmaking (which can divide a district) to casework and bringing home the pork (which uniﬁes support), the marginal districts were rendered safe. The beneﬁts of the current system, for both congressmen and constituents that receive intermediary services, are clear. The bureaucracy also gains from this arrangement, however. Bureaucrats who are interested in expanding their agency services for ideological reasons, as well as those primarily interested in greater personal power and prestige, beneﬁt from larger bureaucracies. Moreover, they can earn the support of congress by their responsiveness to congressional requests, thus consolidating support for continued funding. The end result of this system, according to Fiorina, is that: more and more bureaucrats promulgate more and more regulations and dispense more and more money. Fewer and fewer congressmen suﬀer electoral defeat. Elements of the electorate beneﬁt from government programs, and all of the electorate is eligible for ombudsman services. But the general long term welfare of the United States is no more than an incidental by-product of the system (1977b, p. 49).

The increase in casework and pork provision by congressmen has been facilitated by signiﬁcant increases in personal and committee staﬀ, and by an increase in the number of subcommittees which give members more control of speciﬁc programs which are important to their constituents. The subcommittees form the cornerstones of a series of subgovernments each composed of the standing subcommittee, the agencies under its jurisdiction, and the special interests served by the legislation – the iron triangles. Logrolling between the subcommittees facilitates the passage of legislation which favors organized minorities who then repay their congressmen with reelection. The power of the iron triangles may be diminishing, however. A dramatic increase in the number of interest groups, some of which have competing ends, and speciﬁcally a rise in the number of public interest groups, increases the potential costs to congressmen of satisfying the needs of any speciﬁc group (Rowley et al., 1995). Moreover, committee power has been eroded as proliferating committees have jurisdictional conﬂicts, and as ﬂoor amendments are increasingly being used to circumvent the committee system. Finally, the increasing importance of contributions from PAC, which do not necessarily have strong links with any geographic district, tempers the gains to be made from exclusive focus on district casework and pork provision services.

Government and its bureaucracy

91

THE ROLE OF THE PRESIDENT In analyzing the role of the President, the public choice approach once again incorporates an analysis of the motivations and institutional constraints that inﬂuence the behavior of the oﬃce holder. The impact of the President on the legislative process is of consequence not only because the oﬃce is a powerful one in terms of aﬀecting the passage of legislation and the structure and activities of the bureaucracy, but also because the objectives and constraints of the oﬃce may diverge sharply from those of the legislature. Legislators are beholden to the constituents in their states or districts, while the President is beholden to a more heterogeneous national electorate. Moreover, the President need only be concerned with reelection during the ﬁrst term of oﬃce. Thereafter, eﬀective governance and a historical legacy may be of greater concern (Rowley et al., 1995). The lower priority given to reelection activities, as well as the national scope of the constituency, gives the President autonomy to pursue his own vision for the nation, a vision to be implemented by a centrally directed bureaucracy, responsive to his agenda. The primary role of the executive in the legislative process is derived from the power of the veto. A study of the presidency from a rational choice perspective by Crain and Tollison argues that the veto power raises the cost of reneging on legislative contracts (1976). These authors suggest that increased attempts to renege on previous legislation, or substantively alter it, should lead to more vetoes as the President tries to protect legislation. This argument is based on the idea that larger majorities in the congress will lower the cost of changing existing laws, thus increasing attempts to do so which, in turn, will lead to an increased number of vetoes. The veto power thus increases the durability and thus the value of legislation to its beneﬁciaries, which may include special interests, legislators and members of the bureaucracy. The argument is clearly analogous to the theory of the independent judiciary derived by Landes and Posner (1975), which claims that the judiciary enhances the durability and value of existing laws by rarely nullifying them and generally interpreting them in terms of the intent of enacting legislators. In both cases, the clear implication is that rather than a separation of powers there is a collusion of interests. Crain and Tollison used data on gubernatorial vetoes across state government in the United States to test their hypothesis and concluded that ‘the overall regression explain(ed) over half of the variation in gubernatorial vetoes and (was) highly signiﬁcant at the 1 percent level’ (Crain and Tollison, 1976, p. 565). Clearly this still leaves a great deal to be explained. Moreover, the use of gubernatorial vetoes as a proxy for the presidential veto is highly questionable given the unique nature of the presidency in

92

The political economy of welfare reform in the United States

terms of objectives, constraints and institutional context. Perhaps the most troubling issue, however, is the failure to show what the President has to gain from enhancing the durability of legislation which, according to the iron triangle hypothesis, is developed as a result of bargaining between legislators and special interests. This is particularly problematic when one notes that the proposed new legislation is presumably the result of changes in public opinion of the balance of power of interest groups (see Weingast, 1981). It seems reasonable to assume that the President would generally have a greater interest in supporting the newly powerful legislators and interest groups, than in supporting those now out of power. They are no longer in a position to pay for enhanced durability, and if such durability had been part of an earlier agreement, the incentive to renege is clear. An alternative view of the presidential role suggests that given the limited inﬂuence over policy that the oﬃce entails, the President reserves this power for his highest priority areas such as new initiatives (Weingast, 1981). If the new policy is combined with a change in the composition of active interest groups or with a change in public opinion that supports the President, his action will increase costs for supporters of the status quo, both legislators and special interests. Weingast provides the example of airline deregulation to support his claim. He shows that legislative changes were brought about by, inter alia, increasing public support combined with the interventions of Presidents Ford and Carter. Clearly this argument renders questionable the collusion of powers suggested by Crain and Tollison.

NOTE 1. Ideological is here used in the sense that a Member of Congress is willing to press for legislation even at the cost of constituent interests (see Kau and Rubin, 1979 and Kalt and Zupan, 1984).

6. The origins of the new welfare law – a historical overview This chapter brieﬂy traces the historical development of welfare, from individual charity motivated primarily by moral values and administered within the community or parish, to large-scale governmental charity, administered at both national and local levels, and for which the motivations are far more complicated than simple benevolence. The interests and values of diﬀerent elements within the society and the eﬀectiveness of the interest groups that they comprise are highlighted in the chapter.

HISTORICAL DEVELOPMENT OF WELFARE PROGRAMS IN GREAT BRITAIN Concern for the welfare of the poor has a long history, not only in Western nations but also in the East. William Trattner (1994), who has undertaken a detailed study of the origins of policies aﬀecting the poor, notes that as early as 2000 years before Christ, Hammurabi, the ruler of Babylon, made the protection of orphans and widows, and the weak against the strong, part of his code of ethics. The ancient Greeks and Romans, as well as Buddhists and the authors of the Old Testament, extolled the virtues of giving to the poor, often suggesting that it was a moral duty. Thus, what is often thought to be a Christian tradition has, in fact, a much longer genealogy, and one that spans many cultures. While it is useful to bear this in mind when examining subsequent developments, this chapter will focus only on the growth of welfare policies and institutions in the United States and their precursors in Great Britain. In the early years of Christianity the needy were cared for by family and neighbors in the local community and charitable institutions were unknown. Over time, however, a number of religious and secular institutions emerged to address the needs of growing numbers of poor. The development of monastic orders around the sixth century marks the beginning of this trend, with some orders being formed primarily to aid the needy, both giving to those who came to their doors and taking provisions to the poor in local communities. The monasteries were also among the ﬁrst to 93

94

The political economy of welfare reform in the United States

establish hospitals in medieval times to care not only for the sick, but also to house and care for travelers, orphans, the aged and the destitute. Outside of the monasteries, the ecclesiastical authorities administered medieval poor relief at the parish or diocese level. The bishops and parish priests distributed a percentage of the tithes (which were compulsory payments) to those in need within the parish. Thus, although the redistributive welfare state is often considered a relatively modern invention, its origins can be traced back well over a thousand years! In addition to ecclesiastical institutions, a number of secular agencies emerged to provide for the needy. With the advent of feudalism the lord of the manor took responsibility for the serfs residing on his lands, providing insurance against sickness, old age and unemployment. Those who lived in the city were often aided by guilds, which ‘distributed corn and barley yearly, fed the needy on feast days, provided free lodgings for destitute travelers, and engaged in other kinds of intermittent and incidental help’ (Trattner, 1994, p. 5). Thus, a network of charitable agencies, both urban and rural, secular and religious, coexisted with individual charity even in medieval times. These institutions underwent major upheavals with the demise of feudalism. In the rural areas, the agricultural revolution led to displacement of tenant farmers and agricultural laborers as landlords enclosed their lands in order to raise more proﬁtable sheep. Meanwhile in urban areas the Industrial Revolution and the growth of the factory system led to the decline of craft guilds and the growth of a large pool of low-skilled urban workers, including those who came from the rural areas in search of employment. High levels of unemployment were accompanied by dismal working conditions and long hours of labor for adults and children alike. The seeds of major changes in dealing with the problem of poverty can be traced to the mid-fourteenth century, with the advent of a number of natural calamities, including crop failures, famines, and the Black Death of 1348–1949, which killed almost a third of England’s population. Widespread poverty, migration from rural areas into towns and cities, along with increased crime and begging led to state intervention and the introduction of a series of restrictive statutory measures. One of the most notable was the Statute of Labourers, proclaimed in 1349, which set maximum wages, restricted travel for indigent and unemployed persons and forbade the giving of charity to able-bodied alms seekers. The social and economic upheavals continued, however, and further measures were enacted in the sixteenth century. In 1531 a statute was passed which provided for the public whipping of able-bodied vagrants, while assigning aged or disabled persons areas in which they were allowed to beg. This philosophy of sharply distinguishing between the ‘worthy’ and the

The origins of the new welfare law

95

‘non-worthy’ poor was further developed under the Henrician Poor Law, the Act for the Punishment of Sturdy Vagabonds and Beggars.1 This legislation provided for branding, enslavement and execution of able-bodied beggars, but also ordered local public oﬃcials to gather resources from voluntary contributions to the churches to provide alms for the poor, aged and disabled, and to provide work for the able-bodied. Thus the recognition that some of the poor were unable to ﬁnd work, and the consequent introduction of the ﬁrst ‘workfare’ program, can be traced back over 450 years! Children who were found begging could be placed in apprenticeships if they were between 5 and 14 years old. This act thus provided for the ablebodied and disabled, for the young and old. The administration of the Henrician Poor Law was the responsibility of both civic and ecclesiastical local authorities, such as mayors and churchwardens. Funds were initially raised through contributions to churches. By the late sixteenth century, however, this source of funds was no longer considered suﬃcient, and a law was introduced directing local oﬃcials to assess and tax all the inhabitants of their jurisdiction in order to provide for the needs of the poor. A new public oﬃce was created, the overseer of the poor, who was charged with providing work relief for the needy. In spite of extensive legislation and the increasing allocation of funds to the needy, serious problems remained. The welfare system was put to the test in the 1590s when widespread famine, food scarcity and inﬂation led to rioting and social disorder. Motivated by fear as much as ideology, lawmakers responded to the threat by passing the Poor Law of 1601, a law that remained the basis of England’s welfare system until the 1830s (Butler and Kondratis, 1987).

THE POOR LAW OF 1601 This legislation continued the approach of previous measures such as the Henrician Poor Law, but with more extensive and detailed provisions. It contained harsh measures for vagrants refusing to work – including whipping, branding, being stoned or put to death – and had no provisions for appeal from those who felt aggrieved. It stipulated that parents, if they had the means, were responsible for their children, and children for their needy parents and grandparents. It followed previous legislation in dividing the needy into three major categories and prescribing separate treatments for each. Children were to be placed in apprenticeships, the able-bodied were to be put to work, and the disabled were to be provided either institutional (‘indoor’) relief (such as for the mentally ill), or home (‘outdoor’) relief (Himmelfarb, 1984).

96

The political economy of welfare reform in the United States

While many of the provisions of the Poor Law had clear antecedents in earlier legislation, this law was the ﬁrst comprehensive welfare legislation, making coherent the sometimes inconsistent measures which were in existence, establishing the responsibility of civil authorities, and establishing obligatory ﬁnancing outside of the church. Most remarkably, perhaps, the law clearly reﬂected the view that the needy not only deserved assistance, but had a legal right to it.2 Many of these features subsequently provided the basis for poor laws in the American colonies.

THE DEVELOPMENT OF WELFARE POLICIES IN THE NEW WORLD The historical development of policies and programs to deal with poverty in the American colonies bears remarkable similarity to the development of such policies in England. In the early days of settlement, land was plentiful and the indigent were generally cared for by family, friends or neighbors. Over time, however, the English practice of shipping convicts, vagrants and other undesirables to the colonies, in addition to the hardship of the journey, which resulted in death for some and illness for many, contributed to a growing number of indigents in the New World. Unlike the Old World, in which ecclesiastical institutions such as monasteries provided aid to the poor, the colonies in general did not develop such establishments. (The Dutch Reformed Church in New Amsterdam, which established a system of ecclesiastical poor relief, was a notable exception.) By the mid-seventeenth century, the colonies began adopting legislation based on the English Poor Law. Poor relief was administered at the local level, towns or parishes, and funds were raised through a compulsory tax. Children were placed in apprenticeships. The mentally ill were cared for by the community, and later in institutions. And, as in England, severe penalties were introduced for the able-bodied unemployed that could be jailed, whipped, or bound out as indentured servants. While many of the policies reﬂected the English precedents, however, several new approaches were adopted in the colonies, including the practice of placing the poor in private homes at public expense. In some cases the poor were auctioned oﬀ to the lowest bidder, with the town paying for their upkeep. Free medical attention was provided, and exemption from taxes. The generosity of townsfolk to the poor among them did not extend to strangers, and some towns required newcomers to have a town resident provide for security and guarantee that they would not become a public charge. Since each town was obliged by law to provide for needy inhabitants, the deﬁnition of ‘inhabitant’ gained great signiﬁcance, and generally

The origins of the new welfare law

97

referred to a person who had been resident for three months without being asked to leave. The problem was particularly acute in the large port towns, where many impoverished immigrants were brought. Many of these towns employed oﬃcials to ward oﬀ strangers and required ship captains to post a bond for each person they brought to the colonies. The refugee problem was also exacerbated periodically by those ﬂeeing troubles such as Indian wars in the frontier settlements. The disproportionate burdens falling on towns such as Boston led the authorities of this town to request and receive funds from the colonial treasury, marking the beginning of state funding for welfare. By the early eighteenth century a state funded, locally administered welfare system, not dissimilar to that in England, was in place in the New World. In both the Old and New Worlds the growth in legislation, taxes, and expenditures was accompanied by growth in the numbers of needy. Between 1700 and 1753 the residents of Boston increased the amount spent on poor relief from 500 pounds to 10,000 pounds per year, and it increased further through the 1770s, even when the population had stabilized (Trattner, 1994). Likewise in New York in 1772 one out of four free men were classiﬁed as poor.3 Poor relief is estimated to have cost an average of 10 to 35 percent of municipal funds. In addition to the problems cited above, a rise in illegitimacy contributed to high dependency rates and by 1750 half the poor relief of some of the colony parishes went to families caring for illegitimate children. The growth in public expenditures on welfare was accompanied in the eighteenth century by an increase in private charity from institutions and individuals that had accumulated wealth. In addition to religious organizations, a number of institutions based on ethnicity, such as the Boston Irish Society and the German Society of New York, emerged to provide relief primarily, but generally not exclusively, for persons of the same ethnic origin. From a public choice perspective it is interesting to examine some of the motivations behind the formation of these groups. One of the earliest charitable societies was the Scots Charitable Society, founded in Boston in 1657. According to its charter this organization was founded for the ‘relief of ourselves and any other for which we may see cause’ (Trattner, 1994, p. 37). Clearly, a combination of self interest – providing mutual insurance against calamity – as well as humanitarian principles were motivating factors behind the establishment of this society. William Trattner suggested that the Quakers, who provided a great deal of relief, were motivated not only by humanitarian and religious values, but also by the desire to establish ‘a place in American Society’, particularly since their refusal to bear arms set them apart from their fellow countrymen (Ibid.). He also claims that the

98

The political economy of welfare reform in the United States

philanthropic activities of some of these groups was based on social objectives such as a desire to ‘prove to themselves and to others that they were worthy of respect and admiration, to rub shoulders with distinguished citizens, or to control the poor’ (Ibid., p. 9). While self-interest undoubtedly accounted at least partially for the individual philanthropy of many, humanitarianism as an end in itself was promoted by several movements of the eighteenth century. The Great Awakening, a series of religious revivals which took place from the late 1720s to the 1740s, was led by George Whiteﬁeld, who preached about the misery of the poor, took up collections on their behalf, and distributed aid to individuals as well as orphanages and schools. One noted social welfare historian claims that this movement popularized philanthropy and transformed ‘do-goodism from a predominantly upper- and middle-class activity – half responsibility, half recreation – into a broadly shared, genuinely popular, avocation’ (Bremner, 1960, p. 42). The Enlightenment was also an inﬂuential movement, and John Locke’s treatises on psychology, which held that human beings were created as blank slates, contradicted the determinism of Calvinism, and led to the implication that the elimination of poverty was not only possible but desirable. Locke, himself, devised a plan for pauper schools in which children would be put to work as well as educated, and in which their mothers also would be employed, but this was never adopted (Himmelfarb, 1984, p. 25). Finally, the American Revolution, and particularly the Declaration of Independence, which held that all men were created equal, tended to call attention to the existing inequalities, and may have further spurred the growth of philanthropy. By the end of the eighteenth century, then, philanthropy was ﬂourishing, both from private and individual sources. By this time too public assistance had spread to the Western Territories and the states that were carved from them. In the South, the parish system of poor relief was replaced by civil institutions, following the separation of church and state and dissolution of the parishes. In spite of all these eﬀorts, however, the incidence of poverty continued to increase, and along with it demands on public assistance. Increasing poverty was a feature of both English and American society throughout the nineteenth century, though for diﬀerent reasons. In America, immigration was often blamed, as well as rapid industrialization. While the former seems plausible as new settlers took time to become established develop skills and ﬁnd employment, it seems that industrialization should have led to increased employment as production and commerce increased, more than oﬀsetting the loss of jobs in handicrafts. In England,

The origins of the new welfare law

99

the Speenhamland system was thought to have led to high expenditures on poor relief. This system was developed in the district of Speenhamland in 1795 to address the problem of low wages and high prices (Himmelfarb, 1984). Laborers paid less than a certain amount, which varied according to the price of wheat, were granted supplemental relief. The legislation provided that ‘Every poor and industrious man’ whose earnings fell below a given standard, determined by the price of bread and the size of his family, would receive a subsidy from the parish to bring his income up to a minimum subsistence level. As Himmelfarb notes, this system spread to other districts, particularly in the depressed rural south, with the result that a considerable number of agricultural laborers became dependent, entirely or in part, on the parish. Moreover, Trattner notes that ‘the system seemed to bring with it higher prices, higher taxes, and the belief that, since the poor were guaranteed a minimum income, they not only were becoming lazy, but were encouraged to multiply’, (Trattner, 1994, p. 45) – a refrain strikingly similar to what one hears some 200 years later (Himmelfarb, 1984, pp. 64–5). Growing concern about the increasing costs of poor relief was accompanied by concern about its eﬀects on the motivations of recipients, as well as dissatisfaction with the administration of relief. In England the system of poor relief depended largely on overseers of the poor who were said to be ‘unpaid, untrained and often incompetent’ (Trattner, 1994, p. 49). And, since aid was locally administered, there was great inequality, with those districts having a high proportion of needy being subject to the greatest demands, while also having the most meager resources to address those demands. The practical concerns regarding the eﬃciency and equity of the system were accompanied in the early nineteenth century by the evolution of a philosophy of individualism and laissez-faire. The goal of obtaining the greatest happiness for the greatest number required, according to this philosophy, that each individual be allowed to pursue their own self interest, without interference from the state. Adam Smith, however, a major proponent of laissez-faire, did not call for the abolition of the Poor Law. As Himmelfarb has noted, the self-interest propounded in Smith’s The Wealth of Nations seems to be incongruent with the sentiments in his other major work, The Theory of Moral Sentiments. Regarding the latter, Himmelfarb comments that: ‘the operative word in that book was sympathy. Sympathy was presumed to be as much a principal of human nature as self-interest; indeed it informed self-interest since it was one of the pleasures experienced by the individual when he contemplated or contributed to the good of another’ (Himmelfarb, 1984, p. 47). In fact, Smith insisted that there were many occasions when the interests of the individual had to make way for

100

The political economy of welfare reform in the United States

the interests of others, or the general interest. Moreover, he was, at times, highly critical of the ‘selﬁsh and duplicitous’ mercantile interests: Our merchants and master-manufacturers complain much of the bad eﬀects of high wages in raising the price, and thereby lessening the sale of their goods both at home and abroad. They say nothing concerning the bad eﬀects of high proﬁts. They are silent with regard to the pernicious eﬀects of their own gains. They complain only of those of other people (Smith, 1759, p. 98, quoted in Himmelfarb, 1984, p. 49).

Another highly inﬂuential ﬁgure of the time, Thomas Malthus, was unequivocal about the evils of the Poor Law and was one of the major proponents of abolishing it. He maintained that it enabled families to have more children, thus leading to lower wages and living standards, and ultimately to greater misery for the poor and the society at large. The principle of population which underlay Malthus’ argument was based on what he predicated were two laws of nature: food is necessary to the existence of man and the passion between the sexes is necessary and would remain nearly in its present state. He maintained that population, when unchecked, increases in a geometric ratio while subsistence food production increases in an arithmetic ratio, leading to an enormous discrepancy. In the absence of preventive checks on population growth, such as delayed marriage, this discrepancy would be corrected by positive checks such as starvation, sickness, war, infanticide and other such factors which would reduce the size of the population to the level of the food supply. The power of Malthus arguments has been noted by Walter Bagehot who remarked that Malthus: ‘advertized his notions and ﬁxed them among the men who understood a simple and striking exaggeration far more easily than a full and accurate truth. He created an entirely new feeling on his subject’ (Bagehot, 2001, p. 331). Similarly, Himmelfarb has noted that: ‘Whatever diﬃculties there were in his theory, however faulty the logic or evidence, it gripped the imagination of contemporaries, of all ranks, classes, callings, and persuasions as few other books had ever done’ (1984, p. 127). The concerns about public relief aroused by Adam Smith and Thomas Malthus were reinforced by a number of other prominent ﬁgures, including Alexis de Toqueville, who, writing in the 1830s expressed his conviction that: . . . any permanent, regular administrative system whose aim will be to provide for the needs of the poor, will breed more miseries than it can cure, will deprave the population that it wants to help and comfort, will in time reduce the rich to being no more than the tenant farmers of the poor, will dry up the sources of savings, will stop the accumulation of capital, will retard the development of trade, will benumb human industry and activity, and will culminate by bringing

The origins of the new welfare law

101

about a violent revolution in the state, when the number of those who receive alms will have become as large as those who give it, and the indigent, no longer being able to take from the impoverished rich the means of providing for his needs, will ﬁnd it easier to plunder them of all their property at one stroke than to ask for their help (Drescher, 1968, pp. 1–2).

Clearly, the writings of both Malthus and de Tocqueville not only reﬂected a belief that the system of public relief had negative economic consequences for the poor and for society at large, but it also incited fears regarding the consequences for personal and public safety. The degree to which this was rhetorical rather than based in the genuine beliefs of the authors is open to debate, but undoubtedly the publications of such prominent authors aﬀected the ideology of the time. In addition to concerns about the unintended consequences of the relief system, there were increasing concerns about the costs and eﬃciency of poor relief both in England and in America, where some six million immigrants, mainly German and Irish, landed between 1800 and 1860. In England, a major change was brought about in the system after the electoral reforms of 1832 led to a transference of power from the landed, mercantile classes, who had generally supported the system, to the manufacturing interests, who were highly critical of it. Parliament created a Poor Law Commission which concluded that the prevailing poor law system, and particularly wage supplementation, was responsible for undermining the character and energy of the English laboring classes. This commission recommended the appointment of a national supervisory body to oversee administration of poor relief and the amalgamation of parishes into a smaller number of units in order to improve coordination and eﬃciency. It also recommended an end to assistance for able-bodied persons, except in public institutions. In America, too, there was a growing belief that outdoor relief was ‘more harmful than the pain it (was) intended to relieve’, and there was increasing resentment and perhaps fear, of large numbers of immigrants spending much time in idleness, and, in the case of the Irish and Germans, in the neighborhood bars. In America, as in England, however, the system was already too entrenched to be easily dismantled. Critics had to be content with replacing outdoor relief by indoor institutional relief. This move was supported by the Yates Report, the ﬁrst comprehensive survey of poor relief in the United States, which cited four main methods of public assistance – home relief, the contract system, the auction system, and institutional relief. It reported that where the poor were farmed out, by contract or by auction, they were often abused, and the children neglected, subject to disease and uneducated. Home relief, however, led to idleness, and the able-bodied receiving such relief were often unemployed. The solution,

102

The political economy of welfare reform in the United States

according to the Yates Report, was institutionalization. The nineteenth century, then, saw a major expansion in institutions for the poor. Almshouses were constructed for the disabled, elderly or other ‘worthy’ poor, and workhouses for the able-bodied poor. Along with an increase in local institutions, increases in non-resident populations led to the building of state institutions, and, in 1854, Congress passed a bill that would have provided federal aid for the mentally ill. This was vetoed, however, by President Franklin Pierce, who noted that: ‘If Congress has the power to make provision for the indigent insane . . . it has the same power for the indigent who are not insane . . .’ adding that he could not ﬁnd any authority in the constitution for ‘making the Federal Government the grand almoner of public charity throughout the United States’ (Trattner, 1994, p. 68). Pierce felt that to do so would be contrary to the letter and spirit of the Constitution, and subversive of the theory on which the United States was founded. This strong stance deferred the time when federal aid would gain widespread acceptance as a means of dealing with the problems of poverty. While nineteenth century America was remarkable for the growth in public institutions, these did not supplant private charity, which continued to grow, and was particularly evident in the period of depression following the ﬁnancial panic of 1837, and during and after the Civil War. Private charitable institutions proliferated, especially towards the end of the nineteenth century, when public relief fell into disfavor. While the causes of this are subject to debate, contributing factors included reports of fraud and corruption; a perception that it was ineﬀective; the inﬂuence of Smith, Malthus, and de Tocqueville; and the publications of scholars such as Herbert Spencer, who argued that the government should conﬁne itself to ensuring liberty of the individual by protecting citizens from attack upon their persons or property, and had no role in such matters as health, education or business regulation, let alone public relief. Among the private organizations formed during this period, the US Sanitary Commission was quite remarkable. Organized in 1861, the main objective of this commission was to unite the various local voluntary relief societies into a national organization that would supplement government agencies in meeting the spiritual and physical needs of the men in uniform. Working among the troops in army camps and ﬁeld hospitals, providing education, housing, food and clothing, this organization was so successful that J.S. Mill said of it: ‘History aﬀorded no other example of so great a work of usefulness extemporized by the spontaneous self-devotion and organizing genius of a people altogether independent of the government’ (Trattner, 1994, p. 81).

The origins of the new welfare law

103

Another feature of this period was the growth of the Settlement House movement. Residents of the settlement houses saw themselves as social reformers, who sought to understand the poor by living among them, and to eliminate poverty by understanding its causes. The ﬁrst settlement house, opened in 1886, was based on the idea that groups of residents on a street or a number of streets should: ‘be organized into a set of clubs which are, by themselves, or in alliance with those of other neighborhoods, to carry out, or induce others to carry out, the reforms – domestic, industrial, educational, provident or recreative – which the social ideal demands’ (Trattner, 1994, p. 170). The settlement house workers, generally upperclass and well-educated, published books and magazine articles, held meetings and made speeches, and fought in legislative halls as well as urban slums. They were instrumental in the founding of the ﬁrst Juvenile Court and in passage of state child labor laws. One of the major outcomes of the settlement movement was the large amount of research, and speciﬁcally the founding, in collaboration with other groups of the Charities Publication Committee. This committee undertook studies of social conditions, which inspired a further major investigation – the Pittsburgh Survey. Published in six volumes between 1909 and 1914, and containing page after page of statistics, this report sought to estimate the costs of preventable diseases, industrial accidents, low wages, and other factors associated with the living conditions of the poor in Pittsburgh. These publications led to a reexamination of the causes of poverty and the degree to which they resulted from personal inadequacies or socio-economic factors. The degree to which the latter were gaining credence with voters is evidenced by the adoption in the Progressive Party’s 1912 presidential platform of an eight-hour work day (replacing 12 hours) a six day work week and the prohibition of child labor. Writing about the reforms proposed at this time, Trattner notes that they nearly all involved to some extent ‘limitations on private property rights and the extension of public authority into areas previously regarded as the exclusive reserve of the individual’ (1994, p. 184). The early twentieth century marked a major milestone in the development of the welfare system from a voluntary system of individual private charity to one which was administered and legislated by national and state governments, and ﬁnanced by taxation. While private charity coexisted with public aid, and was particularly forthcoming in times of need such as wars and depressions, the growth in this form of aid was far surpassed by the increase in federal aid in the twentieth century. While the nineteenth century had seen an increasing role for states in taking on previously local responsibilities, there had not been a major federal role, except for wartime activities such as the establishment of the Bureau of

104

The political economy of welfare reform in the United States

Refugees, Freedmen, and Abandoned Lands in the War Department. This entity, which became known as the Freedman’s Bureau, was initially authorized to provide relief during the Civil War and for one year afterwards, but Congress subsequently extended its life for an additional six years. After this period, the federal government withdrew from the social welfare scene and the states took on a greater role both by increasing the numbers of institutions, and by establishing State Boards to coordinate state institutions and improve their eﬃciency. It was not until the twentieth century that the federal government would return to be a major player in the welfare system. Among the factors which led to an increased role for the federal government was the increasingly sophisticated lobbying from the organized charity groups, as well as reform groups such as the Settlement House movement mentioned earlier. A number of studies published by these groups, including one by the New York Charitable Organizations Society in cooperation with Columbia University indicated that, during the period under study, from 1890–1897 (a period of depression), lack of employment was the primary cause of poverty, while sickness and accident were second, and ‘shiftlessness and intemperance’ third, accounting for 10 percent of the cases. It is not clear, however, how a distinction was made between those who had no job due to ‘shiftlessness’ and those who could not ﬁnd any employment. While diﬃculties persisted in determining the degree to which the ablebodied unemployed were responsible for their own condition, and thus support for them was limited, the plight of children generally caused concern. As Trattner has noted, ‘almost every . . . agency or individual working for social betterment saw in children the possibility for constructive altruism. As a result, a broad child welfare movement swept through America from the mid-nineteenth century through the early twentieth, one unlike anything before it or after it’ (1994, p. 110). The increasingly well-organized welfare groups, including the Charitable Organizations Society and the Settlement House movement, espoused the cause of children, who throughout the nineteenth century, had at various times been placed in almshouses (with convicts, the mentally ill and others considered in need of institutionalization), been shipped out to the West to be reared on farms, and been contracted out to homes, where at times they were subject to neglect and abuse. In addition to the publications of welfare groups and academics, writers such as Charles Dickens in England brought the plight of children to the attention of the general public. A major outcome of these eﬀorts, was the ﬁrst White House Conference on Dependent Children in 1909, which gave welfare issues a prominent place in national life, and which resulted, three years

The origins of the new welfare law

105

later, in the creation of the US Children’s Bureau, clearly marking a milestone in the involvement of the federal government, and a complete reversal of the philosophy espoused half a century earlier by President Franklin Pierce. The eﬀorts that led to the creation of the Children’s Bureau illustrate the power of interest groups and the high level of resources they invested in swaying government policies. The proponents of child welfare had long fought for increased legislation, more funding and more government administration and oversight. This group included social workers, civic organizations, public health groups, labor unions and various individuals, all of whom sought to gain, materially or otherwise from increased resources for child welfare. They undertook and published research, held meetings and lobbied extensively. The losers from the measure were business interests, who feared that the Bureau would bring an end to child labor, which they had a vested interest in preserving. The extent of their lobbying is reﬂected in the contentious nature of the ﬁve day debate on the Senate ﬂoor, prior to passage, in which opponents of the bill charged that supporters were working under orders from communists and socialists who wished to control the nation’s youth. The political sophistication of the early welfare administrators is well illustrated by the ﬁrst head of the Children’s Bureau, Julia Lathrop. Trattner describes her thus: (she) believed fervently in the importance of public welfare and in the need to rejuvenate services set up by the taxpayers to help those in need. She also knew how to use the potent constituency network she had at her disposal and, at the same time, recognized the need to avoid controversy and to divorce politics from social welfare in order to gain and retain congressional support for the new bureau. At the outset she thus subordinated studies of child labor in favor of less controversial ones, such as the problem of infant mortality. She therefore won conﬁdence in the new agency and additional resources for it; by 1915, for example, its funds had achieved a sixfold increase, and its staﬀ had multiplied by a factor of ﬁve (1994, p. 219).

By 1918, however, Julia Lathrop was beginning to take on controversial causes, including federal grants to state for public health facilities, hospitals and other institutions falling under the aegis of the medical profession. Lathrop’s proposals were contained in a bill which became known as the Infancy and Maternity, Sheppard–Towner Bill, introduced in 1918. This bill was ﬁercely contested by members of the medical profession, who ran a highly-organized and well-ﬁnanced campaign, accused the bills proponents of socialism and communism, and published pamphlets with titles such as ‘Shall the Children of America become the Property of the State?’, circulated by the Legislative Committee of the Illinois State Medical

106

The political economy of welfare reform in the United States

Society. The bill was passed three years after its introduction, but was rescinded eight years later, again after a forceful campaign by the medical profession. Control of female operated infant and maternal health centers was removed and returned to the hands of private male physicians, and many state facilities lost support. One of the major long-term eﬀects of the bill was in setting the precedent of grants-in-aid to states for welfare programs other than education. Precedents were also set in this period for government intervention in a broader range of activities, particularly pensions and insurance. In April 1911, Missouri enacted the ﬁrst widow’s pension law, which provided aid to full-time mothers with children. Other states rapidly followed suit and by the time the Social Security Act was passed in 1935 all but two states, South Carolina and Georgia, had implemented such programs. These statutes were the forerunners of Title IV of the Social Security Act, which established the federal program of Aid to Dependent Children, the major program of cash assistance for needy families with children. Revised and retitled in 1996, this program is now known as Temporary Assistance for Needy Families, and is discussed in detail in the following chapter. Other measures passed in the ﬁrst two decades of the twentieth century included enactment of worker’s compensation in 43 states between 1909 and 1920 (measures which appealed to employers, who thus avoided the large payments occasionally awarded by courts, and who thus supported the proposals) and old age pension laws, which were generally weak and often inoperative. Eﬀorts to implement health insurance were strenuously and successfully campaigned against by an alliance that included the medical profession, insurance companies, and employers groups. The decade of relative prosperity which followed World War I saw few changes in the nascent welfare system. This was changed abruptly, however, by the stock market crash of 1929, and the Great Depression, although it was not until 1931 that any major action was taken. Between 1929 and 1932, about one-third of the nation’s private welfare agencies disappeared for lack of funds, while millions of workers were unemployed, farmers lost their lands and businesses closed. President Herbert Hoover opposed federal aid, believing it would delay the natural forces at work to restore the economy, impair the solvency of the government, establish more politicized bureaucracies, and undermine free enterprise. ‘You cannot extend the mastery of government over the daily lives of the people without at the same time making it the master of their souls and thoughts’, he declared; and, in response to a measure proposed by House Speaker John Garner and Senator Wagner of New York calling for a public works program to stimulate the economy and create jobs, Hoover remarked that, ‘never before in the nation’s entire history has anyone made so dangerous a suggestion’

The origins of the new welfare law

107

(Trattner, 1994, p. 279). This strong stance did not withstand three winters of depression, in which not only private and local resources, but also state funds were depleted, and in summer 1932, in the face of mounting public pressure, Hoover signed a relief bill providing federal loans to states for general and work relief. The decisive rejection of Hoover in favor of President Franklin D. Roosevelt marked the beginning of a new era. While Governor of New York, Roosevelt had introduced the ﬁrst State Unemployment Relief Act, the Wicks Act. Working on the belief that unemployment, and consequent misery, was generally caused by an uncontrolled economic system rather than by the unemployed themselves, Roosevelt plunged the federal government into the welfare business. Public works programs such as the Public Works Administration and the Civilian Conservation Corps provided employment for millions of citizens, up to a third of all the unemployed. With lower pay and fewer hours than private sector jobs, the programs were constructed so as not to be a threat to private sector interests. The Federal Emergency Relief Act of 1933, which made available $500 million for grants to the states, signiﬁed the beginning of an increasingly large ﬂow of federal dollars to the states for welfare programs. By far the most important and far-reaching piece of legislation to come out of the New Deal, however, was the Social Security Act of 1935. The motivations behind this legislation were multiple. Aside from humanitarianism, the role of fear is apparent in Roosevelt’s statement that: Democracy has disappeared in several other great nations because the people of those nations had grown tired of unemployment and insecurity . . . in desperation they chose to sacriﬁce liberty in hope of getting something to eat. We, in America, know that our democratic institutions can be . . . made to work

providing that the ‘government is equal to the task of protecting the security of the people’ (Trattner, 1994, p. 288). This security was to be provided by the Social Security Act, which was an omnibus measure incorporating both public assistance and social insurance. It provided old-age insurance and public assistance for the aged, unemployment insurance, aid to dependent children in single parent families, to crippled children and to the blind, and federal funds for state and local public health work. The strong and increasing inﬂuence of older citizens was reﬂected in the remarks of the executive director of the committee which developed the program, Edwin Witte, who expressed the belief that without pressure from the elderly he doubted ‘whether anything would have gone through at all’ (Trattner, 1994, p. 290). In addition to serving the needs of the growing numbers of elderly who were increasingly becoming organized as an inﬂuential pressure group, some have argued that employers had a vested interest in encouraging older

108

The political economy of welfare reform in the United States

workers to leave the workforce in order to hire younger, more productive employees. Thus a combination of humanitarianism, fear, and political expediency on the part of the federal government and self-interest on the part of state governments, along with the self-interest of the elderly and employers, and the mix of ideological and expansionary motivations of welfare workers, interacted to give rise to one of the most monumental pieces of welfare legislation ever passed. Shortly after the passage of the New Deal legislation, the outbreak of World War II served to divert the attention of the nation away from domestic problems. Moreover the war years and the decade following were periods of relatively full employment and increasing incomes. It was not until the early 1960s, and the arrival on the scene of President Kennedy, that welfare issues again became prominent in the national policy arena. Among the factors leading to renewed concern about poverty was the mass migration from rural areas to cities, as farming became modernized. Between 1950 and 1965 agricultural production increased 45 percent and farm employment fell by the same percentage, and it is estimated that over 20 million people were forced oﬀ the land, especially in the South. The migration of these unemployed masses came at a time when the need for unskilled labor was declining, leading to the development of even larger and more crowded ghettos and a doubling of the numbers of welfare recipients between 1960 and 1970. In addition to economic factors, social factors such as the increasing divorce rate may have played a role in the increase in welfare caseloads. Moreover, the increasingly well-organized and vocal civil rights movement brought vividly into focus the impoverishment of blacks and the racism with which they had to contend. These issues and the publication of the 1960 census ﬁgures led to a renewed debate on the problems of poverty, and an outpouring of articles in research journals and the popular press. One of the most widely read journalists and social activists was Michael Harrington, who argued that the system created and perpetuated the poverty it was intended to alleviate (Harrington, 1964). Similarly Richard Elman claimed that welfare programs collected the poor into ‘stagnant pools of dependent people who become increasingly separated from the mainstream of economic life’ (Elman (1966), quoted in Trattner, 1994, p. 318). These writers, however, did not call for an end to welfare but for more welfare with no strings attached. The Kennedy and Johnson administrations introduced an array of new welfare measures, known as the New Frontier and the Great Society respectively, in the early and mid 1960s. It is noteworthy that these programs were not primarily the result of popular demand for an expanded welfare system. Rather, as Senator Daniel Patrick Moynihan has pointed out:

The origins of the new welfare law

109

The plain fact, the large and indispensable fact, is that the attempt to address the issue of poverty in the whole of the United States came in the ﬁrst instance from an informal committee of a half dozen persons thinking up themes for President Kennedy’s 1964 reelection campaign. At one point it appeared that the likeliest choice would be the emerging, challenging problems of the suburbs. Poverty held on, however, . . . But the electorate never asked for it; the poor never asked for it (Moynihan, 1996, pp. 82–3).

While presidential politics was undoubtedly a key factor in both the New Frontier and the Great Society, debate continues on the relative impact of socio-economic conditions, the intellectual climate arising from the works of Harrington and others, and the inﬂuence of the civil rights movement and other social activist groups. Such factors may have laid the groundwork for persuasion of the electorate that the poverty issue was in fact an important area for government action. Whatever the relative causal factors behind the program proposals, there is little doubt that a key factor in the passage of much of the Great Society legislation was the power and political savvy of President Johnson, who wanted to bring to life his personal vision of the Great Society. The manner in which he achieved this was quite remarkable. As one student of the period has noted: Johnson’s tireless political crusade for the Great Society’s legislative agenda was a monument to his political skills, but it also sowed the seeds for intractable problems later. It was vintage Johnson. He wheeled and he dealed. Small favors that meant a lot back home were traded for key votes. Insuﬃciently enthusiastic lawmakers were brought down to the ranch for a dose of charm or pressure, whichever was more eﬀective . . . Little time was spent on thinking through the implications of ‘creative federalism’, the nebulous term Johnson used to describe far-reaching changes in the federal relationship, but much was devoted to striking bargains with governors and Congress (Butler and Kondratis, 1987, pp. 9–10).

A similar view of the legislative process and the role of the presidency at this time has been put forward by Lawrence Friedman, who claims that, with the exception of the civil rights legislation, the Great Society legislation was driven by presidential determination, rather than by a social movement pressuring congress, and thus it was riddled with concessions to powerful lobbies (Friedman, 1977). Even after the passage of legislation, Johnson would make concessions to the defeated lobbies to win their acquiescence. Thus, after the passage of Medicaid and Medicare, Johnson invited the AMA lobbyists to the White House and assured them that administration oﬃcials would listen closely to their advice in the development of program rules and regulations. The oﬃcials did so, making numerous concessions that reportedly contributed to vast increases in program costs (Butler and Kondratis, 1987).

110

The political economy of welfare reform in the United States

Support for the new programs was further shored up by incorporating non-proﬁt organizations, thus giving private social welfare advocates and professionals a vested interest in the implementation of the programs. According to a study made by the Urban Institute, the close ties between the federal government and the non-proﬁt sector renders it a virtual ‘third party government’, which, by 1980, received about 58 percent of revenues directly or indirectly from the federal government (Salaman and Abramson, 1982, p. 44). In addition to co-opting special interests in the private and nonproﬁt sectors, Johnson also needed the cooperation of the lower levels of government that would be responsible for implementing programs under the new legislation. Butler and Kondratis (1987) have remarked that the Great Society was an explicit attempt to use federal power to marshal national resources to address problems that some lower levels of government had failed to tackle. In some cases, such as civil rights, this was due to deliberate obstructionism on the parts of local interests, while in others it was due to lack of commitment or resources. The arrangements made between federal, state and local governments – Johnson’s ‘creative federalism’ – were based on political compromises and the promise of increased federal funds. What was not clear, a priori, was the degree to which federal oﬃcials would become involved in the details of local programs ranging from new towns and economic development to education and job training (Butler and Kondratis, 1987). Among the major legislative activities of the Kennedy–Johnson era were the expansion in 1961 of the Aid to Families With Dependent Children (AFDC) program to two parent families if the head of household was unemployed, and, in 1962 Amendments to the Social Security Act, which greatly increased federal funds to the states for welfare programs. Other programs included the Area Redevelopment Act of 1961 and the Economic Development Act of 1965 which were designed to encourage industries to relocate in depressed areas, the 1962 Manpower Development and Training Act, the Food Stamp Act of 1964, Medicare and Medicaid amendments to the Social Security Act in 1965, and the Economic Opportunity Act (EOA), of 1964. The EOA was intended to be part of Johnson’s ‘unconditional war on poverty . . . which threatens the strength of our nation and the welfare of our people’ (Trattner, 1994, p. 323). It established programs such as the Job Corps and Head Start for preschool children, and was generally focused on providing education and training, on the implicit assumption that this would lead to increased employment and reduced poverty. The increased involvement of federal oﬃcials in state and local government programs reﬂects in part a widespread view at the time that these entities were ineﬀective. The situation was described by one former state governor in 1967:

The origins of the new welfare law

111

The states are indecisive. The states are antiquated. The states are timid and ineﬀective. The states are not willing to face their problems. The states are not responsive. The states are not interested in the cities. These half dozen charges are true about all of the states some of the time and some of the states all of the time (Terry Sanford, Storm Over the States, quoted in Butler and Kondratis, 1987, p. 63).

In the South in particular, legislatures met brieﬂy and infrequently, voting rights were eﬀectively denied to blacks in many jurisdictions and the general economic plight of the region resulted in few if any programs directed to the poor and minorities. To force the hand of lower levels of government, federal grants were increasingly accompanied by detailed mandates and regulations. Both the funds and the programs grew steadily in succeeding decades. In 1965, about $10 billion was distributed to the states, a ﬁgure which grew to over $60 billion by 1980, funding almost 500 federal programs (Butler and Kondratis, 1987). These funds, however, did not necessarily end up in those areas of most dire economic need. The beneﬁciaries, including state and local government agents as well as private interests, found that ‘the political dynamics favored them over the proponents of a national interest’ (Butler and Kondratis, 1987, p. 67). There is considerable evidence that political clout and inﬂuence determine the distribution of federal grants, not poverty and distress. For example a study of federal grant disbursements by Randall Holcombe and Asgher Zardkoohi (1981), found no statistically signiﬁcant correlation between discretionary grants and standard determinants of poverty. On the other hand, ‘the data showed the grants to be allocated based upon political power. Per capita grants were higher in those states with more seniority in the Senate, with a larger percentage of majority party members in the House, and with members on the inﬂuential House and Senate committees’ (1981, p. 399). Similarly a 1977 Congressional Budget Oﬃce study of federal economic development programs found that counties in the top ﬁfth of per capita income received considerably more assistance per resident than those in the bottom ﬁfth (Congressional Budget Oﬃce, 1977). Increased federal funding for states during the 1970s and 1980s was accompanied by, though not necessarily causally related to, a transformation of state governments. Voting rights and civil rights legislation undoubtedly contributed to the increased responsiveness of state governments to the electorate. In addition, the federal requirements attached to the multitude of Great Society programs served to build up a professional state bureaucracy. Although disillusionment with the welfare system was widespread, the increased eﬃciency of state government was reﬂected in public opinion polls which showed that by the early 1980s a large majority

112

The political economy of welfare reform in the United States

of Americans believed that state legislatures were better at overseeing the day to day business of government, had more understanding of community needs and were more eﬃcient in managing social programs (Gallup poll, September 1981, Opinion Outlook, February 12, 1982, and Louis Harris Poll, June–July 1979, State Legislatures, November 1979, p. 23, all quoted in Butler and Kondratis, 1987, p. 83). The proliferation of welfare programs led to increased bureaucracy and power struggles over who was to control the programs, and thus reap the rewards. Among the winners were the mayors, who: ‘were able to create high-paying jobs for faithful political supporters and thus strengthen their hold on oﬃce, at the same time preventing meaningful political and social action by the needy, who might otherwise have jeopardized the status quo’ (Trattner, 1994, p. 324). While conservatives called the program a Madison Avenue Deal and the ‘Santa Claus of the free lunch’, radicals attacked it as a ‘huge political pork barrel’ and a ‘prize piece of political pornography’. Pressure from both sides was accentuated by general disillusionment with the program, noted by one of its designers, Senator Moynihan, whose book, Maximum Feasible Misunderstanding, published in 1969, concluded that the war on poverty accentuated doubts about the capacity of social science to plan, and government to deliver, ambitious programs for social betterment. Disillusionment with war on poverty programs, and the escalation of the Vietnam War did not mark the end of welfare expansionism, despite the expectations when Richard Nixon was elected in 1968. While the population in poverty had declined from 40 million, or 20 percent of the population in 1960, to 25 million or 12 percent of the population in 1969, the number of welfare recipients and total welfare expenditures continued to climb. Between 1963 and 1966, for example, federal welfare grants to the states doubled, and one million persons were added to the caseload. This was a period when expectations were rising, partly as a result of the political rhetoric regarding ending poverty and partly due to the rise of increasingly well-organized and vocal civil rights and other advocacy groups, which were already in the 1960s hiring full time paid lobbyists to work in Washington, and organizing mass demonstrations. Perhaps partially in response to these political events, President Nixon not only failed to reduce federal welfare programs, but also, particularly in 1972 when he was running for reelection, contributed to their expansion. Among the major legislation passed by the Nixon Administration was a 20 percent increase in Social Security beneﬁts in 1972, followed by a further 11 percent in 1974 and indexing to the cost of living, expansions of the Food Stamp Program and making it mandatory for all states, a transfer from shared state and federal funding to full federal funding of Old Age

The origins of the new welfare law

113

Assistance, Aid to the Blind and Aid to the Permanently Disabled in 1972, introduction of the Earned Income Tax Credit (EITC), and adoption of the Supplemental Security Program in 1974, which created the nation’s ﬁrst guaranteed national income program to the needy elderly, blind and disabled. These measures led one analyst to assert that ‘the greatest extensions of the modern welfare system were enacted under the conservative Presidency of Richard Nixon . . . dwarﬁng in size and scope the initiatives of (the Kennedy–Johnson era)’ (Trattner, 1994, p. 351). One major piece of legislation that Nixon was unsuccessful in passing was the Family Assistance Plan, a program to extend cash income for the poor who were not aged, blind or disabled. Shortly after his election Nixon selected as his welfare policy advisors: Richard Nathan, a research associate from the Brookings Institution (who was appointed assistant director of the Budget Bureau in the new administration), Marion Folsom, a secretary of Health Education and Welfare in the Eisenhower Administration, Wilbur Smith, the secretary of the Wisconsin Department of Health and Social Services, and Mitchell Ginsberg, chief of the New York Human Resources Agency which administered the nations largest welfare program. Among the proposals of the Nathan team was the introduction for the ﬁrst time of a national minimum AFDC cash payment that would be fully federally funded. While at ﬁrst glance it seems remarkable that such a proposal should arise in a Republican Administration, when one looks at the legislative process under which this legislation was formulated, and the institutional aﬃliations of the persons involved, it is clear that the outcome was relatively predictable. In addition to the Nathan team, the proposal was supported by Daniel Patrick Moynihan, the executive secretary of the President’s Council on Urban Aﬀairs (UAC), a cabinet level group that included Labor Secretary George Schultz and Agriculture Secretary Cliﬀord Hardin. The plan thus had inside support in the White House, Department of Health, Education and Welfare (HEW), and the Budget Bureau. State governors and editorial writers supported it on the outside. It also had enemies, however, most notably economist Arthur Burns, who was appointed chief domestic policy advisor, and other members of the UAC. In February 1969, President Nixon chose sides and remarked that he supported the establishment of national standards (Burke and Burke, 1974). When details of the plan were scrutinized, it became subject to increasing criticisms both from other factions within the administration, and from outside special interests. One of the earliest attacks was from an outgoing HEW oﬃcial, Worth Bateman, (a Johnson appointee) who noted, among other problems, that it provided ﬁnancial incentives to break up intact households, particularly those headed by a working male. Bateman had

114

The political economy of welfare reform in the United States

earlier proposed a negative income tax (which was also supported by Milton Friedman) as a way of dealing with the contradictory objectives of adequacy of payments versus work incentives. Bateman was subsequently appointed to an administration working group on welfare and drew up an alternative to the Nathan plan, most notable for its inclusion of a negative income tax and not conﬁned to families with children. The two welfare plans, with their diﬀerent groups of supporters and detractors, became the center of ongoing and divisive competition and intrigue within the administration, as vividly described by reporters Burke and Burke (1974). Moreover, Arthur Burns, who was vehemently opposed to any plan that would add to the welfare rolls and have signiﬁcant negative budgetary consequences, subsequently developed and presented his own plan to the President. Not being convinced by either plan, the President asked the advice of George Schultz, who devised a plan that contained some elements of both the Nathan and Burns plans. In making a ﬁnal decision regarding which, if any, plan to support, certain political tests had to be applied. These are tests to which any major presidential proposals are submitted before going to Congress. The ﬁrst regards the eﬀect of the proposal on special interests with a political claim on the President. The second test concerns the presidential reputation and legacy. In theory, a bold welfare reform could enhance Nixon’s reputation as a leader willing to confront a diﬃcult issue and put an end to the welfare mess. Moreover, if the Democrat-controlled congress failed to pass Nixon’s proposed legislation, they would be held responsible for the continuation of a discredited welfare system. Evidently, Nixon felt the legislation passed the tests. The proposal, which included a vastly expanded cash welfare program and an enlarged food stamp program, was sent to Congress in October 1969. Proponents of an adequate base payment, as well as proponents of generous work incentives were accommodated by higher guaranteed payments and income disregards. The Secretary of Agriculture and Senator George McGovern, Chair of the select Committee on Nutrition and Human Needs, were accommodated by an expanded food stamps program. Governor Rockefeller, who had complained that New York fared badly in comparison to the southern states, was accommodated by language in the bill that gave guarantees to the states. In spite of all the compromises made during the course of developing the welfare legislation, which came to be known as the Family Assistance Plan, its fate as it went to congress was highly uncertain. In December 1969, Representative Wilbur Mills, chairman of the House Ways and Means Committee which had been studying the bill since its submission in October, predicted that the bill would not pass since liberals thought it

The origins of the new welfare law

115

wasn’t enough and conservatives thought it a guaranteed income. Nevertheless, he managed to shepherd the bill through the House where it was passed by a vote of 243 to 155 on April 16, 1970 (Burke and Burke, 1974). In the interim Mills, with the agreement of the administration, had made further concessions to the nation’s governors, guaranteeing higher federal supplements to states for the new program. In the Senate, the bill ran into very rough weather. The Senate Finance Committee, which considered the bill, was dominated by conservatives who criticized the fact that the so-called reform would do nothing to address one of their major concerns, the anti-work incentives of the existing program. Despite adjustments made and presented by Elliot Richardson, the secretary of HEW, the Senate did not buy the plan. In the end it was mathematically impossible to simultaneously satisfy liberals by providing adequate coverage, satisfy conservatives with adequate work incentives, and do this within budgetary limits acceptable to ﬁscal watchdogs. The bill was buried in committee in late 1970, from where it was never resurrected. Thus after over two years of intensive work on developing this legislation, and numerous concessions made to shore up support, no major reform to AFDC was passed. A number of causal factors may be proposed: the inﬁghting between groups in the White House; the inherent impossibility of simultaneously satisfying liberals with generous beneﬁts and conservatives with strong work incentives and limited budgetary outlays; and the vociferous opposition of interest groups from both right and left. Moreover, it is notable that the idea was hatched by a group of administration oﬃcials aﬃliated with HEW whose goals were a combination of the political – passage of a Nixon welfare reform bill that would enhance the President’s image – and ideological – developing a more eﬀective welfare system. As Burke and Burke have noted: Even after the president proposed it, the working poor never pushed for wage supplements for themselves, perhaps because they were ignorant of the proposal, perhaps because they shunned the stigma of welfare. Among dozens of congressmen during the ﬁrst year of debate on the proposal none was found who received a single letter from a potential beneﬁciary (1974, p. 99).

Lack of grassroots support was compounded by the ﬁerce opposition of some powerful interest groups. The American Federation of Labor and Congress of Industrial Organizations (AFL-CIO) fought against the plan on the grounds that wage supplementation undermined the rationale for a higher minimum wage, and that the bill weakened their case for a government guarantee of jobs for all the unemployed. Some liberal groups wanted the federal government to take over welfare completely to improve treatment of the poor, particularly in the South, and when Nixon withdrew

116

The political economy of welfare reform in the United States

from this initial position, he lost the support of such interests. Other liberal groups, particularly unionized welfare workers on the payrolls of northern states and counties opposed the bill because at worst it might eliminate their jobs and at best it would transfer them to the federal payroll, eliminating beneﬁts such as free pensions and free health insurance, for which under the federal system they would be obliged to make partial payments (Burke and Burke, 1974, p. 145). The National Welfare Rights Organization opposed the bill on the grounds, inter alia that, while the poorest of the poor in the South would have increased beneﬁts, urban welfare mothers in the North would be obliged to work, and future beneﬁt increases might be impeded because of increased costs to their state treasury. Some white southern conservatives felt that by giving cash to six or seven million more recipients, many of them in poor working families, the supply of cheap labor would shrink and taxes would be increased. In the end, despite the best marketing eﬀorts of White House and HEW oﬃcials, they were no match for the highly organized and vocal opposition both within the congress and among the special interest groups. The Nixon presidency was ended by Watergate, and was followed by the generally ineﬀectual administrations ﬁrst of Gerald Ford, and then of Jimmy Carter who had claimed at the beginning of his administration that the welfare system should be scrapped and a totally new system implemented. The failure of Carter to maintain public conﬁdence in the face of stagﬂation in the economy and several foreign aﬀairs crises led to a landslide victory for Ronald Reagan in 1980. President Reagan espoused low taxes, small government, a strong military and a transfer of power from the federal bureaucracies to the states and localities. He was extremely eﬀective in achieving the ﬁrst three goals – increasing the defense budget, cutting taxes, terminating some welfare programs and slashing funding for others. None of the major means tested programs escaped cuts. His attempt to transfer power to the states, the ‘New Federalism’, failed to gain much support from either party but Reagan did succeed in introducing legislation which allowed states to replace some AFDC federal regulations by state designed provisions. (These state waivers were to be extensively used in subsequent years as states experimented with the AFDC program.) Reagan also met with vigorous opposition when he proposed cutting programs such as Social Security and Supplemental Security Income (SSI), the lobbying groups for which, as mentioned earlier, were highly ﬁnanced and well organized. A compromise measure was eventually signed in 1983, by which time the poverty rate was the highest in 20 years, and the budget deﬁcits were the highest in history. In contrast to his position at the beginning of his term of oﬃce, Reagan signed a $9.6 billion relief package in

The origins of the new welfare law

117

1983, almost half of which was earmarked for the creation of 300,000 to 600,00 public service jobs. Five years later, in October 1988, just one month before the presidential election, he signed the Family Support Act, another piece of legislation that was primarily the work of Senator Moynihan. A major feature of the Family Support Act was the creation of the Job Opportunities and Basic Skills (JOBS) program, which required single parents with children over three to work for beneﬁts, or enroll in education or job training. Expenses for child care and transportation were to be paid for recipients for one year after they obtained a job, and Medicaid coverage was also extended for that period. Finally, stiﬀ new child support enforcement measures were introduced in order to aid single parents to become self-suﬃcient. In spite of $6.8 billion in funding for this bill, the measures were not widely implemented by the states, according to a report issued in 1992 (Hagan and Lurie, 1992). Ronald Reagan was succeeded in oﬃce by George Bush, a president who was more concerned with foreign than domestic aﬀairs and who lost the presidency to Bill Clinton in 1992, having made very little mark on poverty or welfare issues, other than an increase in minimum wages from $3.35 to $4.25, and the passage in 1990 of the Americans With Disabilities Act, an act that was relatively uncontroversial at the time, cost the government little, and allowed a large number of disabled people to enter the workforce. The Clinton Administration came to power promising a major break with past policies.

NOTES 1. Passed in 1536, this law coincided with the Reformation and dissolution of the monasteries and other church property, leading to unemployment for many who had made a livelihood in the church, and loss of homes for those who had lived in church institutions. 2. See Trattner, 1994, pp. 10–13 for a fuller discussion of this legislation and its antecedents. 3. See Gary Nash (1976), quoted in Trattner (1994), p. 33.

7. The genesis of the new welfare law – the Personal Responsibility and Work Opportunity Act of 1996 (PL 104–193) Bill Clinton was a populist candidate, who sought to communicate with the people, and spent much time doing so. Sensing the general disillusionment with the federal government, and dissatisfaction with growing social problems such as poverty and crime, Clinton addressed these numerous times in pre-election speeches. At his announcement speech in October 1991, he noted that: ‘we should insist that people move oﬀ welfare rolls and onto work rolls. We should give people on welfare the skills they need to succeed, but we should demand that everybody who can work go to work and become a productive member of society’ (1991a). Several weeks later, noting that middle class families were working more hours for less money and that the inner city streets were taken over by crime and drugs, welfare and despair, Clinton spoke about a New Covenant: ‘to provide opportunity for everybody . . . (and) take government back from the powerful interests and the bureaucracy, and give this country back to ordinary people’ (1991b). He added that ‘people think their government takes more from them than it gives back, and looks the other way when special interests only take from this country and give nothing back. And they’re right’ (Ibid.). While these pre-election speeches did not contain many speciﬁcs for neutralizing special interests, they did include several policies which were intended to ‘break the cycle of dependency and help the poor climb out of poverty’ including an expanded Earned Income Tax Credit for the poor, time limits for welfare without working, education, training, child care and tough new child support enforcement laws (Ibid.). The Clinton message, combined with general disillusionment with the Bush Administration and a strong third party intervention, led to a Clinton victory, albeit a narrow one, in the presidential election of November 1992. The views expressed throughout the campaign continued to ﬁnd voice after the election, however. In his State of the Union Address of February 17, 1993, for example, Clinton ﬁrst made the oft quoted pledge to ‘end welfare as we know it’, and he publicly introduced the 118

The genesis of the new welfare law

119

concept of ending entitlement to welfare without work after a speciﬁed time period: Later this year we will oﬀer a plan to end welfare as we know it. I have worked on this issue for the better part of a decade, and I know from personal conversations with many people, that no-one, no-one, wants to change the welfare system as badly as those who are trapped in it. I want to oﬀer the people on welfare the education, the training, the child care, and the health care they need to get back on their feet. But, say, after two years, they must get back to work too, in private business if possible, in public service if necessary. We have to end welfare as a way of life and make it a path to independence and dignity.

PRESIDENT CLINTON: THE CLINTON BILL The political rhetoric was not converted into proposed legislation until some 18 months into the Clinton presidency. During this time health care held center stage and, much to the consternation of some congressional Democrats, welfare reform received relatively little public attention from the administration, although discussions between the Department of Health and Human Services (HHS) and the administration were going on behind the scenes. The most signiﬁcant action taken by the President in 1993 was the naming, on June 21, of a 27-member task force to develop a welfare reform plan. This eﬀort was led by Bruce Reed, deputy assistant to the president for domestic policy, and two assistant secretaries from HHS, David Ellwood and Mary Jo Bane. The slow pace of welfare reform in the administration during 1993 was matched by relatively little activity on the part of Congressional Democrats. The frustration on the part of some Democratic members was indicated by Senator John Breaux (D-La.) who, on January 18, 1994 called for the Congress to address welfare reform with the same urgency as health care reform (American Public Welfare Association – APWA, 1998). House Republicans in the meantime had been more active, and Minority Leader Robert Mitchell (R-Ill.) introduced a welfare reform proposal (H.R. 3500) that was cosponsored by 160 Republican members in November 1993. Among other provisions this proposal required work in return for beneﬁts, allowed states to convert AFDC matching funds to block grants, required paternity establishment in exchange for beneﬁts and denied beneﬁts to minor unwed parents. In the following six months, some half dozen bills were introduced, (2 Republican, 2 Democrat and 2 bipartisan), including Senator Dole’s (R-Kan.) Welfare Reform Act of 1994 (S. 1795).

120

The political economy of welfare reform in the United States

Arriving rather late upon the scene was the Work and Responsibility Act, or the Clinton Bill, as it was known, which was introduced into the Senate by Senator Moynihan (S. 2224) and simultaneously into the House by Rep. Gibbons (H.R. 4605) on June 21, 1994. The stated purpose of the Clinton Bill was to amend the Social Security Act to revise the federal welfare system for the purpose of making AFDC a transitional gram with the goal of enabling participants to achieve maximum economic independence and self-suﬃciency by, among other changes, imposing time limited AFDC beneﬁts and requiring participation in modiﬁed and new state-administered job training and subsidized employment programs that have been designed to eventually move them into the permanent work force and prepare them for a life without welfare by enabling them to get work experience and by requiring them to perform job searches for suitable nonsubsidized employment. The major elements of the Clinton Bill were: i) a two year time limit on beneﬁts for adults who were not working; ii) a requirement that all recipients able to participate in JOBS do so, on a phased in basis, starting with those born on or after 1972; iii) a requirement that the states establish a new program, WORK, for those without jobs after two years. This would be ﬁnanced with federal matching funds. Wages would be paid for WORK hours, and 15 to 35 hours would be required of participants. AFDC payments could be used to supplement WORK payments where necessary. Mothers of children under 1 year were exempt. iv) a state option to deny beneﬁts for additional children born to a welfare mother; v) a requirement that unmarried minor mothers live at home in order to receive beneﬁts; vi) a state option to implement more liberal rules regarding two parent families and more generous discounting of all recipients earnings; vii) a requirement that states disregard a higher asset limit when assessing eligibility or beneﬁts; viii) a requirement that states implement individualized employability plans for welfare recipients that assessed, inter alia, literacy skills and need for job training, substance abuse, child care or other services: the recipient, on his or her part, was required to sign a personal responsibility agreement; ix) a requirement that the sponsors of legal aliens be ﬁnancially responsible for them for a more extended time period (this was mainly a ﬁnancing measure), and x) provisions to enhance child support enforcement.

The genesis of the new welfare law

121

While some of these provisions constituted fairly radical departures from existing federal regulations, many of them had been implemented by some of the states on an experimental basis under state waivers from AFDC. These waivers, initiated in the Reagan administration, had expanded to a greater number of states during subsequent administrations. Time limits had been introduced in several states. Vermont, for example, under its waiver imposed a 30 month time limit for receiving beneﬁts without work, while Colorado imposed a 2 year time limit. Similarly, several states, including California and New Jersey gave no increase in the beneﬁt payment to a mother who had an additional child while on welfare (the family cap provision). Other states disregarded a higher level of family assets than federal regulations prescribed, in order to encourage savings, and/or disregarded a higher level of earnings to encourage work. Thus, the Clinton Bill was to some extent merely codifying regulations that states had devised in their experimental programs under AFDC waivers. While the Clinton proposal for a time limit on welfare was welcomed by the Republican leadership in both Houses, and a number of bills were introduced with diﬀerent versions of these limits, the Democratic party was initially less than wholehearted in its endorsement. By May 1994, however, the Democratic Leadership Council praised the call for time-limited assistance as a ‘great conceptual leap forward’ and said his pledge deﬁned Mr Clinton as a ‘diﬀerent kind of Democrat’ (Burke, 1996). The Clinton Bill was referred in the Senate to the Finance Committee, and the House referred it to three separate committees and seven subcommittees over the course of the following three months. The 103rd Congress took no action on these bills, however, other than holding several committee and subcommittee hearings, and when the Republicans swept to victory in the mid term elections of November 1994, the window of opportunity for the Democrats was ﬁrmly closed.

REPUBLICAN POLITICS AND IDEOLOGY: THE CONTRACT WITH AMERICA While the Clinton Administration had been focusing on health reform, House Republicans had been pushing their own welfare bill, H.R. 3500, which was endorsed by nearly every House Republican (Haskins, 1999). This bill reﬂected some of the ideas initially put forward by the Wednesday group, a band of about 40 House Republicans led by conservative Vin Weber and moderate Bill Gradison. As early as October 1991 this group had published a paper entitled ‘Moving Ahead: Initiatives for Expanding Opportunity in America’ which recommended, inter alia, mandatory work

122

The political economy of welfare reform in the United States

and time limits. In the Spring of 1992, a member of this group, Clay Shaw, the senior Republican on the Welfare subcommittee of the Ways and Means Committee, had issued a report which declared that: ‘the major goal of the Republican welfare policy is to ensure that families willing to work will be better oﬀ ﬁnancially once they leave welfare and to achieve this goal, not by cutting welfare beneﬁts, but by subsidizing work’ (Haskins, 1999, p. 4). By the early 1990s, making work pay was a major item on the Republican welfare agenda and it was an integral part of H.R. 3500. While H.R. 3500 was generally popular with House Republicans, some conservative interest groups were less than supportive. In fact, Empower America, a think tank headed by prominent conservatives such as William Bennett, Jack Kemp, and Vin Weber who had by this time retired from Congress, publicly denounced some of the proposals made by the task force that was working on the Republican bill (Haskins, 1999). Their major criticism was that illegitimacy and not work requirements should be the major focus of Republican welfare reform. In order to further shore up support for their legislation, Newt Gingrich, the second ranking Republican in the House, asked Republicans on the Ways and Means Committee and the Education and Labor Committee to work with conservative interest groups such as the Heritage Foundation, the Christian Coalition and others that would appeal to the conservative activist base of the Republican party. Illegitimacy turned out to be one of the most divisive issues with which the Republican House had to deal. Conservative Republicans in alliance with conservative interest groups such as the Christian Coalition, the Heritage foundation, Empower America and others, threatened to publicly oppose the legislation unless stronger illegitimacy measures were adopted. A compromise between the extreme right, who would have ended all welfare for illegitimate children, and more moderate groups resulted in a bill which denied beneﬁts to children born to unwed minor mothers or to a mother already on welfare. Both provisions subsequently passed the House but were defeated in the Senate which allowed states the option of excluding these groups. The result of the collaboration between House Republicans and conservative interest groups was H.R. 4, the Personal Responsibility Act, which was incorporated into the Contract With America, the centerpiece of the Republican platform in the 1994 mid term election campaign. The Contract With America was comprised of ten separate bills, of which welfare reform was the third.1 The origins of this document were both ideological and political. The Republicans needed, according to Rep. Dick Armey (R-Texas), ‘to demonstrate that with a Republican majority you would get a contract to actually get certain things done’ (Congressional Quarterly, 1994, p. 2712). An advisor to the Republican campaign, Frank Luntz, advised GOP leaders that they needed ‘a fresh approach’ if they

The genesis of the new welfare law

123

hoped to attract those skeptical voters who had voted for Ross Perot in 1992 (Ibid.). In a memo to the Republican leaders in September 1994 Luntz said: ‘After completing both an extensive telephone survey (. . .) and focus groups sessions with swing voters, I can say with conﬁdence that the Contract is our best hope of winning back Perot voters, disgruntled Republicans and conservative Democrats’ (Ibid.). (Some Perot strategists did not share this assessment, however and John White, the architect of Perot’s 1992 agenda, dismissed the contract as ‘pure chicanery’.) The Contract had something for almost everyone: a balanced budget amendment, a line item veto, tax cuts for some Social Security recipients, and tax credits for families with children, in addition to welfare reform. The power of the document as a political instrument was evidenced by the strong negative response from concerned Democrats in Congress and in the Administration, who labeled its economic content as Voodoo II. It was also conﬁrmed by the Republican success in the 1994 elections.

REPUBLICAN LEGISLATIVE PROPOSALS The Republicans had promised to bring to a vote within the ﬁrst 100 days of the 103rd Congress, the package of 10 bills, which constituted the Contract With America. In a ﬂurry of activity, they achieved considerable success in this goal. However, the third bill, the Personal Responsibility Act, which dealt with welfare reform and which was cosponsored by 110 Republicans and introduced as H.R. 4 on January 4, 1995, was far broader in scope than the Clinton proposals and went through many transformations before a ﬁnal version of welfare reform, was passed. The major provisions of the Personal Responsibility Act (H.R. 4) were: 1. elimination of payments to unwed mothers under 18, or 21 at state option (the savings from which would be returned to the state for speciﬁed purposes including orphanages, group homes and abstinence programs); 2. lifetime limit on assistance of ﬁve years; 3. state option to terminate beneﬁts after two years, after which neither beneﬁts nor subsidized jobs would be available; 4. establishment of work programs and a requirement that by Fiscal Year 2003, 50 percent of recipients would be working in exchange for beneﬁts; 5. consolidation of nutrition programs into a block grant with lower funding; 6. denial of beneﬁts to legal immigrants and children who did not have paternity established, were born while the mother was on welfare or were born to an unwed mother under the age of 18; 7. an overall cap on funding for an array of anti-poverty programs including AFDC; and 8. state option to convert AFDC to a block grant.

124

The political economy of welfare reform in the United States

The net federal savings were estimated at $40 billion over ﬁve years. Overall, the bill was much tougher than the Clinton Bill in terms of budget cutting measures, and much weaker in terms of work and training programs and requirements. Signiﬁcant budget savings were achieved by denying beneﬁts to immigrants and certain categories of children, cutting funding for some programs and introducing a ﬁxed block grant for others. The end of open-ended funding and individual entitlement to AFDC payments were landmark changes in welfare policy and the source of bitterest opposition from liberals. A further major diﬀerence with the Clinton Bill was the diﬀerent emphasis on work. Under the Republican Bill welfare recipients were required to work but there was no guarantee that a job would be available for all those willing and able to work. The Clinton Bill contained several measures dealing with this. The popularity of welfare reform (and, perhaps, the involvement of numerous interest groups) was evidenced by the fact that over two dozen diﬀerent bills dealing with this topic were introduced during the 103rd Congress, most of which died in committee. Not surprisingly, the Republican proposal was strongly opposed by a number of Congressional Democrats including House Democratic Leader Richard Gephardt, who, in a February 10 news conference charged that H.R. 4 as proposed ‘does absolutely nothing’ to replace welfare with work (Burke, March 6, 1995b). In a similar vein, the Secretary of the Department of Health and Social Services, Donna Shalala, said Administration concerns about the Republican plan included: ‘no requirements for job search, education or training, no deﬁnition of required “work activities” and “very low” participation standards’ (Ibid.). (‘Participation standards’ refers to the number of welfare recipients that a state must have enrolled in work or work related activities.) While public hearings on the bill were held in January, it was, according to a Congressional Quarterly report, ‘a series of closed door meetings among inﬂuential Republican House members, aides and governors that changed the face of the welfare plan’ (Congressional Quarterly, 1995a, pp. 7–36). The Republican Governors Association (RGA)was represented by Govs. John Engler of Michigan, Tommy Thompson of Wisconsin, and William Weld of Massachusetts, who agreed to accept limited federal funding in return for vastly expanded state control over welfare and related social services. Transforming AFDC into a block grant fulﬁlled both of these requirements. Democratic governors were not on board, however. While supporting greater state ﬂexibility, they feared that an end to entitlement status might lead to ﬁnancial diﬃculties in an economic downturn. The lack of consensus led to a bill that some governors from both parties criticized, fearing insuﬃcient funds during recessions. Some also objected to denying aid to

The genesis of the new welfare law

125

immigrants, and children born to unwed minors and those already on welfare. The legislation was referred by the House to the three committees of jurisdiction: Ways and Means, Agriculture, and Economic and Educational Opportunities. In the House, the Ways and Means Committee began hearings in January, and approved the welfare bill on March 8. The bill coming out of Ways and Means, HR 1157, proposed two block grants for cash welfare and child welfare, gave states greater control over programs and limited federal funding. Signiﬁcant program savings were made by provisions to deny beneﬁts to immigrants and to reduce eligibility for the Supplemental Security Income (SSI) program. While states had discretion over the manner in which programs were implemented and the determination of who would be eligible, they were to be held accountable for certain end results, including requiring work participation of welfare recipients and reducing caseloads. Thus, 10 percent of recipients had to be in work activities by 1998, and 20 percent by 2003. The block grant could be reduced by 3 percent if a state failed to achieve the work participation standard. A further performance incentive was the provision of a cash bonus for reducing illegitimacy. The committee also changed the funding formula to provide additional funds to states like Texas and Florida that had the fastest growing welfare caseloads. Funding was to be distributed in proportion to funding received for AFDC and related programs in 1994, or an average of 1991–1994, whichever was larger. In the course of working out these provisions several dozen Democratic amendments had been dismissed with little or no debate. The Economic and Educational Opportunities Committee dealt with its area of jurisdiction with as much dispatch as had the Ways and Means committee. Republicans here had an even more uniﬁed front and dismissed nearly two dozen Democratic amendments, many with little debate, before passing HR 999, in a party line vote and over strenuous Democratic objections. They recommended three block grants – for child care, school meals and family nutrition programs. The Clinton Administration and the Department of Agriculture attacked the bill saying it would jeopardize children’s health by restricting funding and eliminating nutrition standards. The department estimated that the two nutrition block grants would provide $7.3 billion less funding over ﬁve years. Republicans argued that consolidating the programs would save money and reduce paperwork (Congressional Quarterly, 1995a). The House Agriculture committee passed a bitterly contested bill, HR 1135, that cut the food stamp program, capped federal funding, denied beneﬁts to most legal aliens, and required recipients to work. Although states were given more discretion and funding was capped, a block grant was not proposed. The degree of rancor on the usually bipartisan committee was reﬂected in an amendment circulated by Harold Volkmer (D-Mo.)

126

The political economy of welfare reform in the United States

to name the bill the ‘Food Stamp and Commodity Reduction to Make Americans Hungry Act’. ‘In my 15 years on this committee, I have never had anything this outrageous’, said a red-faced Bill Emerson (R-Mo.), committee sponsor of the bill (Ibid., pp. 7–42). The three committee bills were amalgamated into HR 4, modiﬁed slightly to, inter alia, make legal immigrants eligible for more programs. The House opened debate on March 21 and passed (245–178) H.R. 4 on March 24 on a largely party-line vote. The House bill contained a few changes from committee provisions, notably allowing more immigrants to be eligible for programs, contingent on their sponsor’s income. The bill promised savings of $66.3 million over ﬁve years. All but nine Democrats opposed it, the administration opposed it and so did some anti-abortion groups who feared that denying welfare to unwed teenage mothers would increase abortions. An indication of the dissatisfaction is reﬂected by the more than 150 amendments that were proposed, most of which were not given consideration. Among the amendments passed was a Republican proposal to allow savings from the welfare bill to oﬀset a proposed tax cut, increased funds for child care, a requirement that states have provisions to suspend driver’s licenses and other licenses from those persons delinquent in child support payments, and restrictions on use of cash welfare funds for medical services including abortions. The Democrats, in the meantime, stood solidly behind an alternative bill crafted by Nathan Deal of Georgia and other moderate to conservative Democrats. This bill, HR 1267, oﬀered as a substitute for HR 4, had a number of similar provisions, including time limits, work requirements and restricted eligibility for immigrants. It would also have increased spending on education, job training, employment services, and day care to facilitate recipients’ participation in the Work First program. A major diﬀerence was retention of entitlement status for cash welfare and other programs. It would also have maintained control of welfare programs at the federal level. It was defeated 205–228 largely on a party line vote. The Republican welfare proposal also had opponents among the ranks of Senate Democrats, and passage through the Senate was considerably more diﬃcult than passage through the House. In the end, the Senate, expected to be a moderating inﬂuence on the legislation, passed legislation that incorporated most of the key provisions of the House bill, including transforming AFDC into a block grant. The proposed legislation initially went to the three committees of jurisdiction, and the reported bills from these committees were then combined into one bill. The Finance Committee, which had jurisdiction over most areas of welfare reform, on May 26 approved a draft welfare bill written by Chairman Bob Packwood of Oregon and subsequently presented as a substitute for HR 4. Like the

The genesis of the new welfare law

127

House bill, the measure would have ended the entitlement to cash welfare, allowed states to determine eligibility and introduced block grant funding. Time limits were similar to those contained in the House bill. This bill did not require states to deny beneﬁts to immigrants and children of unwed minors or those already on welfare. Instead it gave states the option to do so. Suspension of licenses for those delinquent in child support was also rendered optional. However, states were required to guarantee child care to welfare recipients who had children younger than six and who were obliged to participate in work or training activities. The Labor and Human Resources Committee supported a bill (S 850) which basically maintained the status quo for child care, reauthorizing the 1990 Child Care and Development block grant and incorporating additional programs. The bill was passed with bipartisan support. The Agriculture, Nutrition and Forestry Committee endorsed a bill (S 904) to cut food stamp spending, give states more control, and require work. All but one committee Democrat opposed this bill. While all of these bills had been reported out of committee by the end of May, it took several more months before a Senate welfare reform bill was brought to the ﬂoor. The delay was largely caused by serious intraparty disputes, primarily about formulas for distribution of block grant funds and the absence of provisions for reducing illegitimacy. Conservative and Southern senators were among the most prominent critics of the bill’s failure to adequately address illegitimacy, arguing that this ‘allowed the states to subsidize children born out of wedlock, which perpetuates poverty’.2 No such ideological sentiments underlay the funding issues. States had huge sums of money at stake dependent on the formula used to allocate the lump sum block grant allocation among them. At the heart of the dispute was whether states would receive an amount tied to their traditional spending on the programs – in which case states providing higher beneﬁts would stand to gain in comparison to poorer southern states, for example, or whether the formula would be based on demographics such as the number of poor children or the growth of population. A further ﬂy in the ointment appeared when Senator Packwood became the subject of an Ethics Committee investigation which ultimately caused his resignation. Senator Dole had stepped in by this time, and he presented a bill (S 1120) based on the initial three bills but with revisions to attract broader party support from conservative and moderate senators. Funding was based on traditional spending but an additional sum was allocated to states with high population growth and low welfare beneﬁts to appease senators from the south and west. To appease conservatives, states were given an option to deny beneﬁts to children of minors or those already on welfare. These provisions proved inadequate: after bringing the bill to the

128

The political economy of welfare reform in the United States

ﬂoor on August 7, and listening to a day and a half of opening speeches, Dole pulled the bill, which had been oﬀered as an amendment to the House bill, HR 4. This delay made it more likely that parts of the bill would be incorporated into the budget legislation in the fall so that Republicans could count the welfare savings as a contribution towards deﬁcit reduction. The Democratic leadership in the Senate, in the meantime, had introduced an alternative bill (S 1117) on June 8, which retained the federal guarantee of cash aid for the poor, but made it conditional on a recipient working or taking active steps to ﬁnd work. It included additional funding for welfare to work activities and job training and search. The Senate took up the Republican sponsored bill again in September. After voting down two alternative Democratic welfare bills, including the Senate minority leadership bill (S 1117) and a bill by Senator Moynihan, the Senate held two weeks of ﬂoor debate. During this period Senate Republican moderates joined with Democrats to chip away at some of the more conservative provisions of the bill that Dole had crafted. In the ﬁnal proposal, welfare recipients were exempt from work if they had children under age ﬁve and could not obtain the necessary child care. States had the option of denying beneﬁts to unwed minor mothers and children of those already on welfare. States were required to maintain welfare spending of at least 80 percent of 1994 levels for ﬁve years, and child care funds were increased. With these amendments, the Senate passed a revised version of H.R. 4 on September 19, six months after the House had passed its bill. As the vote neared ﬁnal passage, Dole declared: ‘We are closing the books on a six-decade-long story of a system that may have been well intentioned but . . . failed the American taxpayer and failed those who it was designed to serve’ (Ibid., pp. 7–48). Paul Wellstone (D-Minn.), the only one of the bills’ opponents seeking reelection in 1996, said that children would suﬀer if it was enacted. ‘They do not have a lobbyist. They do not have the PACs. They are not the heavy hitters’, he said (Ibid.). The bill went to conference in late October. Democrats were virtually excluded, with Republicans meting out diﬀerences among themselves. By this time welfare reform was moving on two tracks: major elements of the reform, including those that involved budgetary savings, were incorporated into the budget reconciliation bill (H.R. 2491) at the same time as House and Senate conferees were working on a ﬁnal version of HR 4. HR 2491 was vetoed by the President on December 7, 1995. Among the primary areas of dispute between House and Senate was the House proposal to change funding for ﬁve major groups of programs to block grants. In addition to cash welfare, these included: child protection; child care; school lunch and breakfast, and the Nutritional Program for Women, Infants and Children (WIC). Denying beneﬁts to the children

The genesis of the new welfare law

129

of unwed minor mothers or those already on welfare were also contentious issues. Senate moderates generally wanted only two block grants – cash welfare and child care; they wanted to give states more choice on eligibility decisions, increase child care funds and ensure states did not make major cuts in welfare spending. Joining with Senate Democrats, Senate Republicans succeeded in achieving signiﬁcant compromises. Despite strong opposition from both the administration and a signiﬁcant number of congressional Democrats, Congress passed the conference agreement on H.R. 4 in December, 1995. In the bill that came out of committee, block grants were limited to cash welfare and child care and some child protection programs, and states had the option to deny beneﬁts to children of minors and those already on welfare. States were required to maintain funding at 75 percent of their 1994 levels on AFDC and related programs. The conference agreement on H.R. 4 contained several elements that diﬀered substantially from the original bill. A major innovation was the compulsory replacement of AFDC and JOBS (which were based on federal-state matching funds) by a ﬁxed block grant to the states for the TANF program. This was a state option in the original bill. Beneﬁts were conditional on work after two years, but under the terms of the conference agreement, states themselves had more latitude to deﬁne what constituted work activities, and they were obliged to enroll only 4 percent rather than the proposed 8 percent of families in work or work activities. Both versions maintained the lifetime limit on welfare beneﬁts of ﬁve years. States, which had always had wide latitude in setting beneﬁt levels, were also given more discretion over eligibility rules, such as limits on resources or income. The actual levels of block grants were based on funds the states received in FY1994, FY1995 or an average of FY1992–1994, whichever was greater. Since caseloads were declining from an all-time high in March 1994, the states stood to gain millions of additional federal dollars under TANF compared to what they had received under AFDC. Most of the provisions agreed upon by the conference committee working on the free standing welfare bill had been incorporated into the budget reconciliation bill which was vetoed by the President on December 6. One provision that had remained highly divisive was whether states should be given control over child nutrition programs by funding them with block grants. The chairman of the House Economic and Educational Opportunities Committee wanted to give states this control, while the Senate Agricultural Committee Chairman was adamant that child nutrition programs such as school lunches should remain under federal control. The ﬁnal compromise allowed seven states to be funded with block grants. Other last minute changes increased funding and

130

The political economy of welfare reform in the United States

reduced spending cuts for several programs. The House adopted the conference report on December 21, and the Senate the following day. As promised, it was vetoed on January 9, 1996, by President Clinton, who found it ‘tough on children and weak on work’ (Burke, 1996). Within a month after the Presidential veto, the NGA voted unanimously to propose modiﬁcations to H.R. 4, including more federal funds for child care, and grants to states with high unemployment. These proposals addressed the issues of the bill being ‘tough on children and weak on work’. A new version of the bill, introduced in May 1996, the Kasich Budget Reconciliation/Welfare Reform Bill (HR 3734), was introduced June 27, 1996. This legislation moved through the House and Senate with amazing speed, perhaps due in large part to the desire of Members to return to their districts to campaign in an election year. (And the fact that caseloads had been declining for almost two years, and if they continued to do so and the block grants were tied to a readjusted level, the governors would lose substantial funds.) The bill passed the House only three weeks after it was introduced on July 18, 1996 by a vote of 256–170. The bill was received in the Senate on July 18 and read twice. It took only 5 days for the bill to pass in this chamber. On July 23, the Senate passed H.R. 3734 by a vote of 74 to 24, with one Republican and 23 Democrats voting against the measure. Following a motion on July 24 that the House disagree with the Senate amendment, the Bill was sent to conference on July 25. A conference report was ﬁled in the House on July 30, and the House agreed to the conference report by 328–121 on July 31, while the Senate did likewise by a vote of 78–21 on August 1. The bill was presented to the President on August 19, 1996, and signed by him on August 22, to become Public Law No: 104–193. This new document, like the original H.R. 4 and the Clinton Bill, required work after two years of beneﬁts and a lifetime limit of ﬁve years of assistance. A major diﬀerence between the two was the replacement of matching funds and entitlement status of AFDC by capped block grants. This had the eﬀect of ending entitlement to cash welfare for the ﬁrst time in decades. The changes which had been wrought in the welfare system of America were brought about by a combination of economic factors, ideological and political motivations, and the activities of interest groups. Traditional theories of public ﬁnance which hold that government intervention arises in response to an externality or public good problem, such as (in the case of welfare) the inability of local communities to deal with large numbers of poor, justiﬁes government intervention on the grounds that transactions costs render these activities too costly for the private sector, and thus they

The genesis of the new welfare law

131

would be underprovided or not provided at all. In the early years of welfare relief this may indeed have been the case. However, as the preceding account of the development of welfare policies and programs suggests, the forces that currently shape government policies are far more complicated than a simple desire for the common good. The process is, to a large extent, the result of the interplay of numerous complex and often conﬂicting interests. The next chapter attempts to shed some light on the objectives and interactions of those individuals and institutions involved in the policymaking process and presents an analysis of the eﬀects of interest group activities on the American welfare system.

NOTES 1. Other bills included: The Fiscal Responsibility Act, which proposed a balanced budget amendment and line item veto for the President; the Taking Back Our Streets Act, which dealt with crime; the Family Reinforcement Act dealing mainly with tax credits for elderly care at home and adopting children; the American Dream Restoration Act which provided expanded IRAs and tax credits for families with children; the National Security Restoration Act dealing with defense issues; the Senior Citizens Equity Act which increased the earnings disregard for seniors and reduced the amount of Social Security beneﬁts that were taxable; the Job Creation and Wage Enhancement Act which proposed reducing capital gains and providing more generous business tax deductions; the Common Sense Legal Reforms Act dealing mainly with product liability; and the Citizen Legislature Act, providing for a term limits constitutional amendment. 2. Statement of Senator Faircloth of North Carolina, quoted in Congressional Quarterly, 1995a, pp. 7–47.

8.

Institutional analysis

In the early chapters of this book two alternative theories of government were discussed. The more traditional economic theory of welfare economics argues that government exists to correct an economic problem such as an externality or the existence of a public good which the private market would not or could not address. This theory holds that government activity in such a context provides a mechanism for moving the economy to a Pareto superior position, and, in equilibrium, a Pareto optimal situation is achieved. Certain assumptions are clearly implicit in this analysis, including both the assumption that government has all the requisite information to design appropriate policies and programs and that the political will exists to use the power of government for this objective. The alternative theory of government which has been presented in this book evolves from the ﬁeld of public choice. In contrast to the normative public welfare perspective, this view takes a positive approach, and directly addresses the institutional context of governmental activity. Applying the standard assumption of economic theory as applied to the private sector, that rational individuals pursue their own self-interest, the theory can be used to explain or predict the eﬀect of such behavior on a particular outcome in the public sector. Thus, in the case of welfare, the legislative outcome would be predicted to be a result of the inputs of self-interested individuals acting either individually or, more often, in the context of a particular interest group and constrained by the institutional context of the legislative process. This research was undertaken to examine the validity of these two theories and speciﬁcally their usefulness in explaining government welfare policy. The focus of the analysis has been the major federal cash welfare program in the United States, formerly known as Aid to Families With Dependent Children (AFDC), and since 1996 renamed Temporary Assistance to Needy Families (TANF). The program provides cash welfare to families and children with income and assets below a certain threshold (and fulﬁlling certain other requirements). The 1996 New Welfare Law described in the previous chapter transformed this program in a manner unknown in its 30 year history. A close examination of this legislation, and particularly the legislative process that gave birth to it, serves to highlight the usefulness of the two theories being examined. 132

Institutional analysis

133

In order to assess the validity of alternative theories it is necessary to compare the outcomes that would be predicted by each theory. This in turn requires an assessment of the objectives sought from the legislation. We turn ﬁrst to the traditional welfare economics. The classical welfare economic theory posits that government intervention in the welfare arena may be justiﬁed and explained by a number of factors – all generally related to the disutility of poverty in society. This disutility may arise from several sources, including the purely economic – such as the underutilization of human capital when people are not gainfully employed, altruistic sources such as the desire to improve the lot of others, or simply discomfort at observing poverty in the society. According to this line of argument the primary objective of government policy would be the elimination of poverty and putting people to work. In fact, the stated goal of both the Clinton Administration and the Republican leaders of the Congress emphasized getting people oﬀ welfare and into the workforce rather than reducing the numbers in poverty, thus emphasizing the economic rather than the ultraistic. A further major source of disutility, particularly for Congressional Republicans, was the belief that federal funds spent on welfare programs represented a misallocation of resources since such programs were ineﬀective in reducing poverty, and according to some analysts, actually worsened the situation by providing disincentives to work, creating an underclass that led to the perpetuation of poverty. Traditional economic welfare theory would predict that government would remedy the disutilities outlined above, and bring the economy closer to a Pareto optimal equilibrium. The implicit assumptions are that suﬃcient information is available to the government and that it has the capacity and will to implement an eﬀective solution. While the plethora of research from all ideological perspectives renders the ﬁrst assumption more or less plausible, there are many who would claim that the government by its very nature has limited capacity to eﬀectively implement welfare programs (and that the mechanisms it uses to do so in fact exacerbate the problem). The second theory that we are examining is that of public choice. In contrast to the ﬁrst, the emphasis here is on the individuals and institutions involved in the process, as well as on the outcomes of the legislation. Assuming rational self-interest, the objectives of each agent, individual or institution, may diﬀer and may conﬂict. Rather than some societal good, the ultimate outcome of the process will depend on the objectives of the agents involved and the balance of power among them. According to Mancur Olson’s theory of interest groups, smaller groups and those that can oﬀer speciﬁc beneﬁts to members are likely to be relatively more eﬀective in achieving their goals than are larger groups with free rider

134

The political economy of welfare reform in the United States

problems. When this theory is applied to the legislative arena, the successful groups can be deﬁned as those that succeed in obtaining legislation that furthers, directly or indirectly, their speciﬁc ends – whether ideological, political, or pecuniary. In order to elucidate what outcomes would be predicted by public choice theory in the case of government welfare policies, it is necessary to examine in more detail the nature of the principal agencies and individuals involved. For simplicity, we group them into three categories: the ideological, the political and the pecuniary. Clearly there are overlaps; those whose goals are primarily political, such as reelection, may obtain success by pursuing a particular ideological goal to which supporters are sympathetic, for example. Moreover, those that we deﬁne as primarily ideological may stand to reap pecuniary rewards from policies that reﬂect their ideology. The political group consists of the President and Members of Congress who, it is assumed, are primarily interested in getting reelected, increasing the party majority, or in the case of the President, personal legacy. Clearly, gains to one political agent or agency may or may not be at the expense of the others. The pecuniary group includes those who stand to gain from reallocation of federal welfare funds or the allocation of additional funds, including HHS, state governments, non-proﬁt and private sector welfare service providers, members of the business community and welfare recipients. Admittedly the water is somewhat murky here in the sense that state governors, for example, would like to increase the ﬂow of funds going to their state in order to enhance their own political status. The pecuniary motive precedes the political, however. Similarly with HHS oﬃcials, many of whom undoubtedly have an ideological commitment. The ideological group we deﬁne as those whose primary motivation is employing the legislative process to promote an ideological rather than purely economic outcome. This group includes two very diﬀerent categories. On the one hand are the faith organizations, liberal think tanks and others who are committed to the belief that the government has a moral responsibility to guarantee a minimum set of beneﬁts for the poor. At the other end of the philosophical spectrum are those conservative groups whose emphasis is on personal responsibility rather than government responsibility, such as the Christian Coalition and the Heritage Foundation, and who strongly believe that moral values, and particularly illegitimacy, should be at the center of any welfare reform legislation. The following section examines the speciﬁc goals of selected institutions in each of these categories, and the ways in which they strove to achieve these goals by participating in the legislative process that resulted in the new welfare law, P.L. 104-193.

Institutional analysis

135

SELECTED INSTITUTIONS THAT INFLUENCED THE LEGISLATIVE PROCESS OF WELFARE REFORM The story of the new welfare law is full of ironies, not least of which is the transformation of the President’s welfare reform proposal into a piece of legislation designed largely by the most conservative Republican Members of Congress and Governors and reﬂecting the values of some of the country’s most conservative interest groups. Clinton’s original proposal, made during his campaign, focused on work, but also on expanding education and job training and providing public sector employment for those unable to ﬁnd private sector jobs. It would have increased federal expenditures. The Republican Bill that was ﬁnally passed cut expenditures, emphasized work or work activities such as job search and, most radically, ended the entitlement of individuals to welfare. In order to understand how the legislative process transformed welfare reform it is necessary to look brieﬂy at the internal workings of the Clinton campaign and presidency. When Bill Clinton was pledging to reform the welfare system he was sending a politically popular message to the voters. It was a message that had diﬀerent meaning for diﬀerent audiences, however. The ﬁne print of campaign documents such as ‘Putting People First’ revealed that the Clinton team meant giving more assistance to welfare recipients looking for work, not less (Clinton Campaign Economic Plan, 1992). Some of the Clinton rhetoric insisting that, ‘If you can go to work, you ought to go to work’ gave a diﬀerent impression, however, and voters in general felt that too much money was spent on welfare, that the system was abused by many welfare recipients, and that there should be stricter work requirements (Stephanopoulos, 1999). This ambiguity allowed the Congress to transform the meaning of welfare reform. The ability of the President to advance his welfare agenda was severely damaged by three circumstances: a continuing series of internal scandals and external crises, internal disarray among his advisors, and the huge budget deﬁcits and sluggish economy that he inherited. The scandals began during the 1992 campaign with the Gennifer Flowers stories and the draft dodging revelations. Problems continued during the transition, with several Clinton nominees facing problems, and Zoe Baird forced to resign when, it was learned she had hired illegal immigrant household help and had not paid social security tax. The litany continued as the new administration took over; in May 1993 the ﬁring of a White House travel oﬃce employee and subsequent replacement by a Clinton cousin caused a furor in the press, surpassed in July by the suicide of White House Counsel Vincent Foster which led ultimately to the fateful Whitewater investigations. Five months later the Paula Jones story broke and in

136

The political economy of welfare reform in the United States

February 1994 at the annual convention of the Conservative Political Action Committee she publicly accused Clinton of sexual harassment (see Woodward, 1994, and Stephanopoulos, 1999, for fuller discussion of these events). By Summer of 1994, the Republicans in Congress were having a ﬁeld day, mounting rapid ﬁre attacks on the administration and holding numerous public hearings on Whitewater which drew constant media attention and served to distract senior administration staﬀ from working on legislative issues. While the House Republican leadership had assigned committee members and their staﬀ to focus on developing welfare reform legislation and enlisted a great deal of help from the nation’s Governors in doing so, the White House staﬀ was preoccupied with damage control activities, and was itself hampered by internal conﬂicts and disarray. The disarray was ﬁrst evident in the transition during which there was no central authority, the President was working on his cabinet selections and ‘a lot of people were doing a lot of separate work’ (Woodward, 1994, p. 80). On the ﬁrst Monday of the ﬁrst full week in the White House, a column in Time carried a challenge from a ‘top administration oﬃcial’ to Senator Daniel Patrick Moynihan, chairman of the powerful Senate Finance Committee. ‘He’s not one of us . . . We’ll roll right over him if we have to’ (Stephanopoulos, 1999, p. 121). Moynihan’s committee was responsible for all the major programs on the Clinton agenda – health care, welfare and the budget. In that same week, the President’s statements on gays in the military caused a showdown with the Joint Chiefs of Staﬀ, and three people were killed in a shooting at the CIA. This was not an auspicious start for the new administration, but it was to be a precursor of much of the same. Further disarray resulted from conﬂict between the political advisors, who had worked on the Clinton campaign and were focused on carrying out the agenda, and his economic advisors, who were concerned about the budgetary impacts of campaign pledges. Clinton had promised during his campaign to have an economic recovery plan within the ﬁrst 100 days of his Presidency. In preparation for this he met with Federal Reserve chairman Alan Greenspan in Little Rock on December 3, 1992 (Woodward, 1994) (the following account is from Woodward, 1994). During this encounter Greenspan made the case for giving the highest priority to reducing the deﬁcit. He argued that long-term interest rates were an unusually high 3 to 4 percent higher than short term rates because bondholders and traders felt that the budget deﬁcit would continue to soar, and history had shown that high inﬂation would result. The gap in the rates represented an inﬂation premium, Greenspan explained, and could only be eradicated by changing market expectations regarding inﬂation, which meant that Clinton needed a credible economic plan for reducing the deﬁcit. Lower long-term rates

Institutional analysis

137

would in turn lead to expanded business investment and consumer spending thus leading to economic growth. Further, it would encourage a switch from bonds to stocks thus also helping the stock market. Although an economic plan had been promised in the ﬁrst 100 days, the ﬁrst oﬃcial meeting was not held until January 7, 1993. Over the following weeks, as it became clear that it would not be possible to pass a middle class tax cut, increase investments and cut the deﬁcit, arguments increased between the economic team, mainly deﬁcit hawks, and the political advisors concerned about political fallout. In the meantime, what many people in the White House did not know was that Treasury Secretary Lloyd Bentson had also met with Greenspan and together they had agreed that a deﬁcit reduction of $140 billion would be required to be credible in the bond market. It was Lloyd Bentson who ﬁrst suggested dropping welfare reform in order to achieve $3 billion in saving towards the $140 billion target (Woodward, 1994). Budget Director Leon Panetta, and National Economic Council Director Robert Rubin supported Bentson, and Clinton, much to the dismay of Labor Secretary Robert Reich, George Stephanopoulos, James Carville and the political advisors, ﬁnally agreed to drop welfare reform at a meeting on Saturday, February 13, 1993 (Ibid.). Clinton was said to have remarked a few weeks later that Senator Moynihan was very angry that welfare reform was dropped out of the budget, ‘and he’s right’ (Ibid., p. 165). Clinton did eventually pass a deﬁcit reduction budget, which scraped through the House after months of threats, promises, carefully crafted compromises and ‘absurdly trivial deals’ such as a promise to play golf with the President (Stephanopoulos, 1999, p. 179). Similar eﬀorts were required to get through the Senate, and it was only when Senator Moynihan’s wife, Liz, used her powers of persuasion on her friend intransigent Democrat Senator Bob Kerrey, that passage was achieved in that chamber on August 6, 1993 (Woodward, 1994). Even after the passage of the budget the dissent continued. In the fall of 1993, the economic team wanted the North American Free Trade Association (NAFTA) to have priority, Hillary Clinton wanted to focus on health care, and Gore wanted to reinvent government (Stephanopoulos, 1999). As the internal disarray continued, economic growth soared in the last quarter of 1993 to an annual rate of 7 percent. Fearing inﬂationary tendencies, Greenspan made a series of short-term rate hikes between February and April. By May Clinton’s poll numbers were plummeting, the legislative agenda was ﬂoundering and Clinton felt he was a prisoner of Congress (Stephanopoulos, 1999). In that same month Kentucky’s second district elected a Republican congressman for the ﬁrst time since 1865, after campaign ads identiﬁed the Democratic candidate with Bill Clinton.

138

The political economy of welfare reform in the United States

When Congress adjourned in August 1994 without voting on health care, it was eﬀectively killed. Clinton had failed to deliver on his campaign promises of a middle class tax cut, welfare reform or health care. His major accomplishment, cutting the deﬁcits won few votes. And the persistent scandals as well as perceived ﬂip-ﬂops on issues ranging from gays in the military to policy on Haitian refugees led to widespread voter dissatisfaction which was reﬂected in the landslide Republican victories in the mid term elections. In November 1994, not a single Republican incumbent running for Governor, House or Senate lost. Republicans took control of the House for the ﬁrst time since 1954, won a majority of governorships for the ﬁrst time since 1970 and took control of the Senate. Seeing the writing on the wall for 1996, Clinton called in Dick Morris, a political consultant who had advised him in earlier campaigns and who knew how to win but ‘wasn’t scrupulous about how he did it or whom he did it for . . . His other clients were Republicans and his attack ads were the roughest in the business’ (Ibid., p. 332). His strategy for Clinton to win in 1996 was to ‘neutralize’ the Republicans and ‘triangulate’ the Democrats (Ibid., p. 334). ‘Neutralizing’ the Republicans meant passing a good deal of their proposed legislation, including a balanced budget, tax cuts and welfare reform, to relieve the frustrations that led to their election and push them further to the right on issues such as gun control and abortion rights. Triangulation meant picking a point also distanced from the liberal Democrats who were now out of favor, thus favoring a centrist position – the famous median voter position. Dick’s cardinal rule, according to George Stephanopoulos, was that if 60 percent of Americans were for it, the President had to be for it. The arrival of Morris in the White House only added to the disarray: ‘From December 1994 through August 1996, Leon Panetta managed the oﬃcial White House staﬀ, the Joint Chiefs commanded the military, the cabinet administered the Government, but no single person more inﬂuenced the President of the United States than Dick Morris’ (Stephanopoulos, 1999, p. 329). By Spring of 1995, the whole White House had become dysfunctional, according to Stephanopoulos, and rumors were rife that Morris was feeding inside information from the White House to his most prominent former client, Republican Senate Majority Leader Trent Lott (Ibid., p. 339). Leon Panetta called Morris ‘a spy in our midst’ (Ibid., p. 386). Under the direction of Morris, President Clinton started to implement the new legislative strategy of neutralization and triangulation. In 1995 neither the Office of Management and Budget nor the Congressional Budget Oﬃce analysts realized that the economy would grow so fast in the following years,

Institutional analysis

139

and given their economic assumptions it was clear that to balance the budget would require major cuts in programs that aﬀected a broad slice of the population. Clinton’s budget proposal had extended deﬁcit reduction but was not close to balance. The idea was to force the Republicans to specify cuts that they would make to balance the budget. The Republicans had indeed proposed deeper cuts and Congressional Democrats were attacking these cuts with some success. Morris wanted Clinton to oﬀer a more detailed proposal and a balanced budget. According to Stephanopoulos: It wasn’t enough for the President to balance the budget; Dick wanted to make our friends howl. He insisted, for example, that the President contrast his plan to the ‘Congressional’ rather than the ‘Republican’ budgets; and his draft praised the civility of Speaker Gingrich – acid words to our allies who had been campaigning against the Republican budgets (1999, p. 358).

The President complied on June 13, 1995. Not surprisingly the Democrats were hurt, angry and confused. Not surprisingly some questioned whose side Morris was on. The war in the Balkans and the bombing of Sarajevo slowed down the budget process over the summer, but Morris kept strategizing, and in August proposed a major ad campaign against the Republicans attacking their Medicare cuts. As a side beneﬁt Morris would get a healthy commission for every dollar they spent on television (Stephanopoulos, 1999). Morris felt that such ads combined with his behind the scenes negotiations with Trent Lott would lead to a deal. He didn’t count on the strength of the freshman Republicans. Neither did he count on increasing press scrutiny as his role became more evident. Starting in late October a series of articles in the press revealed the shadowy side of his past, and particularly the racist ads that he had been behind in previous campaigns. As Morris was losing power, the Republicans were faced with intransigent freshmen and a Speaker who was increasingly unpopular, not least of all when he complained that he would not compromise on the budget because he felt he had been snubbed by the President on a trip to Prime Minister Rabin’s funeral (Ibid., p. 404). President Clinton meanwhile was taking a strong and popular position in defense of Medicare, refusing to sign Republican legislation which cut the program. When the impasse ultimately led to two successive government shutdowns, most Americans put the blame squarely on the Republicans. Clinton was able to veto a Republican budget bill (which included welfare reform) and a separate welfare reform bill with little fear of negative consequences in the polls. The situation was diﬀerent six months later. Clinton was facing reelection in three months and was vulnerable to Republican attacks that he had failed to deliver on campaign promises. When the Republicans sent him a

140

The political economy of welfare reform in the United States

third welfare reform bill he faced a diﬃcult choice. He had been opposed to ending the welfare entitlement and to a number of provisions such as cutting aid to legal immigrants and cutting food stamps, but another veto could hurt him in the polls. If the bill passed he would have kept his campaign pledge to reform the welfare system. Leon Panetta, Robert Rubin, Laura Tyson, Robert Reich, George Stephanopoulos and other major administration oﬃcials urged a veto. Bruce Reed, the domestic policy advisor who had worked on welfare policy since 1992, felt that the bill contained much of what Clinton had originally proposed and although the food stamp and legal immigrant cuts needed to be ﬁxed, there probably would not be a chance to get a better bill from the Republican Congress. Dick Morris urged Clinton to sign. He claimed that his polls showed a Clinton veto would transform a projected 15 point November win into a three point loss. Clinton signed the welfare bill.

THE CONGRESS While Clinton had opened the door to welfare reform, it was the Republicans and not the Democrats in Congress who enthusiastically took up the issue. When they swept into power in the 1994 mid term elections, the newly energized Republican members brought with them a determination to pass the 10 bills that constituted the Contract With America. The third bill, the Personal Responsibility Act was to be the basis of welfare reform. The Republicans at this time had a popular mandate, a majority in both Houses, and opinion polls showing widespread dissatisfaction with the existing welfare system, expanding welfare rolls and budgets, and an overall huge budget deﬁcit, all of which contributed to a strong momentum for passage of welfare reform legislation. Republican leaders in the House sought out the help of the Governors almost immediately after their 1994 landslide election (see below for fuller discussion of these meetings). At the urging of Haley Barbour, Newt Gingrich and Bob Dole met with the RGA, laid out their agendas for welfare and Medicaid and requested their help saying they had neither the staﬀ nor the understanding of the programs that the Governors had. Other motivations may also have been involved, however. According to political scientist Elizabeth Drew, Bob Dole wanted the endorsement and campaign machinery of the Governors for his upcoming 1996 presidential bid and Barbour felt the Governors could help sell the Republican message to the country (Drew, 1996). One of the most contentious issues arising out of the proposed Republican legislation was the idea of funding a number of welfare programs,

Institutional analysis

141

including AFDC, by giving block grants to the states. Under existing federal regulations for AFDC, certain categories of individuals were entitled to cash beneﬁts, the levels of which varied widely and were determined by the states. Funding for the program was shared by the states and the federal government which gave matching funds to the states to cover beneﬁts and administrative costs. Thus if the rolls went up, or beneﬁt payments increased, both state and federal funding would increase. The idea of capping the federal contribution by funding the program with a ﬁxed federal block grant (states were, of course, still free to spend their own funds as they wished) gained prominence in late 1994 in discussions between Trent Lott, Newt Gingrich and the Republican Governors. Initially a broad range of federally funded programs was going to be block granted, including food stamps, child care, child welfare, AFDC, Medicaid, school meals programs, WIC and the JOBS program. This carte blanche transfer of funds to the states with few federal regulations regarding implementation of the programs provided fodder for the House Democrats who were quick to point out the possible consequences of transferring billions of dollars to the states with no strings attached. A strong message from Leon Panetta indicating that the administration was opposed to these block grants, re-enforced the criticisms of congressional Democrats. To shore up support, the Republicans were obliged to make concessions both to their supporters, including the Governors and conservative Christians, as well as to Democrats such as Congressman Tony Hall. Congressman Hall lobbied his fellow Democrats to oppose block grants for school lunches and child nutrition programs, and held a press conference at which he displayed a ten foot aluminum plate, knife and fork, and argued that children would go hungry if the legislation passed. This eﬀort generated a groundswell of support from angry parents and teachers who successfully lobbied against it. A similar eﬀort to defeat the food stamp block grant was supported by private sector interests such as the agricultural and grocers lobbies. There were neither middle class nor private sector interests who could be called upon to lobby against the cash welfare block grant, however. Congressional Democrats were, however, not in a very strong position. It was diﬃcult to defend a welfare program that almost everyone believed as seriously ﬂawed. At a time when budget deﬁcits were still high, any bill that increased funding for welfare was clearly not politically feasible, and cutting welfare beneﬁts would alienate some major Democratic constituencies. Moreover, practically speaking the Republicans in the House were operating largely behind closed doors and proposed Democratic reform bills, such as the Deal bill, were never seriously considered by the

142

The political economy of welfare reform in the United States

Republican majority who clearly stood to gain from passing their own legislation. An additional diﬃculty for the Democrats was the ambiguous position of some of their potential allies, including Democratic Governors and the White House. The NGA, which included both Democratic and Republican Governors, was supportive of the Republican legislation, with the result that a Member opposing the legislation might ﬁnd himself opposed to the Governor of his state. Finally, it was unclear what position would be taken by the White House, which had sacriﬁced its own welfare reform proposal to the budget deﬁcit. While the Republican Party suﬀered setbacks resulting from public disillusionment with Newt Gingrich and the shutting down of the Government in 1995 due to failure to pass appropriations bills, the Republican leaders were able to maintain pressure for passage of welfare reform legislation. By the time the bill arrived on the desk of the President virtually the only active opponents were the poverty groups (discussed further below) and some Democratic Congressmen. When the bill was brought to the Senate it faced opposition from Senator Moynihan and a group of Senators who were adamantly opposed to any bill that ended entitlements (Smith, 1998, p. 177). Not all Democratic Senators were opposed, however. Further, many Democratic governors believed that the political impetus for a welfare bill was so great that it was inevitable that such legislation would pass and the best strategy would be to attempt to inﬂuence the process as much as possible (Ibid.). Governors Chiles, Romer, Carper and Dean worked with the Senate Democrats to educate them on the technicalities of the programs and advise on legislation. The battle over entitlements was thus replaced in the Senate by a battle over funding, and speciﬁcally the formula that Governors Engler and Thompson had used to calculate the block grants. They chose as the basis of calculation an average of the amounts states spent on AFDC between 1992 and 1994, or the amount spent in 1994. Only Michigan and Wisconsin would gain from the ﬁrst formula. Even given a choice, other states would receive a cut in beneﬁts while Engler and Thompson’s states received an increase. Governor Chiles, who had served as chair of the Budget Committee while serving in the Senate was well aware of the implications and came to the attack, focusing on the implications for states with high population growths, and thus enlisting their support. This tack proved eﬀective in the Senate because the Governors could talk states rather than districts, and the audience was 100 rather than 435. After a great deal of acrimonious debate, Senator Dole announced a revised funding formula on July 31, 1995 (Congressional Quarterly, 1995b, p. 2371). In the meantime slowing the process had opened the door to other interest groups and Dole, on reintroducing the bill in September had added looser immigrant

Institutional analysis

143

requirements, exemptions from work for mothers of children under one, maintenance of eﬀort regulations for the states and an additional contingency fund for the states. In mid September, an additional $3 billion for child care was added and $1 billion state contingency funds to ‘break a deadlock and ensure the support of moderate Republicans and some Democrats’ (NGA, cited in Smith, 1998, p. 183). The bill which was passed by the Senate diﬀered in some notable ways from the House-passed bill. The House bill, reﬂecting the inﬂuence of Robert Rector and the religious right, required the states to implement family caps and deny federal funding to unwed teen mothers. The Senate bill allowed states to opt out of these provisions. The Senate bill also, reﬂecting the concerns of those Democrats who feared there would be a rush to the bottom in terms of cutting beneﬁts, required the states to maintain state welfare expenditures of at least 80 percent of 1994 levels, while the House bill had no such provision. The Senate had fewer cuts for immigrants as a result of the inﬂuence of Governor Chiles as described above. Finally, the Senate bill had work participation rates of 25 percent in 1996 and 50 percent in 2000, while the House had only required 10 percent in 1996 and 50 percent by 2003. Both bills had a ﬁve year lifetime limit on beneﬁts and a two year limit on beneﬁts without work. After some compromises in the conference committee, the welfare bill, which was initially attached to a budget bill, was vetoed by Clinton. A stand alone bill was subsequently passed which allowed states to choose 1994, 1995 or an average of 1992–1995 to calculate their grants, allowed states the option of family caps and (through legislation) denying beneﬁts to unwed teen mothers, allocated $800 million for states with high population growth and 1.7 billion for contingency funds, set a participation standard of 50 percent by 2000 and maintenance of eﬀort at 75 percent of 1994 levels. This bill too was vetoed by Clinton on January 9, 1996. At the NGA meeting in February the Governors took up the issue of welfare reform again and Democratic Governor Carper and Republican Thompson agreed to work together to craft a compromise bill. The proposal that emerged contained an additional $4 billion for child care and an additional $1 billion for the contingency fund, made the family cap a state option, and allowed the states to exempt up to 20 percent of their caseload from the ﬁve year requirement as a hardship limit. This bill, together with a Medicaid policy, was unanimously adopted by the Governors. Although many Governors were unhappy with the welfare bill, the need to pass a Medicaid bill which would allow them to control Medicaid expenditures was of paramount importance to them. When brought to the House the proposal was attacked by some conservative interests represented by Robert Rector of the Heritage Foundation

144

The political economy of welfare reform in the United States

and subsequently some changes were made including adding $50 million per year for abstinence education programs, and reducing beneﬁts by 25 percent for recipients not cooperating in establishing paternity. Most of what the Governors proposed was accepted, however, and the House passed the revised bill. The Senate passed bill had few diﬀerences, the major one being the required family cap unless states legislated to not implement it, but this was again taken out in conference under the Byrd Rule. When the Republican leaders in the House and Senate decided to split the Medicaid and Welfare reforms, in spite of opposition from Republican governors, the way was open for a Clinton signature, which was given on August 22, 1996. Governor Engler, although invited, did not attend the signing ceremony at the White House.

THE NATION’S GOVERNORS The Governors were in an extremely strong position and played a key role in the welfare reform legislative process. They derived their power from several sources. In the ﬁrst place, they had the advantage of experience. Under the multitude of waivers granted to the states to experiment with alternative welfare regulations, the states had been implementing a wide variety of welfare reforms since the time of the Nixon administration. Many of these waivers had provisions such as time limits and work requirements that bore similarities to the proposed welfare legislation. In implementing these waivers many states had gained substantial expertise and had increased their capacity for implementing the provisions of welfare reform. Moreover, a number of states had done so within the limitations of balancing their budgets. Republican leaders in Congress would be relying on these Governors both for political support of the law and eﬀective implementation of it since, as the 1988 welfare legislation had shown, laws on the books meant little if the Governors were not enrolled in the process. Of even greater signiﬁcance was the ability of the Governors to help craft the legislation. They were able to provide a level of expertise on the intricacies of the legislation and implementation of the programs which was lacking for most Members of Congress and their staﬀs. The Republicans wished to push the legislation through as quickly as possible, and to do so were dependent on the expertise of the Governors. Finally, the Republican ideology of devolution required Governors to take a greater role in policy implementation. The landslide victory of Republicans in the Congress was mirrored somewhat in the states, where a majority of Republican Governors were elected, clearly making devolution

Institutional analysis

145

a particularly attractive prospect at this time. The Republicans now had the opportunity to shape a Republican welfare reform plan which would be implemented largely by Republican Governors, and which would deﬁne welfare policy for years to come (to what extent this also served the Republican agenda by transferring power away from the bureaucracies, such as the Department of Health and Human Services, and particularly a Democratically controlled bureaucracy, is not clear). The role of the Republican Governors in inﬂuencing the ﬁnal outcome of the welfare legislation can hardly be overestimated. At the Republican Governors’ Annual Conference in November 1994 (two and a half weeks after the election) a group of House and Senate leaders, including Bob Dole, Newt Gingrich and budget committee chair John Kasich, came to meet with the Governors and made a commitment to work more closely with them on crafting and implementing budget cuts.1 The two sides had a common enemy – the Administration, particularly the Department of Health and Social Services. The Governors and Congressmen agreed to reconvene in January 1995. Central to the collaboration on the welfare legislation were Governor Thompson of Wisconsin, Governor Engler of Michigan, and Governor Leavitt of Utah. These three Governors negotiated out a lot of details for the states, and the director of the Social Services Department in Wisconsin, Jerry Miller, was responsible for drafting signiﬁcant parts of the proposed legislation.2 Governor Thompson had implemented a number of welfare reform provisions in Wisconsin under federal waivers and was well positioned to know what worked and to negotiate terms beneﬁcial for the states. On January 6, 1995, a key group of Republican governors came to Washington, DC, and presented their proposed legislation to House and Senate Republican leaders. A major element of the proposed legislation was an agreement by the states to accept ﬁxed funding for ﬁve years in return for ﬂexibility in implementing their programs and eliminating the requirement to have HHS authorize any changes they wished to make in their welfare programs. The governors proposed instituting seven or eight block grants with no strings attached. The only requirement would be certain objectives fulﬁlled: the Governors would have total discretion over implementation. Probably the most important outcome of this meeting was the directive from House and Senate leadership to committee chairs instructing them to have the Governor’s staﬀ sit side by side with them in drafting the welfare legislation. This power conferred to the Governors reﬂected the conﬁdence felt by Newt Gingrich, Clay Shaw, Senator Packwood and others in the ability of the states to implement reform in line with Republican ideals.

146

The political economy of welfare reform in the United States

Working steadily in the Spring of 1995, Governors Thompson, Engler, and others crafted legislation incorporating the end of entitlements and the introduction of no strings block grants. While they had the support of House and Senate Republican leaders, however, they did not have the support of another inﬂuential group of Republicans. The conservative wing of the party, and inﬂuential conservative groups felt that the Governors were too liberal and that certain mandates should be included in the legislation. Thus Robert Rector of the Heritage foundation wanted illegitimacy provisions. Congressman Talent wanted to tighten work requirements and limit job search time. John Ashcroft wanted drug testing. All of these were opposed by governors, but compromises had to be made, and some were dropped and some left as a state option. A further area of conﬂict related to the number of block grants, and this conﬂict was among the Governors themselves as much as between them and other groups. In the January 1996 NGA annual winter meetings there were reportedly more closed door sessions than ever before. The Association ﬁnally came up with a proposal that linked Medicaid and Welfare Reform, reﬂecting the importance they attached to Medicaid, which was growing faster than welfare, took about 20 percent of many state budgets and was riddled with federal requirements. Strong opposition from health interest groups made passage of a block grant for Medicaid highly improbable, however. With Clinton and many Republicans running for reelection, a political decision was made in July 1996 to decouple Medicaid from the welfare legislation and block grant only the latter. One area where Governors were notably unsuccessful was the deposition of unspent funds. These funds, known as rainy day funds, are deposited in the Federal Treasury, rather than in the states. This issue was to become the subject of considerable animosity when Congress, seeing large unspent sums sitting around, proposed to take back a certain percentage of these funds. Since states are required under the legislation to use their own funds before drawing down federal monies, and since they are also obliged to spend 75 percent to 80 percent of their historic funding on welfare programs, it is diﬃcult for some states to eﬃciently allocate these large federal sums. The threat of Congress to recoup the funds clearly sets up an incentive for the states to ensure that they are spent, and thus to some extent confounds the legislative goal of reducing tax dollars spent on welfare programs.

THE NATIONAL GOVERNORS ASSOCIATION (NGA) The National Governors Association traces its roots to 1908 when President Theodore Roosevelt invited the nation’s governors to the White

Institutional analysis

147

House to ask them to lobby their state legislators to support a bill that had stalled in Congress (Smith, 1998, p. 113). While the organization originally focused primarily on intrastate activities, it started to include federal issues in the 1930s. It was not until 1975, however, that it established an oﬃce in Washington DC. It currently devotes about a third of its resources to federal issues. About 35 states have separate state oﬃces in Washington, DC, and the larger states work outside the aegis of the NGA in lobbying activities for their states. The NGA is composed of an Executive committee and three standing committees: Human Resources, Natural Resources and Economic Development and Commerce. Any policy proposal is initially considered by these committees which must approve them by a two-thirds vote. They are then voted on by all the Governors, and if approved by two thirds adopted as oﬃcial NGA policy. Policies automatically expire after two years (Ibid., p. 111). While the Governors were wooed when they had something to sell, after it was acquired they were to a large extent discarded. Though actively involved in the crafting of the legislation, two of the NGA’s main goals – block grants for Medicaid and maintenance of direct control over unspent TANF funds – were not achieved. Moreover, to appease the Christian conservatives, several mandates on such issues as illegitimacy were incorporated into the welfare legislation. And when in 1997 the Republican governors asked to meet regularly with congressional leaders Senate majority leader Trent Lott reportedly told them that ‘he hates to have meetings just to have meetings’ (Ibid., p. 125).

THE AMERICAN PUBLIC HUMAN SERVICES ASSOCIATION (APHSA) The American Public Human Services Association (APHSA), founded in 1930, is a bipartisan, nonproﬁt organization, representing all of the state human services departments as well as local agencies and individuals concerned with social welfare policies. According to the 1998 Annual Report of the Association, its mission is to develop, promote, and implement public human service policies that improve the health and well-being of families, children and adults. The bipartisan nature of this group, and the central role it would be required to play in implementation of welfare reform virtually assured this group some voice in the proceedings. It did not turn out to be a very eﬀective voice, however. APHSA managed to keep a foot in both camps of the Congress, passing a resolution in 1994 that laid out changes they felt

148

The political economy of welfare reform in the United States

would be eﬀective if the entitlement status of welfare was maintained and those that would be required if the entitlement status was ended. The organization did not endorse welfare legislation, and, in fact, opposed lifetime time limits and many of the restrictions on eligibility for food stamps and SSI, including most restrictions for non-citizens. The involvement of APHSA in inﬂuencing the ﬁnal legislation was facilitated in large part by its close connections with state Governors. The president of APHSA, Gerald Miller, was also the Commissioner of Human Services for Michigan Governor Engler, one of the two or three leading Republican Governors responsible for crafting a great deal of the ﬁnal welfare legislation. Drafts of legislation drawn up by the Human Services Subcommittee of the Ways and Means Committee could be sent out through APHSA to the Democratic and Republican members of APHSA in the states. These members could then relay feedback to their members, further tightening the bonds between the Congress and the state governments.

CHAMBER OF COMMERCE The US Chamber of Commerce represents 215,000 businesses in the United States, 3,000 state and local chambers of commerce, 1,200 trade and professional associations and 72 American Chambers of Commerce overseas (Josten, 1995). It was rated by Fortune magazine as the sixth most politically inﬂuential organization in the US in 1999 and, according to a CNN report, pledged to spend ‘at least $5 million to (elect pro-business candidates to) Congress’ in the 2000 elections (Cable News Network (CNN), 1999). Clearly, it has weighty political clout. Holding round table discussions with members and their staﬀs, inviting Members of Congress to speak at Chamber events and granting awards to members who have been supportive of business interests, the organization is a very visible presence on Capitol Hill and its leaders are on a ﬁrst name basis with a number of inﬂuential Republican congressmen.3 Moreover the Chamber publication, ‘How They Voted’ (1997), which evaluates the support or opposition of each senator and representative by tabulating ﬂoor votes on issues important to the business community puts members on notice that their every vote will be published and distributed widely to business and other interests. The Chamber was a strong a proponent of welfare reform. In an August 1995 letter to members of the Senate it noted that a survey of its members revealed that welfare reform was the second highest priority (after unfunded mandates) on a list of 64 issues (it did not state the sample size or what percentage of those receiving the survey were suﬃciently interested in the issue to respond).

Institutional analysis

149

The major reason was employment. Since the early 1990s businesses had been reporting to the Chamber that the shortage of labor, not capital or technology, was a major problem. The Chamber believed that a reformed welfare system was ‘a transitional system leading to work’ (US Chamber of Commerce, 1997, p. 8). Clearly, welfare reform could potentially result in the formation of a vast pool of cheap laborers who would be compelled to work or face destitution. Businesses had a clear stake in the imposition of strict time limits and work requirements, and wanted to be involved in the legislative process. In a January 1995 letter to key House and Senate committees, Bruce Josten, a Senior Vice President of the Membership Policy Group of the Chamber, noted that the Chamber was ‘anxious to work with members of Congress and to lead the ﬁght for business to reform welfare in 1995’ and expressed the hope that Congress would call on Chamber oﬃcials ‘to assist (their) eﬀorts by way of testimony, brieﬁngs and grassroots support’. In a further letter to members of the Senate in August 1995, Josten expressed the belief that it was ‘essential that business be involved in the design, development and operation of any changes in America’s welfare system’. This correspondence also noted the Chamber’s strong support for time limits and work requirements. Time limits and work requirements were, predictably, widely supported by businesses. In a June 1995 poll the Chamber reported that of over 6,000 respondents, 75 percent felt that the ﬁve year lifetime limit on beneﬁts was too long and that welfare recipients should be forced to work within a shorter period. A requirement that able-bodied food stamp recipients between the ages of 18 and 50 be required to work or be in training within 90 days or lose beneﬁts was supported by 69 percent of respondents, while 29 percent said it was not tough enough. In a similar vein, denying beneﬁts to legal immigrants was generally supported and providing tax credits for businesses that hired welfare recipients were strongly supported. These last two issues were reﬂected as priorities in the Chamber’s 1999–2000 National Business Agenda which supported legislation increasing the number of employer sponsored visas for both high tech and low skilled immigrants, and advocated extension of business tax credits such as the Work Opportunity Tax Credit, available to businesses that hire members of targeted groups including welfare recipients (US Chamber of Commerce, 1999). The Agenda did not specify why, given the extreme shortage of labor in the US, it would be necessary to give businesses tax breaks for hiring available employees. The impetus of business interests to increase the labor supply was perhaps most clearly summed up by Richard Lesher, President of the US Chamber of Commerce, in a March 1995 article entitled, ‘The Welfare

150

The political economy of welfare reform in the United States

Trap’ (Lesher, 1995). Lesher stressed the need to move people oﬀ welfare and into the work force, help business fund on-the-job learning by providing ‘a window of ﬂexibility in which the usual litany of employment regulations and potential liability is lifted’, and have a ‘hardnosed commitment to impose a deadly serious cutoﬀ for beneﬁts, without appeal or bureaucratic runaround’. He concluded that this was a diﬃcult thing to do ‘but we must – if we really care about people’ (Lesher, 1995).

PUBLIC SECTOR WELFARE SERVICE PROVIDERS The American Federation of State, County and Municipal Employees (AFSCME), like the NGA and APHSA, represents state interests, but with one crucial diﬀerence; it represents the employees of the NGA and APHSA, which gives it a quite diﬀerent perspective and very diﬀerent issues of concern. In the discussions and negotiations which preceded passage of the new welfare legislation AFSCME was aligned more with the poverty groups than with the state interests, worked in coalition with the various poverty groups such as the Children’s Defense Fund (CDF) and Catholic Charities, and had monthly meetings throughout 1994 and 1995 with the Coalition on Human Needs. This organization shared with the poverty groups an opposition to block grants and to time limits and other aspects of the law which they felt were too punitive, but one of their greatest concerns was that welfare to work legislation would result in displacement of the employees they represented as states tried to ﬁnd jobs for welfare recipients. Under workfare regulations, a welfare client can be put to work with a private or public employer and works oﬀ their welfare beneﬁt, rather than getting wages. Clearly, there are pecuniary advantages to an employer who replaces wage workers with welfare recipients. While it is diﬃcult to prove that this has in fact occurred, AFSCME has several cases under investigation and currently has litigation pending against the State of New York for engaging in unfair labor practices. A further concern was that states would not remunerate welfare recipients at the minimum wage, as Department of Labor guidelines stipulated. This concern proved well founded when Governor Wilson refused to enforce these guidelines in California. Under the new welfare law, private contractors can be hired not only to provide services such as child care programs, but also to actually administer certain aspects of the law. The legislation has provided a bonanza to several contractors, most notable among whom are Lockheed Martin Corporation, more widely known as a defense contractor, and Maximus, a ﬁrm which has been granted substantial contracts to carry out evaluations

Institutional analysis

151

of the new welfare law. Firms such as Lockheed Martin can take advantage of the subsidized labor of welfare recipients in their own company, as well as the $2,000 fee they are paid to place each recipient in a job. Established in 1975 by Dr David Mastran, Maximus has over 3200 employees providing management services to Health and Human Services Departments of State and local governments. The company went public in 1997 and is listed on the New York Stock Exchange. After focusing mainly on child support enforcement programs in its early years, in recent years the company has broadly diversiﬁed. It now oﬀers technical assistance in a range of areas including managed care enrollment, waiver applications, welfare to work programs, referral and monitoring of drug addicts and alcoholics receiving SSI and Disability Insurance (DI), and information technology services for federal, state and local governments. In addition to program assistance, Maximus provides revenue maximization services to state and local governments, and notes among its achievements that it has ‘helped our clients to obtain nearly $1 billion in federal funding for systems eﬀorts’ (Company brochure). Moreover, according to their Annual Report: ‘the states have received more than $350 million in additional federal revenue as a result of the Company’s eﬀorts and expect current projects to yield another $300 million in new federal revenue’ (Maximus, 1998, p. 8). Clearly the relationship is mutually beneﬁcial to both the Company and the states. As outsourcing of government functions has proliferated, Maximus has ﬂourished. Its growth in the past ﬁve years has been remarkable. In FY1994 the company had revenues of a little over $50 million, a ﬁgure which increased 600 percent by FY1999, when revenues exceeded $300 million. During the same period net income increased from $2.1 million to $14.4 million. Thus, while it took the company 23 years to attain a level of proﬁtability of $2 million, in the following ﬁve years it increased proﬁts sevenfold to $14.4 million. In addition to ballooning proﬁts, Maximus made several strategic acquisitions in 1998. These included David M. Griﬃth and Associates, which provides consulting services in cost accounting, wage and compensation evaluation, and executive recruiting, Spectrum consulting, which focuses on system planning, Carrera consulting, a software company, and Phoenix Planning and Evaluation, specializing in electronic commerce. These acquisitions increased the number of the Company’s professional consultants from approximately 120 to over 600. The advantages to Maximus include not only more diversiﬁed and experienced consultants, but also ‘valuable relationships with members of the executive and legislative branches of state and local governments’ (Maximus, 1999, p. 3).

152

The political economy of welfare reform in the United States

Relying entirely on government contracts for its revenue, Maximus places great emphasis on establishing relationships with past and present local and state government oﬃcials. The 1999 Annual Report notes that: ‘since state and local government administrators are subject to changing legislative and political mandates, the company has developed strong relationships with experienced political consultants who inform and advise the company with respect to strategic marketing opportunities and legislative initiatives’ (Ibid., p. 4). Its growth strategy includes recruiting top government management professionals and middle level consultants with a ‘network of political contacts . . . to leverage the company’s . . . client relationships’ (Ibid., p. 4). Focusing on government contracts shows promise of being an increasingly lucrative endeavor. The 1998 Annual Report states that: The Company believes that providing program management and consulting services to government agencies represents a signiﬁcant market opportunity. Federal state and local government agencies in the United States spend more than $250 billion annually on the health and human services programs to which the company markets its services, including Medicaid, Food Stamps, Temporary Assistance to Needy Families, Child Support Enforcement, Supplemental Security Income, General Assistance, Child Care and Child Welfare. The state operated programs alone cost an estimated $21.0 billion annually to administer (Maximus, 1998, p. 1).

In addition to these programs, the report notes other areas of opportunity, including the Balanced Budget Act of 1997 which established new programs, including the Children’s Health Initiative Program (CHIP) which provided $20 billion in federal matching grants to states, and in June 1998, the Clinton administration mandate of extended eligibility for Medicare. The expansion of programs in times of budget surpluses means more opportunities for private companies such as Maximus. The corollary does not hold, however. In bad economic times the number of beneﬁciaries of such programs grows which also means increasing business for the company. As the report notes, ‘the Company believes that state and local governments will continue to seek its services despite the eﬀects of economic cycles on government budgets’ (Maximus, 1998, p. 2). In order to capitalize on these opportunities, the company takes a proactive approach. While it follows the traditional approach of obtaining contracts by responding to government requests for proposals (RFP), whenever possible, prior to the issuance of an RFP, it dispatches senior executives ‘to work with senior government representatives such as a state’s governor, members of the governor’s staﬀ, and the heads of health and human services agencies to encourage them to outsource certain health and human service functions’ (Ibid., p. 9). Moreover:

Institutional analysis

153

to identify opportunities to work with government oﬃcials at early stages and to optimize the government’s receptivity to the company’s proposal to provide program management services, the company establishes and maintains relationships with elected oﬃcials, political appointees and government employees. The company engages market consultants, including lobbyists to establish and maintain relationships with these client representatives. The company’s consultants and lobbyists provide introductions to government personnel and provide information to the company regarding the status of legislative and executive decision making (Ibid.).

Clearly, the modus operandi of interaction between the private sector and the public sector diﬀers little whether the issue is utility regulation or welfare recipients. The Children’s Defense Fund (CDF) is a private nonproﬁt organization established in 1973 by Marion Wright Edelman. The stated goal of the organization is ‘to provide a strong and eﬀective voice for all the children of America, who cannot vote, lobby, or speak out for themselves’ (Children’s Defense Fund, 1998, p. 1). Supported by foundation and Corporate grants, and individual donations, the Fund has assets of about $50 million. It does not accept government funds. CDF undertakes research and analysis on issues relating to children’s welfare and engages in advocacy and program development at the state, local and national level. The CDF was strongly opposed to the welfare reform proposals of both the Clinton administration and the Republican-led Congress. They engaged in advocacy at all levels of government, from direct lobbying of administration oﬃcials to grassroots eﬀorts encouraging their state and local aﬃliates, including local service providers and church groups, to lobby their Congressional Representatives. The success of their eﬀorts was limited by a number of factors. Internally, many of the organization’s resources were devoted to the battle against the Balanced Budget Constitutional Amendment which was seen as a larger threat to the goals of CDF than was welfare reform. With a limited budget and limited staﬀ time, welfare concerns became secondary. Externally, opportunities for providing input to the debate were severely restricted. According to a CDF staﬀ member, the opportunity for minority members of a committee to select witnesses to give testimony at hearings was sharply curtailed.4 (The observation that members of liberal interest groups were generally given little or no voice by the Republican-led Congress, was conﬁrmed by a representative of one of the larger groups involved in the process, who noted that requests to Republican staﬀers for meetings were routinely disregarded).5 And, as mentioned earlier, the Democrats were excluded from many of the welfare reform negotiations. The frustration felt by the liberal organizations such as CDF, the Institute

154

The political economy of welfare reform in the United States

for Research on Poverty (IRP), and other such groups was reﬂected in a statement by a former director of the IRP who complained that this Congress was immune to research. He cited an analysis of 76 leading researchers, both liberal and conservative, showing that welfare programs were not among the primary reasons for rising out-of-wedlock births as many Republicans asserted (CBPP, 1995). The contrary view was espoused by conservative organizations such as the Heritage Foundation, whose director, Robert Rector, worked closely with Congressional Republicans. The mission statement of Network describes the institution as a Catholic Social Justice lobby whose goal is to educate, lobby and organize to inﬂuence federal legislation to promote economic and social justice. Founded in 1971, the organization has a membership of about 10,000, and an annual operating budget of about $800,000. Like other ideologically based groups, Network, while not opposed to welfare reform itself was strongly opposed to ending the entitlement status of AFDC. This group engaged in extensive grassroots campaigning as well as collaborating with other groups on media campaigns and lobbying the administration and the Congress. A major constraint on their eﬀorts, according to oﬃcials of the organization, was restricted access of lobbyists and an inability to plead their case. Other church-based organizations were more successful. The politically weighty Catholic Charities USA presented testimony to several House and Senate committees in 1995 and 1996. According to the testimony of its president, Rev. Fred Kammer, S.J. before the Senate Labor Committee in March 1996, Catholic Charities is a national association of 1,400 independent local Catholic Charities agencies with a combined budget of $1.9 billion in 1994. The agencies comprise 234,000 staﬀ members and volunteers, and in 1994 served 11 million people of diverse religious and ethnic backgrounds. Services provided range from homeless shelters to adoption and psychological counseling. Although its voice was heard, however, the organization had limited success in achieving its goals. The major goal of maintaining AFDC as an entitlement was not obtained. It was more successful in removing the obligatory requirement that states implement a family cap and deny payments to unwed mothers under 18. These provisions were left as state options in the ﬁnal legislation. The success in these areas was in part attributable to collaboration with conservative groups such as the national Right to Life organization and some Congressional Republicans who felt that cutting oﬀ beneﬁts for children born to unwed teen mothers or mothers already on welfare would lead to increased abortions. State governors also objected to the provisions being obligatory and fought for them to be eliminated or left as a state option.

Institutional analysis

155

Two other areas in which Catholic Charities was successful – dropping the block grant for food stamps and Medicaid – were also achieved in collaboration with other groups. The food industry, as well as anti-hunger groups such as Bread for the World, opposed block granting food programs, fearing that such funds would go into general state funds and not necessarily be used for food. There was also much stronger popular support for food programs, especially for children, and members such as Democrat Tony Hall took a very strong stand against these block grants. Medical providers had similar fears that block grants would go into general state coﬀers rather than being dedicated to the purchase of the goods and services they provided, and such institutions lobbied strongly against the Medicaid block grant. Thus, although Catholic Charities, with its huge budget and its position as representative of a large Catholic vote, had access to inﬂuential Republican lawmakers, it is not at all clear what, if any, inﬂuence on the ﬁnal outcome this organization would have been able to achieve without the collaboration of other powerful pecuniary and ideological interests. Founded in 1910, the National Urban League (NUL) is a social service and civil rights organization whose stated mission is to help African Americans attain social and economic equality. With an annual budget of about $35 million, and headquarters in New York, the NUL operates at the national, state and local levels, engaging in advocacy, research, policy analysis, and provision of social services. The NUL has aﬃliates in 115 cities in 34 states and the District of Columbia. While this organization was extremely active in the legislative activities that preceded the 1988 Family Support Act, working with a Democratic Congress, like other groups with similar goals, it was largely sidelined in the proceedings that led to the 1996 welfare reform. Giving Congressional testimony only once prior to the 1996 legislation, the NUL was rarely called upon to advise the Republican members and their staﬀs that dominated the legislative process. Several of the liberal organizations opposing the welfare reform legislation, and particularly the ending of entitlement status, expressed frustration that their ability to inﬂuence the process was severely constrained not only by a tightly controlled Republican Congress, but also by the fact that both the Democratic Party and the President were moving toward the center. These groups had paid little attention to the Republicans when the Congress was controlled by the Democrats and were thus easily relegated to non-players by the Republicans. This miscalculation on the part of the poverty groups was compounded by their miscalculation of the possibility that Clinton would actually sign the welfare reform law, although this latter was an easier mistake to make, given his apparent indecisiveness, even up to the last minute.

156

The political economy of welfare reform in the United States

At the other end of the ideological spectrum from the anti-poverty groups are the conservative groups such as the Heritage Foundation, the Christian Coalition, the Family Research Council, the Coalition for Traditional Values and Empower America. Led primarily by Robert Rector, a senior policy analyst at the Heritage Foundation, a conservative Washington think tank, these groups were principally interested in raising the issue of moral values, and inﬂuencing the legislative process to produce a bill that would reﬂect their values and what they believed should be the values of others. The Christian Coalition (CC) was founded in 1989 by Pat Robertson. Ordained a Baptist minister in 1960, Robertson started a television station in Norfolk, Virginia in 1961 and subsequently built a business empire that includes the Christian Broadcasting Network (CBN), a commercial cable television channel, a relief agency, Regent University, and other business, ministry and educational ventures. The objective of CC is to make government more responsive to: ‘the concerns of evangelical Christians and pro-family Catholics by activities such as providing information on pending legislation, providing training on social and political action, lobbying local state and national leaders and speaking out through the media’ (Watson, 1997, pp. 52–3). Robertson gained national prominence in 1988 when he sought the Republican nomination for the presidency. His ambitions were broad. A campaign oﬃcial claimed that: this campaign is not a one-shot attempt to win one oﬃce, though that is the focal point of our eﬀorts. It is designed to start a permanent restructuring of American politics, especially Republican politics. There are a lot of folks out there who want to reclaim control over their lives and their government. And we’re determined to help them succeed (Ibid., p. 42).

Although Robertson lost, his campaign served to establish a foothold for the religious right in state party committees throughout the country and marked an increasing degree of political sophistication in the movement. Robertson used his campaign’s mailing list of donors and activists to solicit help in starting up the CC which was designed to be a new grassroots political organization. He hired Ralph Reed, who had worked on the Reagan presidential campaign, and the Jesse Helms reelection campaign, to be executive director of the CC. Reed was able to obtain a donation of $64,000 from the Republican Senatorial Committee as seed money for the new organization (Isikoﬀ, 1992, A14). The CC has headquarters in Virginia with a network of state aﬃliates and local chapters. It grew from a membership of 25,000 in 1990 to an estimated 1.9 million when Reed resigned in 1997 (Reed, 1997). The organization uses

Institutional analysis

157

its three publications, television channel and Internet site to get its message out, inform members of current issues of importance to the pro family movement, and oﬀer guidance on what local members can do to move public policy. Guests on the television program have included Bob Dole, Newt Gingrich and William Bennett. In the ﬁrst six months of 1996, a crucial time for welfare reform, the CC reportedly spent $5.9 million on lobbying activities (Watson, 1997, p. 66). At the grass roots level, it holds seminars on political activism, including voter registration drives in member churches, establishing telephone trees and lobbying of legislators and distributes voter guides. The voter guides compare candidates positions on selected issues and are distributed on the last Sunday before an election, giving candidates little time to refute any issue. CC claimed it distributed 46.3 million voter guides in more than 100,000 churches nationwide just prior to the 1996 election. Reed supported the Republican Contract With America, although it did not address some of the major issues of concern to CC. He explained that his objective was ‘to win a big victory on the Contract With America, thereby building up political capital that we (can) later spend on social issues’ (Ibid., p. 73). On May 17, 1995 the CC published their own issues agenda ‘Contract with the American Family’, which was developed through focus groups and polls, even hiring the same pollster, Frank Luntz, who worked with Gingrich on the Republican Contract. In comparison with earlier manifestos this avoided many radical positions and Reed was accused of political expediency by other Christian Right leaders. Martin Mawyer of the Christian Action Network complained that the CC was ‘so locked into Republican politics, they are continually forced to redeﬁne themselves based on the current political climate and who’s in charge of the Republican party’ (Ibid.). After the Republican landslide of 1994, Reed was credited with delivering the evangelical vote to the Republicans, and the CC’s annual conference in September 1995 was attended by all the Republican presidential candidates except the pro choice Pete Wilson and Arlan Specter (Wilson, 1989, p. 81). In early 1995 Reed claimed the 1994 elections were important because: they gave people of faith what they have always sought: a place at the table, a sense of legitimacy, and a voice in the conversation that we call democracy. We have become a permanent ﬁxture on the American political landscape, too large, too signiﬁcant and too diverse to be ignored by either major party (Ibid., p. 163).

The honeymoon was short lived, however. After helping Dole to gain the nomination, the CC and other religious conservatives were treated as outsiders at the 1996 Republican National Convention. There was a notable absence of religious conservative speaking from the podium, and although

158

The political economy of welfare reform in the United States

their views were incorporated into the party platform, Dole made it clear he did not feel bound by that document. Despite delivering millions of voter guides, the CC was not able to deliver the election for Dole. In addition to problems with the Republicans, rifts were present between the diﬀerent groups which constituted the religious right. Reed had alienated some of these with his Contract With the American Family, which they felt was a sellout, and he further alienated other Christian Right organizations by not appearing with their representatives at a press conference organized by the Buchanan campaign in May 1996 to denounce attempts to change the abortion plank of the Republican platform. In negotiations with the Congress on the new welfare law, the self appointed leader of the Christian right groups was not one of their leaders but Robert Rector of the Heritage Foundation. Rector had worked on welfare policies for several decades and had long advocated a pro family position, similar to that shared by the majority of the Christian right. He used his close ties to the Republican Congress to insert this agenda into the legislative process. The Heritage Foundation, played a central role in the development of the 1996 welfare legislation. Robert Rector was highly critical of the 1988 legislation which he felt was a smokescreen basically intended to preserve the status quo. In late 1993 Senators Faircloth and Talent approached Rector for advice on crafting welfare reform legislation. This collaboration resulted in the Real Welfare Act of 1994, a piece of legislation that reﬂected Rector’s primary concerns: controlling aggregate welfare spending; ending entitlements and establishing block grant funding; requiring work provisions, and, most importantly to Rector, focusing on illegitimacy as a cause of many social ills. In the Talent–Faircloth legislation work requirements were very strong, reﬂecting, as Rector noted, that such requirements have a 92 percent public approval rating. Although Rector’s most acute concern was illegitimacy, he knew that the public in general was far less overwhelming in their support for illegitimacy provisions in welfare legislation, and strong work requirements were more likely to win popular support.6 A substantial part of the Talent–Faircloth bill was subsequently incorporated into the Contract With America, including the illegitimacy provision, work requirements, limits on aggregate welfare spending and removal of entitlements and block granting of a number of welfare programs. One of the major battles that arose among conservatives pitted Robert Rector against the Governors, and particularly Governor Engler who was strongly opposed to the illegitimacy provisions in the welfare legislation. In the winter of 1994 Rector met with Engler in a Washington hotel room, explaining to him that the family cap and the provision regarding

Institutional analysis

159

unwed teenage mothers did not necessarily limit state control since states could circumvent them by funding beneﬁts for such recipients with state money. From the point of view of Rector and other social conservatives, the main objective was to symbolically raise the issue of illegitimacy in order to generate public attention for what they felt was a phenomenon destroying the society and creating an underclass. Whether or not the States actually implemented the provisions was not of primary concern. Rector believed that legislation would be the vehicle to generate debate about illegitimacy, and was proud of the fact that by December 1994 liberals such as Eleanor Holmes Norton, the Democratic Congressional delegate representing Washington, DC, were expressing public concern about the social eﬀects of the collapse of marriage. He attributes this attention on the part of liberals to his longstanding eﬀorts to publicize the issue as well as the eﬀorts of Charles Murray, who in the fall of 1993 published a very inﬂuential article in the Wall Street Journal on the rapid rise of white illegitimacy and what he perceived as the development of a white underclass (Murray, 1993). In this article, Murray drew attention to the fact that the overall illegitimacy rate in the US in 1991 was about 30 percent, four points higher than the black illegitimacy rate in the early 1960s which had motivated Senator Moynihan to write about the breakdown of the black family. Murray claimed that ‘the new trend that threatens the US is white illegitimacy’, and he referred to the development of a white underclass (Ibid.). By casting illegitimacy as a white as well as a black problem, the taboo felt by some against speaking aloud on the issue was removed. There was no longer the same danger of being cast as racist by talking about an underclass, and the social evils that resulted from illegitimacy. The Murray article was important not only because it facilitated a more open debate, but also because it publicized the issue of illegitimacy, especially white illegitimacy, more widely, and indirectly contributed to the belief that the welfare system not only was not working, but that it may be contributing to the development of the white underclass and accompanying major social upheavals. The issue of illegitimacy was a very divisive one in the Senate. A number of interest groups including the Christian Coalition, Empower America, the Family Research Council, the Traditional Values Coalition, and the Heritage Foundation weighed in strongly and continuously from the summer of 1994 to passage of the ﬁnal bill in August 1996. On the other side were the Governors, including Governor Engler, who were strongly opposed to incorporating illegitimacy provisions in the welfare legislation. Ultimately a compromise was worked out in which the language was included, but there was no requirement for states to pay any attention. The

160

The political economy of welfare reform in the United States

states had the option of giving beneﬁts to children of unwed teens and children born to a mother already on the welfare rolls. The states kept control over their program rules: the anti-illegitimacy interest groups had successfully raised the issue at the highest levels of national debate. A further area of contention between Rector and the Governors was that of work provisions, including the percentage reductions in caseloads that states would be required to make, and the exemptions permitted for mothers of young children or other categories. While the Governors wanted minimal mandates, Senators Faircloth and Talent, as well as Rector wanted very stringent requirements and incentives to force the states to drastically cut the welfare rolls and put recipients to work. In contrast to the Christian Coalition, which could mobilize huge voter turnouts, and the Heritage Foundation, which had become a very inﬂuential Conservative think tank, Empower America was a relatively small and unknown entity. Founded in 1993 by Former Department of Housing and Urban Development (HUD) Secretary Jack Kemp, former Education Secretary William Bennett (who was also a John Olin fellow at the Heritage Foundation), former Congressman Vin Weber, and Ambassador Jeanne Kirkpatrick, this group used its inside inﬂuence to aﬀect the course of welfare legislation. With Malcolm S. Forbes, Jr. as its Chairman, and Congressman Newt Gingrich and Senator Trent Lott on its Board of Directors, access to Congressional decision-makers was not a problem. The founders of Empower America, particularly William Bennett and Jack Kemp, held press conferences, gave testimony at Congressional hearings and wrote press articles to promote their views. A major objective of the group was to throw its weight behind the Talent–Faircloth bill, rather than the more moderate bill cosponsored by Finance Committee chairman Bob Packwood, and supported by other members of the Senate Republican leadership and Senate Democrats. The pressure to move the legislation to the right was founded on both ideological principles and political astuteness. In a memorandum to Congressional Republicans, dated April 13, 1994, Bennett, Kemp and Weber urged House Republicans to ‘fashion a bold, principled, and fundamentally diﬀerent alternative to the current House Republican bill’. They advocated a ‘comprehensive welfare approach’ based on 12 principles, of which the ﬁrst four related to work. Noting that ‘Economic institutions are some of society’s most important mediating structures, creating hope as well as values’, the authors suggested that meaningful reform should include policies to promote job creation with pro-growth, asset-based, entrepreneurial strategies. In addition they advocated incentives to take and hold entry-level jobs, including tax exemptions for low income working families, and work requirements for able-bodied welfare recipients, the nature of

Institutional analysis

161

which should be at the discretion of the state. Other principles proposed in the memorandum ranged from capping overall welfare spending, to easing restrictions on adoption and limiting welfare payment to women who have children out of wedlock. The memorandum went on to note that welfare reform also represented a political opportunity: ‘This is a moment when good public policy is also good politics. Republicans must begin by preventing the President from packaging status quo policies as major reform. This can only be accomplished by sharpening policy diﬀerences, not blurring them with tepid legislative compromises’ (Bennett et al., 1994). The memorandum continues: The President ﬁnds himself in a political bind. He can maintain his liberal political coalition, essential for this health care plan, by proposing liberal welfare policies. If he goes down this path, however, he will (a) betray his campaign promise to ‘end welfare as we know it’, (b) shatter once and for all the myth that Bill Clinton is a ‘new Democrat’, and (c) reveal his true political ideology. Or, the President can risk fracturing his liberal political alliance by adopting a marginally more conservative approach – one that does not go far enough to reform welfare, but one that does go far enough to oﬀend his allies on the left. For a sneak preview one need only consider the angry response from the Children’s Defense Fund in response to the Clinton task force. We do not believe it is the duty of House Republicans to rescue Bill Clinton from problems of his own making.7

Thus, as William Bennett and Peter Wehner noted in a Washington Times article in August 1995, Clinton could sign the Republican legislation, alienating many of his supporters and giving the Republicans a policy victory, or he could veto it, reneging on his promise, ‘perpetuating the status quo’, and giving Republicans ‘a political, if not a policy, victory’ (Bennett and Wehner, 1995). The preceding description of the institutional context within which the welfare legislation was produced sheds light on the validity of the alternative theories with which we have been dealing. Both the legislative process and the legislative outcome of P.L. 104-193 cast doubt on the validity of traditional welfare economic theory. Given the powerful inﬂuence of some of the institutions discussed above, and their integral involvement in the legislative process, it would be diﬃcult indeed to accept that the legislation produced by the processes described in this and the preceding chapter was a product designed to move the economy to a better place, unbiased and untrammelled by the activities of the various individuals and groups involved. This is not to deny, however, that certain outcomes of the legislation may be Pareto superior. An increased number of welfare recipients in the work force, for example, would ceteris paribus tend to make both the

162

The political economy of welfare reform in the United States

former recipient and the economy in general, better oﬀ. The point is, however, whether this is a direct objective and outcome of the legislative process or a side eﬀect resulting from the political market activities of the interests concerned in the process. It is important to note in this context that the issue of poverty, presumably a major disutility to be dealt with by any national welfare policy, was notable for its absence. While anti-poverty and church groups attempted to engage in the process, they were, for the most part excluded (except for those with political clout such as Catholic Social Services) and had minimal voice or eﬀect on the legislation ultimately passed. The degree to which the outcome (as opposed to the process) of the legislation supports the traditional economic welfare theory is discussed more in the following chapter. The alternative public choice theory would predict an outcome to the legislative process that reﬂects the relative bargaining power of the institutions involved, acting within speciﬁc institutional constraints. It would predict a high level of involvement by interest groups that would be motivated by the pursuit of individual self-interest rather than a notion of the common good, such as elimination of poverty. In this context, the political market place is a forum in which legislation is traded and those individuals, groups or coalitions of groups willing and able to pay the highest price are those who are successful in obtaining the product. In the framework outlined above, legislation is an intermediate product that provides diﬀerent ﬁnal goods to diﬀerent agents. These ﬁnal goods we have classiﬁed as political, ideological or pecuniary. It is important to point out that the inclusion of ideology renders this framework diﬀerent from that generally used by public choice theorists, who ascribe pecuniary or political motivations to most transactions in the political market place. In the case of the welfare legislation under consideration, ideology held center stage in the marketplace. We turn now to assess what outcomes would be predicted for the diﬀerent categories of groups by a public choice model. In the marketplace for legislation, the Congress and the President have to negotiate to produce a legislative product that will garner for each of them the highest price. If we assume what is being sought by these two entities is the political objective of reelection, then votes are the ultimate price that must be delivered, either directly from members and aﬃliates of a particular group, or indirectly through the use of funds, publicity or other resources a group can provide. The potential purchasers of the legislation, the pecuniary and the ideological groups, will be more or less successful depending on their ability to deliver probable success at the polls to the primary sellers of the legislation, in the case under discussion the Republican Congress and the Democratic President. Whether the potential buyer of the legislation seeks pecuniary

Institutional analysis

163

or ideological goals is irrelevant; what is traded is legislation in exchange for political support. On the supply side, it is clear that the end product will be the result of trade-oﬀs between the legislative and executive, particularly where the two are from opposing parties. A Democratic President can shore up support from his core constituents by providing them with legislation supported by Democratic interest groups and Democratic voters at large. To the extent that he can attract Republican support without losing this base he can maximize his support by moving to the right. The corollary is true for the Republican Congress. In this we agree with the basic tenet of the median voter theorem. An important variable not addressed by this theorem, though, is the importance of the initial political strength of the political entity. Thus, where the entity is initially in a very strong position, as was the case with the Republican Congress in 1994, there is little necessity to trade. With a strong popular mandate, extra votes were of little marginal value and Congressional Republicans could reward their supporters with little need to make concessions to fringe supporters. President Clinton had far less security and thus more incentives to compromise in hopes of regaining some centrist voters. Given this balance of power, in which Congressional Republicans had a good deal of monopsony power, a legislative product more amenable to conservative Republicans than liberal Democrats is precisely what public choice theory would predict, and precisely what, in fact, materialized. On the demand side of the market, public choice theory would predict that those groups able to pay the highest price in political support, either through the strength of their voting membership, the size of their Political Action Committees (PAC) or their inﬂuence through other channels, would be most successful in purchasing the legislation they sought. Among the potential buyers of welfare legislation were two opposing sets of ideological groups: the liberal pro-entitlement groups such as Network and the Children’s Defense Fund, and the religious right and aﬃliated groups including the Christian Coalition and the Heritage Foundation. The strong inﬂuence of the Christian Coalition and other conservative groups on the 1994 election clearly showed their ability to bring out the vote and provided evidence of their strength in the political marketplace. Not only were these groups able to deliver large numbers of votes through their grassroots networks, they also had signiﬁcant ﬁnancial resources to oﬀer. Moreover, the religious based groups such as the Christian Coalition and Empower America worked closely with the Heritage Foundation, which in turn has very close links with the Chamber of Commerce and the business community in general and receives substantial funding from businesses. The power of the conservative groups is reﬂected in the comment by a

164

The political economy of welfare reform in the United States

senior Republic Congressional aide that Robert Rector was ‘the single most important outside person (involved in crafting the legislation), including the Governors’ (Smith, 1998, p. 172). While the conservative groups had both votes and ﬁnances to oﬀer, the liberal groups were not so fortunate. Public opinion had generally become more conservative and there was less support for the positions taken by liberal groups such as the Children’s Defense Fund, Network and Catholic Charities. Moreover, many of these groups refused on principal to accept government funds, but had far fewer aﬃliations with the business community, thus rendering them at a disadvantage vis-à-vis the conservative groups both in votes and in funds. While well-funded organizations such as Catholic Charities were also in this group, the very diverse political viewpoints of Catholics meant that the organization could not be relied upon to deliver a vote with the same consistency as, say, the Christian Coalition. While the ideological groups competed against each other to inﬂuence the legislation in favor of their ideological viewpoint, the pecuniary groups were able to take a diﬀerent approach. Whereas the passage of legislation favorable to the religious right, such as cutting oﬀ beneﬁts for children of unwed teen mothers, was, of necessity, a defeat for liberal groups, the various pecuniary groups could all be accommodated by simply increasing the size of the pie. This was becoming increasingly true as the federal budget was changing from deﬁcit to surplus over the period that this legislation evolved. Among the primary beneﬁciaries of the new welfare law were the states. What did the Governors gain? A central role in the crafting of the legislation which resulted initially in the elimination of most of the bureaucratic restrictions in return for accepting a ﬁxed block grant rather than federal matching funds. With this freedom from Washington bureaucrats and regulations, the capacity of the governors to design programs more acceptable to their local constituency was enhanced, and with it their reelection prospects. While the block grant held certain risks, particularly in recessions when federal matching funds would have been increased, there was concern among governors that the federal government had been trying to improve its budgetary situation by passing the buck to the states (Smith, 1998, p. 21). As one political scientist remarked in 1994, ‘the political viability of Congress in today’s budget climate rests heavily on its ability to meet interest group demands through unfunded mandates’ (Kincaid, 1994, p. 576). Governors feared this process might continue, and a block grant would hedge against this risk. As it turned out, not only did they gain a great deal of control over the programs in their states, they also received vastly greater funding than they

Institutional analysis

165

would have under the former legislation. According to Governor Leavitt, it was the governors who, when meeting with Republican leaders in Williamsburg, ﬁrst proposed block grants: We were sitting around a table. I don’t remember exactly who said it but (a Governor) stated ‘we would be willing to trade a level amount of funding over the next ﬁve years if we could have the ﬂexibility to manage (the programs)’ . . . I remember John Kasich getting very excited about it, saying, ‘If we could do that, we could balance the budget’ (Smith, 1998, p. 155). The programs the Governors were talking about were not just the AFDC programs, however, they were also including Medicaid, which was devouring state budgets far more ferociously than was AFDC. Between 1980 and 1990, for example, Medicaid costs to states grew from $11.2 million to $31.4 million, and reached $61.9 million in 1994. During the same period AFDC grew from $6.2 million in 1980 to $9.5 million in 1990 and $11.9 million in 1994. [See Table 8.1.] The growth in Medicaid costs far outweighed the growth in the number of Medicaid recipients and the Governors felt that with greater ﬂexibility in the programs they could provide services at signiﬁcant savings to state budgets (Smith, 1998, p. 157).

As Table 8.1 shows, the rate of increase in federal expenditures also provided an incentive for the Republican leaders to cap funding for the program. When Medicaid was detached from welfare in order to avoid a Presidential veto of welfare reform, it was a major blow for state Governors. Those Governors who took the lead on welfare reform had committed many resources to helping the Republicans formulate policies and craft legislation. They assigned senior staﬀ to do work that Congress would normally do, including drafting bills, in response to a promise from Republican leaders that they would have inﬂuence. In this quid pro quo the expertise of the Governors would help propel the legislation a swift passage before interest groups or opponents could intervene. But once the Governors had provided assistance in formulating the bills, and incidentally, made it unlikely that they would attack the legislation later or blame the Congress, the Republican leadership had little further need of them. Thus, it is not surprising to hear Governor Leavitt remark, when reﬂecting on the Williamsburg RGA meeting, ‘From that point forward our actual inﬂuence on the process diminished’ and the product that emerged 18 months later was ‘substantially diﬀerent’ from what the Governors had originally proposed (Smith, 1998, p. 159). The change was brought about, to a large extent, by the increasing inﬂuence of the ideologically conservative interest groups in early 1995, under the informal leadership of Robert Rector of the Heritage Foundation. Although both the Governors and the interest groups were

166

Table 8.1 Fiscal Year

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996

The political economy of welfare reform in the United States

State historical spending on AFDC and Medicaid programs State AFDC Expenditures Beneﬁts

Administrative Costs

Medicaid Program Costs

1,443 2,469 2,942 3,138 3,300 3,787 4,418 4,762 4,890 4,954 5,508 5,917 5,934 6,275 6,664 6,763 6,996 7,409 7,538 7,807 8,390 9,191 9,988 10,016 10,286 10,014 9,613

186 254 241 296 362 529 527 583 617 668 729 814 878 915 822 889 967 1,052 1,159 1,206 1,303 1,300 1,342 1,438 1,612 1,754 1,796

2,235 2,802 4,074 4,113 4,396 5,578 6,332 7,389 8,269 9,489 11,231 13,303 14,931 15,971 17,508 18,262 19,856 21,909 23,654 26,642 31,389 38,987 50,339 56,236 61,885 67,193 71,585

Source: Committee on Ways and Means (1996).

political conservatives, the former had a vested interest in limiting federal mandates, while the latter wanted to promote behavioral changes through legislation. Robert Rector did not feel that the Governors could be relied upon to implement policies that would promote his agenda of reducing illegitimacy and strict work requirements and thus sought to force their hand through legislation (personal interview, 11 December 1999). The House passed bill reﬂects the power of the conservative groups who won out over the Governors in many of these battles. It also represents the power of the

Institutional analysis

167

deﬁcit hawks who made budgetary savings by cutting oﬀ many immigrants versus the governors and Congressmen from states such as Florida with high concentrations of immigrants (Smith, 1998, p. 177). Although the Governors may not have obtained as much control over their programs as they had sought, ﬁnancially they ended up receiving considerably more federal dollars than they would have received under the old welfare law. This was, to some extent, in spite of the Republicans rather than because of them. The funding formulas were based on average caseloads from 1992 to 1994, or actual caseloads in 1994 or 1995, whichever was largest. The historic rapid rise in caseloads came to a halt in March of 1994, however, and declined signiﬁcantly over the following years. In 1998, when Congress realized just how much federal money the states had gained as a result of the block grant formulas they tried to recuperate some of these funds, causing an outcry from the Governors who complained that the block grants were an entitlement to the states and successfully blocked their removal. The era of welfare entitlements was evidently not yet over.

NOTES 1. 2. 3. 4. 5.

Personal interview with Mary Kay Mantho, member of Governor Thompson’s staﬀ. Personal interview with Leeann Reddick, member of Governor Engler’s staﬀ. See the Chamber publication ‘1996 How They Voted’ for illustration of this point. Personal interview with Arloc Sherman. Telephone interview with Sharon Daly, formerly of CDF and currently with Catholic Social Services. 6. Personal interview with Robert Rector. 7. Memorandum from Empower America Co-Directors William Bennett, Jack Kemp and Vin Weber to Congressional Republicans on the subject of the House Republican Welfare Bill, dated April 13, 1994.

9. An econometric analysis of the variables aﬀecting changes in welfare caseloads In this chapter the comparative analysis of the public interest and public choice theories of government shifts both in focus and methodology. The focus of the previous chapter was on the legislative process and the methodology was institutional analysis. In this current chapter the focus is on legislative outcomes and the methodology will be an econometric analysis of a speciﬁc outcome – changes in welfare caseloads. A primary objective of those interest groups shown to have most power in the political market, and thus most inﬂuence on the legislative process of welfare legislation, was the reduction in welfare caseloads. Business interests, represented by organizations such as the Chamber of Commerce, were already feeling the eﬀects of a tightening job market in the early 1990s, and had an interest in increasing the available supply of labor, which could be achieved by moving people from welfare to work.1,2 The Christian conservative groups and their allies had a strong ideological belief in the value of work and the detrimental eﬀect of welfare dependency and for this reason wanted people oﬀ the welfare rolls. As discussed earlier, those groups opposed to welfare reform had neither the organizational and ﬁnancial resources, nor the popular support, of the reform proponents. If, indeed, the welfare legislation can be shown to have had a substantial, measurable eﬀect on welfare caseloads when other factors such as economic growth and demographic change are controlled, this would add further strong support for the eﬀectiveness of powerful interest groups and for public choice theory more generally. It is important not to overstate the case, however. Clearly, a public interest model could have a similar outcome. The point is that the primary focus of the latter is a Pareto preferred situation in which some individuals are better oﬀ and no individuals are worse oﬀ. A reduction in the number of persons in poverty and/or increases in employment would be predicted outcomes of this model. Such outcomes might or might not be associated with a reduction in welfare caseloads. There is no clear evidence that moving from welfare to work per se makes individuals eco168

Variables aﬀecting changes in welfare caseloads

169

nomically better oﬀ in the short term. Individuals who move into minimum wage jobs, losing not only welfare beneﬁts but also Medicaid, child care and other beneﬁts, may indeed be less well oﬀ ﬁnancially, at least in the short term.3 While an analysis that incorporated the eﬀects of welfare legislation on these other outcomes might provide a more ideal test of the alternative theories it is clearly well beyond the scope of this book. This ﬁrst part of the chapter provides descriptive background on historical changes in welfare caseloads since the inception of the federal welfare program. This historical perspective provides a background for the next two parts of the chapter which examine existing econometric studies of the determinants of caseload, and present an alternative model which incorporates a longer time series and an alternative speciﬁcation of explanatory variables. The final part of the chapter extends the analysis to examine several areas of speciﬁc interest – the relationship between cash beneﬁts and births to unmarried mothers, the median voter model, and the inﬂuence of special interest groups on welfare beneﬁts. Econometric models which examine these issues, as well as a model which combines the eﬀects of special interests and median voter interests, are speciﬁed and tested, and the implications of the results are discussed. The growth in the number of persons receiving cash welfare beneﬁts is presented in Figure 9.1 After relatively rapid growth in the late 1960s and early 1970s, a period during which eligibility was expanded and information on the program more widely disseminated, caseloads were quite stable until the recessions of the early 1980s. A sharp upturn at this time was curtailed by the introduction of legislation in President Reagan’s ﬁrst term that ended eligibility for over 10 percent of the population. As states took measures to oﬀset some of these eﬀects, and as others were reversed by the federal government, caseloads returned to their levels prior to passage of the legislation and again remained relatively stable until 1989. The period 1989 to 1994 saw a sharp increase in the caseload, with the maximum being reached in March 1994. The reasons for this increase are not clear; a mild recession in 1990–1991 would not explain the magnitude and persistence of the caseload growth. This period of rapid growth gave impetus to the welfare reform movement. From an all-time high in March 1994, the welfare caseload fell dramatically. As of June 1999, 2.5 million people were receiving welfare, lower than at any time since 1967 and a drop of almost 50 percent over a ﬁve year period. The data in Figure 9.1 represent the total number of households receiving AFDC, and, since 1996, TANF payments. The AFDC program had two components, AFDC-Basic, which served single parents and their children, and AFDC-Unemployed Parent (AFDC-UP), which is paid to

1970

Figure 9.1

0

1,000,000

1978

1977

1976

1975

1974

1973

1972

1971

Historical trends in caseloads

June 1999 Caseload 2.5 million

1990

2,000,000

1991

3,000,000

1992

4,000,000

1993

5,000,000

March 1994 Caseload 5.1 million

1994

1989

1988

1987

1986

6,000,000

1995

170

1998

1997

1996

1985

1984

1983

1982

1981

1980

1979

Variables aﬀecting changes in welfare caseloads

171

low-income married couples and their children. The AFDC-UP program was only available in about half the states prior to 1990, when all states were obliged to implement it. In several states the program was started and dropped during the 1970s and 1980s. Even after 1990, the program accounted for less than 10 percent of the total number of AFDC cases. It is clear that numerous factors are at work which aﬀect the number of people moving on and oﬀ the welfare rolls at any point in time. A primary objective of this chapter is to examine a range of variables that might plausibly be hypothesized to aﬀect caseloads, and by doing so shed light on the eﬀectiveness of the 1996 legislation. The central hypothesis being tested is that the TANF legislation, so strongly lobbied for by the interest groups described earlier, had a strong negative impact on welfare caseloads. Before specifying a model of caseload changes, however, it is useful to examine some of the studies that have already been carried out. There are few existing econometric studies of models of changes in AFDC caseloads, and most of those that do exist are at the state rather than the national level. Studies of changes at the state level have been carried out for California (1988), Delaware (1991), Massachusetts (1989, 1990), New Jersey (1988) and Washington (1987, 1989), as well as for New York City (1976, 1987). National models have been developed by the General Accounting Oﬃce (GAO, 2002), the Congressional Budget Oﬃce (CBO, 1993), the Council of Economic Advisors (CEA, 1997), Rebecca Blank of Northwestern University (1997), and Ziliak et al. of Oregon University (1997). The following discussion examines in some detail two of the major national studies, that by the CEA and that by Rebecca Blank, which are most instructive for the current analysis.

THE CEA STUDY Amongst all the studies described above, one of the most widely discussed, and controversial, was that of the CEA. This report, entitled ‘Technical Report: Explaining the Decline in Welfare Receipt 1993–1996’ (1997), updated in 1999, is the outcome of a longer term study of AFDC caseload trends by the Council. The objective of the study was to analyze the eﬀect of economic and policy factors on variations in the share of the population receiving AFDC. The principal conclusion of the study was that: ‘over 40 percent of the decline (in the aggregate national AFDC caseload from 1993–1996) can be attributed to economic growth and that almost onethird is related to waivers, particularly those that sanction recipients who do not comply with work requirements’ (1999, p. 2).

172

The political economy of welfare reform in the United States

Econometric Speciﬁcation Two basic models were developed by the CEA to estimate the eﬀects of policy and economic conditions on caseloads during the period 1976 to 1998. In both models the dependent variable was the ratio of AFDC recipients to the state’s population. And both speciﬁed the minimum wage and maximum welfare beneﬁts in identical ways. The remaining policy variables had alternative speciﬁcations. In Model 1, dichotomous variables were used to capture the period when a major waiver from federal AFDC policy was in eﬀect in the state and the period when TANF was in eﬀect. The speciﬁcation of this model was as follows: lnRst WaiverstBw TANFstBtanf lnBenefitsstBb lnMinWagestBmw UnemploymentstBu ys yt trend*ys est

(9.1)

where the variables are deﬁned for state s, in calender year t, as follows: R is the ratio of the number of recipients to the population under 65 years of age (the number of recipients is obtained from administrative reports on AFDC/TANF); the model estimates the natural log of this ratio. Waiver is an indicator variable that takes the value of one if the state had a major waiver in eﬀect; the indicator is turned oﬀ when TANF is implemented in the state. TANF is an indicator variable that takes the value of one if TANF was in eﬀect in the given state. Beneﬁts is the maximum monthly beneﬁt for a family of three on AFDC/TANF. MinWage is the value of the state-speciﬁc minimum wage expressed as a monthly amount (to make comparable with the beneﬁts variable) assuming employment for 30 hours per week for 4.33 weeks. Unemployment is the unemployment rate (current, lagged 1 year, lagged 2 years) Ys is state ﬁxed eﬀects. Yt is year ﬁxed eﬀects. Trend*ys is linear state speciﬁc time trends. The study was based on annual data from 1976–1998 from the 50 states and the District of Columbia, providing a total of 1,173 observations. The authors of the study note that state and year ﬁxed eﬀects variables and state-speciﬁc time trends are included to capture unobserved factors, such

Variables aﬀecting changes in welfare caseloads

173

as family structure and other policies that may be correlated with the observed variables (CEA, 1999, p. 12). All dollar values are expressed in constant 1998 dollars using the CPI. In the second model the eﬀect of speciﬁc welfare policies was incorporated, regardless of whether the policies were implemented under waivers or TANF. Model 2 was speciﬁed as follows: lnRst XstBx lnBenefitsstBb lnMinWagestBmw UnemploymentstBu ys yt trend*ys est

(9.2)

where Xst represents a vector of variables that describe speciﬁc policies that are in eﬀect in state s, in year t. Although a broad range of policies could have been tested, the authors choose ﬁve policy areas including the imposition of time limits, a family cap, work exemptions for mothers of very young children, work sanctions and earnings disregards. Results Table 9.1 contains the estimates of Models 1 and 2. The results for Model 1 imply that states that implemented a major waiver experienced a decline in participation that was 8 to 9 percent greater than in states that did not. The eﬀect of TANF was twice as great, with a decline in participation of 18 percent. All other variables had the expected sign. Higher beneﬁts increased participation, as did lower minimum wages. While the current unemployment rate was not statistically signiﬁcant, the 2 year lag of this variable was highly signiﬁcant and associated with an increase in caseloads of over 4 percent. The results for the speciﬁc policy variables included in Model 2 show, as the authors note, ‘mixed results’ (CEA, 1999, p. 17). The time limit variable is negative, as expected, but not precisely estimated, and the higher earnings disregard is positive, but the eﬀect is relatively small. The eﬀect of work sanctions was negative and signiﬁcant, as expected. The family cap variable, however, is surprisingly positive and strongly signiﬁcant. The implication that cutting oﬀ beneﬁts for additional children born to mothers already on welfare increases caseloads is hardly tenable. It is noteworthy that Rebecca Blank, who also incorporated a waiver variable, concluded that the variable represented other changes going on in states that applied for waivers, rather than the speciﬁc eﬀect of a particular waiver. Noting the unusually large eﬀect of waivers in her regression results, she questions whether the waiver variable ‘is measuring the eﬀect of the direct program implemented by the waiver or whether it is acting as a proxy for a whole set of changes that occurred in states where waivers were implemented’ (Blank, 1997). She tests this by including not only the waiver

174

Any waiver TANF Log maximum monthly beneﬁt Log monthly minimum wage Unemployment rate: Current 1 year lag 2 year lag Speciﬁc welfare policy variables (X) Termination/work req. time limit Family cap Work exemption based on age of youngest child: Traditional AFDC & JOBS Exemption (reference group) Child as old as 6 months to 3 years Child newly born to 6 months old No exemptions base on age of youngest child Work sanctions: Traditional AFDC or JOBS (reference group) Partial/Partial Partial/Full Full/Full Log earnings disregard State-speciﬁc trends?

2.90 4.37 1.93 4.02 0.74 2.40 8.92

9.40 18.84 14.98 39.59 0.36 1.50 4.27

Yes

t-stat

Beta

Model 1

No

0.20 1.70 5.13

7.99 18.12 51.74 63.91

Beta

0.30 1.88 7.40

2.90 1.75 6.20 3.61

t-stat

Model 1A

Table 9.1 CEA Models 1 and 2 (coeﬃcient estimates are multiplied by 100)

0.76 2.19

2.46 1.53 0.77

2.52 3.76 5.57 2.40

12.37 11.56 4.86 9.71 18.14 39.36 5.38 Yes

0.61 2.06 8.34

2.37 2.27

t-stat

3.75 6.71

0.30 1.29 3.94

15.01 25.59

Beta

Model 2

1.36 22.76 33.53 5.86 No

2.79 3.05 0.81

4.30 8.21

0.13 1.65 4.77

53.84 51.95

Beta

0.32 4.20 4.51 2.00

0.57 0.40 0.12

0.73 2.35

0.20 1.92 7.39

7.63 2.74

t-stat

Model 2A

Variables aﬀecting changes in welfare caseloads

175

variable but a lead of the waiver variable, noting that ‘lead values cannot possibly signal program eﬀects, and must indicate that the waiver variable is correlated with other changes in the state’ (Ibid.). Her results show that the lead variable is larger and more signiﬁcant than the current waiver variable, suggesting that ‘something was changing in those states prior to the implementation of the waivers that reduced caseloads, and that the waiver programs themselves are not the primary cause of the caseload reductions’ (Ibid.). These other things could include administrative changes or media publicity about the states new ‘get tough’ laws that would discourage new applicants and encourage current ones to look for alternative sources of income.4 Finally, work exemption waivers based on the age of the youngest child are not strongly associated with changes in caseloads, and the one signiﬁcant eﬀect is of unexpected sign. On the basis of the above estimates, the authors of the study attempt to assess the relative contribution of economic and policy factors to changes in caseloads from 1993–1996 (the waiver period under the Clinton Administration) and 1996–1998 (the TANF period). The method used for determining the share of caseload change attributable to each of the factors was as follows. First, using the coeﬃcients estimated in the model, the share attributable to change in unemployment was calculated, had all other variables stayed constant at their 1993 values. This unemployment-eﬀects-only change in the utilization rate was then divided by the actual change in the utilization rate to produce an estimate of the share (expressed in percent) of the change attributable to reductions in unemployment. This computation was performed state by state and then averaged to produce a national estimate, using state population as weights. A similar procedure was used for waivers. All other variables, including the unemployment rate, were held constant at 1993 levels, so that the predicted change was the result only of growth in the number of approved waivers, as measured by the waivers indicators. Using this methodology on the estimates in Model 1, the authors ﬁnd that waivers accounted for about 14 percent of the decline in caseloads from 1993–1996, while the lower unemployment rate was responsible for about 26 percent (see Table 8.1). A decline in cash beneﬁts contributed to a 6 percent decline in caseloads during the same period, while the fall in the real value of the minimum wage reduced the caseload decline by almost 10 percent. For the period 1996–1998, when TANF was in eﬀect, over 36 percent of the caseload decline is associated with TANF. The unemployment rate accounts for almost 8 percent during this period, while higher minimum wages account for almost 10 percent and lower cash beneﬁts for an additional 1.4 percent.

176

The political economy of welfare reform in the United States

THE REBECCA BLANK STUDY Noting the ‘sparse research literature on the determinants of aggregate caseload changes’, Rebecca Blank investigates ‘the role of macroeconomic forces, public policies, and demographic change in explaining caseload changes over time’ in her paper, ‘What Causes Public Assistance Caseloads to Grow?’ (1997). She uses more extensive data, both across states and over time, than any previous research, and in extensions of the model incorporates the eﬀects of programs related to AFDC/TANF, including Food Stamps and Medicaid. Econometric Speciﬁcation Blank estimates a series of models of the following form: Caseloadts y1Xts y2Dts y3Pts vt rs wts

(9.3)

where Caseload is measured as the log of the AFDC-Basic caseload/ female population ages 15–44. Blank explains that this measure is chosen since almost all AFDC households are headed by women of childbearing age. (Blank also estimates a model in which the denominator is not used.) X is a vector of economic factors, including the unemployment rate with two lags, median wage and the twentieth percentile wage. D is a vector of demographic factors, including race, education, marital status, immigrant status, and age. P is a vector of policy parameters, including waivers, beneﬁt levels, whether a UP program is being implemented, Medicaid beneﬁts, and the party aﬃliation of state governors and representatives. As in the CEA model, state and year ﬁxed eﬀects are included; the v vector represents year ﬁxed eﬀects and the r vector represents state ﬁxed eﬀects. The subscripts s and t refer to states and years. The w is a random error term. Equation 3 is estimated with an OLS procedure on annual data from 1977 through 1995 for 50 states and the District of Columbia, providing a total of 969 observations. Results A summary of the results obtained by Blank is presented in Table 9.2.

177

Variables aﬀecting changes in welfare caseloads

Table 9.2

Rebecca Blank Model (summary of results)

Basic Speciﬁcation

Unemployment Rate Unemployment Rate 1 Unemployment Rate 2 Log (Median Wage) (0.118) Log (20th Wage Percentile) % Black (0.725) % Single Female Heads Years of Education (0.027) % Elderly (0.404) % Immigrants1 (x100) % Immigrants2 (x100) Party of Governor (1Republican) Both State Senate and House Democratic Both State Senate and House Republican Log (Maximum AFDC Beneﬁt Level) AFDC-UP program Log (Avg Family Medicaid Expenditures)1 Any Major Waiver (0.020) State Eﬀects Year Eﬀects State Time Trends Number of obs.

Column 1 With state time trends

Column 2 With no time eﬀects

0.007 (0.005) 0.014 (0.006) 0.017 (0.005) 0.774 (0.094) 0.104 (0.084) 0.307 (1.079) 1.353 (0.466) 0.046 (0.023) 1.502 (0.322) 0.031 (0.024) 0.074 (0.025) 0.030 (0.008) 0.026 (0.012) 0.019 (0.017) 0.559 (0.046) 0.113 (0.014) 0.039 (0.009) 0.107 (0.016) Yes Yes No 969

0.001 (0.003) 0.011 (0.004) 0.017 (0.003) 0.324 (0.124) 0.260 (0.057) 3.742 (0.819) 0.584 (0.332) 0.025 (0.020) 0.107 (0.439) 0.011 (0.016) 0.022 (0.017) 0.030 (0.005) 0.013 (0.009) 0.008 (0.011) 0.218 (0.040) 0.085 (0.012) 0.008 (0.008) 0.041 (0.021) Yes Yes Yes 969

Notes: Standard errors in parentheses. All regressions based on data for 51 states from 1977–1995. 1 Average state expenditures for a family with one adult and two children.

178

The political economy of welfare reform in the United States

The dependent variable is the log of the ratio of the number of AFDCBasic households divided by the female population ages 15–44 in state s at time t. Blank uses this ratio because it ‘presents caseloads as a share of the population group that is most likely to be available as a household head for an AFDC-Basic case’ (1997, p. 14). This caseload share varies from 6 to 8 percent of the young female adult population over this time period. Basic results are presented in column 1, with economic variables listed ﬁrst. The unemployment rate has a major impact on caseloads, and is associated with a 0.7 percent increase in the current year, 1.4 percent with a one year lag and 1.7 percent with a two-year lag. The log of median wages has a strong negative eﬀect on caseloads, and the log of the twentieth percentile has a negative eﬀect, but much weaker. One of the most signiﬁcant demographic variables is the share of femaleheaded households. As Blank notes: ‘if you believe that the formation of single mother households is inﬂuenced by the presence and level of AFDC, then this variable is endogenous and its coeﬃcient is biased upward, although the evidence suggests this is not a major problem . . . excluding the variable has little eﬀect on the other coeﬃcients’ (1997, p. 15). The relationship between cash welfare payments and family structure and size, particularly the number of children born to unwed welfare mothers, is highly controversial, however. While Blank does not believe there is signiﬁcant correlation, Charles Murray and others believe there is a strong causal relationship from AFDC/TANF beneﬁts and beneﬁt levels to unwed births. The existence and signiﬁcance of such a relationship will be modeled and tested in the final part of this chapter. A second demographic variable in the Blank model, the percent elderly, is negatively correlated with caseloads, a result that Blank does not try to explain, but which is discussed further below. Race, education and immigration in the current year show little association, although immigration lagged two years is positively correlated, suggesting increased immigration leads to increased demand for AFDC beneﬁts over time. Of the political variables, the party of the governor is strongly signiﬁcant, with lower caseloads under Republican governors. Blank suggests this may indicate that ‘governors are able to shape the administrative processes by which public assistance is provided’ (Ibid.). States in which both houses are controlled by the same party have lower caseloads, regardless of whether the party in control is Republican or Democrat. She attributes this to the diﬃculty of passing welfare reform legislation when there is split party control. The policy variables are generally signiﬁcant. An increase in AFDC beneﬁts is associated with an increasing caseload. However, at least a part

Variables aﬀecting changes in welfare caseloads

179

of this is automatic since a higher beneﬁt level automatically makes more families eligible as a result of the way the beneﬁt formula is calculated. The sign on the AFDC-UP variable is unexpectedly positive, which Blank interprets as reﬂecting that this variable is a proxy for the generosity of the state in AFDC-Basic programs. To the extent that the generosity of the state indicates a liberal ideology, this provides further evidence for the inﬂuence of ideology, and re-enforces the argument that ideology should be incorporated into such analyses. In a similar vein, family Medicaid expenditures are positively associated with caseloads, indicating, according to the author that ‘states whose Medicaid policies have been more expansive have seen greater public assistance usage’ (Ibid., p. 16). Finally, states that received a waiver had signiﬁcantly reduced caseloads. This last result is somewhat surprising since waivers were generally introduced relatively slowly and initially aﬀected only a small share of the caseload, since they were usually demonstration projects in the ﬁrst phase. Blank concludes that waivers are acting as a proxy for ‘other changes occurring and even preceding their implementation that are causing caseloads to decline in states that seek waivers’ (Ibid., p. 20). Blank does not indicate what the speciﬁc changes might be, but the ideological climate in the state is an obvious candidate. Column 2 uses the same speciﬁcation as column one, but includes state speciﬁc time trends, which provides some indication of how many of the signiﬁcant variables ‘are signiﬁcant simply because they are trending up (or down) in a linear way’ (Ibid., p. 16). Blank notes that the inclusion of this variable reduces the magnitude of most of the estimated coeﬃcients, although the sign and signiﬁcance is not greatly changed. Other speciﬁcations that Blank tests include removing all time trends and time eﬀects and splitting the data into two time periods, 1977–1985 and 1986–1995. She concludes that the results: ‘suggest that the data estimation results for AFDC-Basic caseloads are generally robust to changes in speciﬁcation and to various time periods. Caseloads are strongly aﬀected by both macroeconomic factors and by programmatic and political factors. Demographic factors are important, but their signiﬁcance varies across speciﬁcations’ (Ibid., p. 18).

CRITIQUE The CEA study has been criticized on a number of grounds, including the speciﬁcation and interpretation of the variable measuring the eﬀect of waivers. The restriction of the analysis to waivers issued after 1991 disregards some important innovations that were approved and implemented by

180

The political economy of welfare reform in the United States

the states prior to this cutoﬀ, according to Martini and Wiseman (1997). Moreover, regarding the relationship between waiver approval and caseload decline, Martini and Wiseman question the direction of causality and argue that state requests for waivers may result from an unanticipated reduction in the caseload, which frees up resources for greater welfare activism, including waiver requests. Other criticisms of this model include the choice of dependent variable and the returns to alternative opportunities. The CEA model uses ratio of persons using AFDC to the entire population, thus including many people not currently or potentially eligible for AFDC, and in combinations that diﬀer across states. An alternative procedure used by Blank, is to use the ratio of AFDC recipients to adult women aged 15–44 years, since virtually every AFDC case is associated with a mother (Blank, 1997). The CEA model includes a measure of the gains from welfare, but no measure of alternative gains from employment. Blank (1997) controls for wage trends using wage distribution data from the monthly Current Population Survey. While this measure is somewhat incomplete, ignoring, for example, changes in the Earned Income Tax Credit (EITC), it is nevertheless an improvement over completely ignoring the inﬂuence of earned income. Overall, the Blank model includes a wider range of descriptive variables, several of which add signiﬁcant explanatory power to the analysis. However, while the CEA model uses an annual data series that includes data up to 1998, the analysis undertaken by Blank terminates in 1995, prior to the passage of the new welfare law (Dr Blank has recently updated and extended her earlier work, but these additions do not signiﬁcantly aﬀect her earlier ﬁndings). The model presented below seeks to address these issues by incorporating data up to 1999, thus reﬂecting three to four years of implementation of TANF, and by also including several of the variables Blank found to be signiﬁcant, as well as others that are hypothesized to be signiﬁcant. A Proposed Alternate Model An econometric model is speciﬁed and tested using panel data from 50 states for the 20 year period 1980 to 1999, the most recent year for which a full set of data for all the variables in the model was available. This data set provided a total of 1,000 observations. Unlike the CEA and Blank studies cited earlier, this study did not incorporate data from the District of Columbia. When data sets included data from the District of Columbia and the states and territories, such as Puerto Rico and Guam, they were cleaned of these data before any regressions were run. The unique character of the District,

Variables aﬀecting changes in welfare caseloads

181

particularly in terms of demographics and legislative structure, when compared to the states, rendered it inappropriate to specify the relationships under consideration by the same model. Moreover, the use of data containing extreme outliers can lead to regression results which diﬀer substantially and have greater variance than would estimates calculated without the outliers. The model was therefore estimated on the basis of state only data in order to provide a more eﬃcient estimate. Following the procedure used by the CEA, Blank and others, the methodology to be employed in this study is that of multiple regression analysis. All dollar-denominated variables were converted to constant 1999 dollar values. Econometric speciﬁcation The model to be tested is the following: Caseloadts y1Xts y2Dts y3Pts y4Its vt rs wts

(9.4)

where the dependent variable is the total number of cases (households) receiving beneﬁts, divided by state population. X is a vector of macroeconomic variables, including median wages and unemployment with two lags, D is the demographic variable, births to unwed mothers, P is a vector of federal and state policy variables, OBRA, TANF, waivers and beneﬁt levels, I incorporates the proxy for ideology, party of the state governor, Vt is year ﬁxed eﬀects, Rs is state ﬁxed eﬀects, and Wst is a random error term. The policy variables incorporated into this model are those that reﬂect major federal and state policy changes regarding the AFDC/TANF programs. At the federal level these include: the Omnibus Budget Reconciliation Act (OBRA) of 1981, which sharply curtailed eligibility for beneﬁts, as well as the new welfare law (PL 104–193) which introduced the TANF program in 1996. At the state level, the introduction of waivers marked changes in welfare policies, although measurement of these eﬀects is diﬃcult since many waivers were introduced in stages over the course of several months or even years, and in some cases certain provisions were later rescinded. These diﬃculties and the interpretation of the waiver variable are discussed further below. The ﬁnal policy variable to be included is the level of cash beneﬁts paid by states to welfare recipients (clearly, this could also be classiﬁed as an economic variable, aﬀecting the individual’s economic choices).

182

The political economy of welfare reform in the United States

The economic variables in the model include the unemployment rate in the state, with a one year and a two year lag, and median income in the state. The availability of alternative sources of income to welfare would clearly be expected to aﬀect caseloads, and thus a strong positive relationship between the unemployment rate and the caseload level would be expected. Moreover, to the extent that those who become unemployed spend some time in job search activities prior to seeking welfare beneﬁts, there would be an anticipated lag between changes in the explanatory and dependent variables. Thus both current and lagged unemployment data were included in the model. The other economic variable included in the model is median income. Even where employment opportunities are available, employment may not be the economically rational choice for a welfare mother. The cost-beneﬁt calculus between welfare and work is clearly aﬀected by the level of wages in jobs available to her in comparison to the welfare payments she is entitled to receive. A negative correlation between median wages and caseloads is hypothesized, reﬂecting the increased preference for work as median wages rise relative to welfare payments in a state. The use of median income as a measure was selected in part because of the theoretical issues related to the median voter theorem discussed earlier in this chapter. The analysis is extended in the final part of the chapter, with the development of a supplementary model. It was originally intended to include two demographic variables: the number of immigrants entering the state since children of immigrants born in the United States are eligible for cash beneﬁts if the household income is suﬃciently low and other requirements are met, and the growth in female-headed households, since growth of this category would cause a compositional shift in the overall population that would lead to a higher caseload. The diﬃculty of obtaining accurate and timely data on immigrants precluded the use of this variable at this time, however. Among other issues is the problem that while immigrants must record a state in which they intend to reside, it is not always possible to verify changes in residency. Data at the state level on female-headed households with children also proved illusory. The closest approximation, which was used by Rebecca Blank, was a ﬁgure for ‘the share of households headed by a single woman and including other related persons in the households’ (Blank, 1997, Data Appendix). The source of this data was the Outgoing Rotation Group of the Current Population Survey. In this survey there is a rolling sample and the outgoing group each month is surveyed and the 12 monthly ﬁgures (which comprise diﬀerent samples) are then aggregated to obtain an annual ﬁgure. In addition to statistical limitations from the survey method, the data obtained is problematic in that it incorporates a far broader category than the one sought. Included, for instance, are households consisting only of adults, households in which there may be children but where those

Variables aﬀecting changes in welfare caseloads

183

children are not the children of the household head, multiple family units in which a female of one generation lives with the (married or unmarried) members of the next generation and their oﬀspring, and so on. Clearly a proportion (which is unknown and which could diﬀer by state and over time) is not eligible for the program. An alternative, which was suggested by Professor Blank, was the use of the percentage of births in the state in which the mother was unmarried. The ﬁnal variable to be included in the model is a dummy for the party of the state governor. It is hypothesized that ideology plays a signiﬁcant role in the legislative process, just as it plays a signiﬁcant role in individual behavior, and a conservative ideology would be negatively associated with caseload growth (reﬂecting both the impact on the behavior of individuals and the degree to which the state welfare program is liberal or conservative in its eligibility and beneﬁts policies). A Republican governor is a proxy for a conservative ideology. Results The results of the regression are presented in Table 9.3. All of the economic variables are signiﬁcant and have the hypothesized sign. The current unemployment rate as well the rates with a one and two year lag are strongly positively associated with caseload levels, suggesting that where employment opportunities are available a certain percentage of those otherwise on welfare will take advantage of these opportunities. This result is in accord with the relationship hypothesized and the results of the other studies discussed above. Real median income is strongly negatively associated with higher caseloads, as hypothesized. This result reﬂects both a change in the cost beneﬁt calculus of potential welfare recipients facing a work or welfare choice, and the fact that at higher overall income levels, which are likely correlated with median income levels, fewer households are eligible for cash beneﬁts. The two major policy variables, TANF and beneﬁt levels, were also signiﬁcant and of the expected sign. Implementation of the new welfare law, TANF, is strongly signiﬁcant and, as hypothesized, negatively associated with caseload levels, reﬂecting a legislative output that rewarded those powerful groups seeking a cut in the welfare rolls. This clearly lends strong support for the public choice hypothesis. The second major policy variable, the maximum beneﬁt level for a family in a state (which is also an economic beneﬁt in the sense that it enters the individual’s economic cost/beneﬁt calculus) was positively related to caseload level and was highly signiﬁcant. This is perfectly in accord with rational choice theory, and provides further evidence for the power of policies to inﬂuence economic and social behaviors.

184

Table 9.3

The political economy of welfare reform in the United States

Reintsma Basic Model

Variable

coeﬃcient

t-statistic

GOV1 OBRA TANF WAIVERIMP2 AMB3IN993 NWEDBRTH REALMDIN99 UNEMPRATE URATE1LAG URATE2LAG STFIX YRFIX

0.000188 0.001224 0.004860 0.001104 1.21E-05 0.049654 7.42E-08 0.000304 0.000291 0.000323 3.39E-05 0.000107

0.885766 2.870010*** 11.251770*** 2.837493*** 15.90039*** 23.54542*** 3.060389*** 2.738993*** 1.993585** 3.142788*** 4.506424*** 2.99732**

Notes: R-squared 0.583513 Adjusted R-squared 0.578449 F-statistic 115.2351 Prob. F statistic 0.000000 ** Signiﬁcant at 95 percent conﬁdence level *** Signiﬁcant at 99 percent conﬁdence level 1 Party of State Governor. 2 Dummy1 if waiver implemented. 3 The maximum beneﬁt for a family of three in the state.

The two other policy variables, waivers and OBRA, did not have the expected sign for reasons suggested below. The sign for the waiver variable is positive, as is the sign for OBRA, and both of these are signiﬁcant. There are several possible explanations. One is the usual that these variables may not in fact represent what they are supposed to represent. The waiver variable, for instance, which reﬂects only the date of introduction of a major waiver, may not reﬂect the extent to which the waiver was introduced throughout the state, the speed with which it was fully implemented, or the extent to which the provisions aﬀected the majority of the caseload. Rebecca Blank concluded in her study that the waiver variable was largely a proxy for other related policy changes in the states. The OBRA variable reﬂects two policy initiatives, one in the ﬁrst year of implementation of this federal policy, when rapid caseload declines were caused, and a second when states took measures to counteract the aﬀects of the federal legislation, and introduced their own oﬀsetting provisions. To the extent that a state policy counteracted the federal policy, the caseload would not be expected to decline in that state over the longer term when the state policy was fully implemented. In fact, the states may have overcompensated for federal cutbacks leading to the observed outcome for this variable.

Variables aﬀecting changes in welfare caseloads

185

An alternative explanation for the diﬀerences in the eﬀectiveness of the policy variables relates to public choice theory. The role of interest groups in the development and passage of TANF was tremendous. As explained in the preceding chapter, some of the most highly ﬁnanced and best organized groups participated in intensive lobbying for TANF. Their primary objective was to reduce caseloads. The regression results suggest they were very successful. In contrast, state waivers received far less attention, especially from the national lobbying interests, and caseload reduction was not necessarily the overriding immediate goal. It is noteworthy that in several of the other major studies the statistical eﬀect of this variable was not as expected. Rebecca Blank suggested that the waiver variable ‘acts as a proxy for a whole set of changes that occurred in states where waivers were implemented’ (1997, p. 19). She tests this by running a regression that includes not only the waiver variable but also a lead of this variable, which could not signal program eﬀects. Her results show that the lead variable is almost as large and signiﬁcant as the current waiver variable, which she interprets as indicating that ‘something was changing in these states prior to the implementation of the waivers . . . the waiver programs themselves are not the primary cause of the caseload reductions’ (Ibid.). The fact that policies such as state waivers and OBRA were far less inﬂuenced by interest groups than TANF (being more a function of the internal policy-making of bureaucrats and politicians at the state and federal levels) and were far less eﬀective in reducing the welfare caseloads provides further evidence for the public choice theory of interest groups. The demographic variable, the percentage of births to unwed mothers, was positive and signiﬁcant. The interpretation of this result is somewhat ambiguous, however. To the extent that eligibility criteria limit beneﬁts to needy mothers with children, and that single mothers as a group have a lower income than married mothers, a positive correlation would simply reﬂect the eligibility criteria of the program. However, to the extent that higher beneﬁts provide a greater incentive for unmarried women to have children in order to qualify for ﬁnancial assistance, a positive correlation would reﬂect the causality between the two variables. As discussed earlier in this chapter, Charles Murray and others have long held that higher welfare beneﬁts, and in fact the welfare system in general, provide positive incentives for welfare mothers to have additional children. Given the central importance of this issue to an understanding of changes in welfare caseloads a speciﬁc model is developed in the final part of this chapter to examine the relationship between caseloads and births to unmarried mothers. The only ideology variable to be incorporated in this model was a dummy variable for the party of the Governor of the state. This variable, reﬂecting political ideology, was not signiﬁcant, no doubt in part due to the

186

The political economy of welfare reform in the United States

fact that the legislative outcomes depend not only on the governor but also on the composition of the state legislature. Moreover, party, in itself, is not a precise reﬂection of ideology, since certain southern Democrats may, in fact, be somewhat more conservative than some northern Republicans, and, moreover, during the period in question both parties may have tended to become more centrist. Before leaving the subject of ideology a couple of comments must be made. In order to more fully explore this issue a more accurate proxy for ideology needs to be deﬁned and it may be more fruitful to examine characteristics that would correlate with a conservative or liberal ideology in the electorate rather than in the elected oﬃcials. Such characteristics might be social or demographic: religious aﬃliation or number of churchgoers, or percentage of elderly persons in the state, for example. Since the elderly population is generally more conservative, one could hypothesize that the higher the percentage of elderly persons in the state the more likelihood of both a legislative and social climate which fosters low caseloads. In Rebecca Blank’s study there was a highly signiﬁcant negative correlation between the percentage of elderly persons and caseloads, a result which she does not explain, but which supports this interpretation, and suggests that indeed ‘ideology matters’. In a similar vein, the percentage of births to unmarried mothers may, at least in part, reﬂect social ideology (in the sense of social values) and the acceptability of childbirth outside of marriage. To the extent that the percentage of births to unmarried women reﬂects a more liberal ideology, a positive correlation between percentage of unwed births and caseloads tends to give further support to the recurring theme that ‘ideology matters’. As mentioned above, this relationship is explored later in this chapter.

EXTENSIONS OF THE BASIC MODEL The ﬁnal section of this chapter uses subsets of the data along with some additional data to examine three areas of speciﬁc interest in more depth: the relationship between welfare beneﬁt levels and births to unwed mothers; the inﬂuence of interest groups and ideology on the level of beneﬁts, and the median voter theorem, or more speciﬁcally, the relationship between median income and the level of beneﬁts. The Relationship between Welfare Programs and Births to Unwed Mothers The existing literature on the relationship between welfare programs and births to unwed mothers is inconclusive. Charles Murray, Robert Rector

187

Variables aﬀecting changes in welfare caseloads

and others claim a strong causality from higher beneﬁt levels to higher births and thus caseloads, as discussed in the preceding chapter. However, after completing an extensive literature review and after analyzing time series data over several decades, Robert Moﬃtt concludes that ‘the evidence does not support the hypothesis that the welfare system has been responsible for the time series growth in . . . illegitimacy’ (Moﬃtt, 1992, p. 29). To shed further light on this relationship, a model is developed below, using a subset of the variables for the full model, to examine the relationship between unwed births and cash beneﬁts. The model used data from 50 states over the period 1980 to 1999, and incorporated state and year ﬁxed eﬀects. The speciﬁcation of the model was as follows: NWEDBRTHts C AMB3IN99ts STFIXt YRFIXs Wts

(9.5)

The purpose of this analysis was to test whether, in fact, higher cash beneﬁts provided an incentive for unmarried women to have children in order to increase their income from cash welfare payments, which generally are higher for each additional child. Surprisingly, the regression results showed a signiﬁcant negative correlation between these two variables, suggesting that higher beneﬁt levels, per se, do not increase caseloads (see Table 9.4 below). Whether or not the negative correlation indicates that higher beneﬁts actually lead to a lower percentage of births to unwed mothers is not clear. A higher level of beneﬁts could, theoretically, provide a level of support that enabled unwed mothers to transition to the workforce more easily than those who resided in states with lower beneﬁt levels. The validity of such an interpretation is partially dependent on the importance and eﬀects of missing variables in such a simple model. Among the missing variables that one would expect to inﬂuence unwed births are social or ideological trends that aﬀect the social acceptability of unwed births. To Table 9.4 Variable C AMB3IN99 STFIX YRFIX

Regression results Coeﬃcient 0.235584 9.74E-05*** 0.000941*** 0.008265***

Notes: R-squared 0.566686 Adjusted R-squared0.565381 F-statistic434.1882 Prob. F statistic 0.00000 *** signiﬁcant at 99 percent conﬁdence level.

t-Statistic 35.51529 11.40208 8.465472 29.05039

188

The political economy of welfare reform in the United States

the extent that these trends would be captured by the year ﬁxed eﬀects variable the strong signiﬁcance of this variable indicates that such factors may play a signiﬁcant explanatory role. There is a further technical issue related to the relationships in the basic model, and that is the problem of endogeneity. If the variable representing welfare beneﬁts is correlated with the variable representing births to unwed mothers then ordinary least squares (OLS) estimation may give biased results, since the error terms associated with each of the variables will not be independent. An alternative statistical procedure, two stage least squares (2SLS), in which the beneﬁts variable was replaced by a series of instrument variables (mainly those interest group variables in equation 7), was also run. The results of this procedure were not signiﬁcantly diﬀerent from those using the OLS method. Interest Groups and Ideology The role of interest groups and ideology has been a major focus of this study and is considered of central importance in understanding the legislative outcomes of economic issues such as welfare reform. Unfortunately, it was not possible to obtain data on some of the major interest groups. The amount and types of data collected is limited, and access to that data for nonmembers is often restricted. The Chamber of Commerce, for example, does not make available even such basic data as the number of members in each state. It was, however, possible to obtain data on the number of federally registered corporate PACS in a state, and the expenditures of these organizations, variables which would be expected to correlate with the strength of the Chamber of Commerce and with business interests in general. The per capita expenditures by state was chosen as the most appropriate measure of the strength of business interests in a state. Given the need for additional workers, it is hypothesized that this variable would be negatively correlated with the level of beneﬁts in a state, since lower beneﬁts would provide greater incentives for potential workers to join the labor force. It was also possible to get information on union membership by state, and union membership as a percentage of total oﬀ-farm employment in the state was the variable used to represent the interests of workers. This variable would be hypothesized to be positively correlated with beneﬁt levels, since this would reduce the supply of labor and, ceteris paribus, increase the equilibrium wage. In view of the inﬂuence of the Christian Coalition on the legislative process, a variable was sought to represent this ideological group. The closest proxy available was the percentage of a state’s population aﬃliated with a Christian church. Clearly this is a more inclusive categorization than

Variables aﬀecting changes in welfare caseloads

189

membership in the Christian Coalition, and would thus represent a less homogeneous ideology. For example, members of some of those interest groups most adamantly opposed to the welfare legislation, claiming that it was punitive towards welfare recipients, and most in favor of more generous beneﬁts, would be included in this category, along with the Christian Coalition members who lobbied for the opposite policies. In spite of these diﬀerences however, it seems reasonable to hold that the relative strength of the Christian Coalition in a state would be highly correlated with the percentage of the population aﬃliated with a Christian church, thus rendering the variable a reasonable proxy. A pure interest group model was speciﬁed in which the levels of cash welfare beneﬁts in the states were a function of the strength of union membership, corporate PAC expenditures and the percentage of the state population aﬃliated with a Christian church. Since this last variable was available only for the years 1980 and 1990, it was not possible to do a time series analysis and instead a cross sectional study for the year 1990 was undertaken. A total of 50 observations was used, one for each state. The model was speciﬁed as follows: AMBIN90s C CHRADSs CORPACPCs UMEMPs

(9.6)

where: CHRADS is percentage of state population adherents of a Christian Church CORPACPC is the per capita expenditures of Corporate PACs in the state UMEMP is Union membership as a percentage of employment in the state. The results of running this regression indicated that each of the variables had the expected sign, i.e., corporate PACS and adherents of Christian Churches were associated with lower welfare beneﬁts and union membership was associated with higher beneﬁts, but only union membership was also statistically signiﬁcant. These results are presented in Table 9.5. The Combined Interest Group-Median Voter Model The ﬁnal model to be speciﬁed was one in which the explanatory variables included both interest groups and the median voter as represented by median income. The median voter theorem maintains that the median voter plays a crucial role in the allocation of government expenditures. The income eﬀect would suggest that a higher median income would lead to a preference for higher expenditures, and in the case of welfare programs, a

190

Table 9.5

The political economy of welfare reform in the United States

Regression results

Variable

Coeﬃcient

t-statistic

C CHRADS CORPACPC UMEMP

417.9198 2.179181 7630573 1759.176

3.054342 1.203003 1.276235 4.677968***

Notes: R-squared 0 .423836 Adjusted R-squared 0 .386260 F-statistic11.27948 Prob. F statistic 0 .00001 *** signiﬁcant at 99 percent conﬁdence level.

higher level of beneﬁts. The hypothesis to be tested is therefore, that higher median income in a state is positively correlated with beneﬁt levels. The speciﬁcation of the model is as follows: AMBIN90s C CHRADSs CORPACPCs UMEMPs MDINC$90s

(9.7)

Where the variables are identical to those in equation 9.6, supplemented by the median income variable, MDINC$90. The results of this regression are presented in Table 9.6. The median income variable was indeed positively correlated with the beneﬁt level, and was strongly signiﬁcant, providing supportive evidence for the median income theorem. Moreover, the variables representing both corporate and union interests were statistically signiﬁcant and of the expected sign. As in the previous model, the Christian adherents variable was not signiﬁcant. Two Stages Least Squares Model As discussed earlier, a two stage least squares (2SLS) procedure was used as an alternative to OLS in order to address potential problems of endogeneity in the basic model, and avoid inconsistent regression results. This procedure employs an instrument variable, the value of which is estimated statistically, to replace the values of the original variable contained in the original data set. The ﬁrst stage of the procedure involves using variables exogenous to the system to arrive at an estimated value for the variable. Thus, for instance, if the welfare beneﬁts variable is suspected of being endogenous, it could be replaced by an instrument variable which is estimated by regressing welfare beneﬁts on those public interest variables such as union membership and corporate PAC contributions exogenous to the original model. In the second

Variables aﬀecting changes in welfare caseloads

Table 9.6

191

Regression results

Variable

Coeﬃcient

t-Statistic

C CHRADS CORPACPC UMEMP MDINC$90

258.4054 0.363399 12426105 1044.977 0.023521

1.706461 0.255730 2.730376*** 3.420978*** 6.071193***

Notes: R-squared 0.683269 Adjusted R-squared 0.655116 F-statistic24.26915 Prob. F statistic 0.00000 *** signiﬁcant at 99 percent conﬁdence level.

Table 9.7

Results from 2SLS Model

Variable

Coeﬃcient

t-Statistic

C AMBIN90 NWEDBRTH UNEMPRATE UNEMPRATE1LAG UNEMPRATE2LAGS

345107.0 249.6416 969293.1 21226.7 62345.06 29262.02

2.927388*** 2.367558** 3.303400*** 0.910150 1.536502 1.309630

Notes: R-squared0.287865 Adjusted R-squared 0.206941 F-statistic3.213055 Prob. F statistic 0.014723 *** signiﬁcant at 99 percent conﬁdence level. ** signiﬁcant at 95 percent conﬁdence level.

stage of the procedure an OLS regression is estimated using the estimated value of the variable in place of the original value. Several models were tested, none of which gave signiﬁcantly diﬀerent results from those described above. All the 2SLS models were run on only a subset of the data, since some of the needed exogenous data were available for census years only, i.e. 1980 and 1990, and thus only 1990 data were used. The 2SLS models are thus based on a substantially lower number of observations. Results from one representative model are given in Table 9.7. This model estimates average monthly caseload as a function of unwed births, the unemployment rate with two lags, and an instrument variable estimate of welfare beneﬁts calculated by regressing beneﬁts on median income and the interest group variables, corporate PAC contributions, union membership and adherents of Christian churches.

192

The political economy of welfare reform in the United States

As in the OLS models, caseload is shown to be positively and signiﬁcantly related to unwed births and median income. The relationships with unemployment are less clear. While two out of three are of the anticipated sign, none is statistically signiﬁcant. The overall conclusion from running this and other 2SLS models is that there is little to be gained and the OLS models are robust.

CONCLUSIONS It is evident from the foregoing analyses that the interrelationships between the variables aﬀecting welfare beneﬁts and welfare caseloads are extremely complicated. Moreover, it is undoubtedly true that some variables have been omitted, either because of unavailability or because their eﬀect is unknown, and the measurement of some of the variables is most probably less than 100 percent accurate due to the diﬃculties of compilation or extrapolation. Bearing these caveats in mind, however, it seems clear that the statistical analysis presented above provides substantial support for the public choice argument that the inﬂuence of interest groups on economic policies such as welfare legislation is both substantial and eﬀective. What is also clear is that it is far easier to measure and analyze the eﬀects of economic interest groups, such as unions and corporations, than to measure and analyze the eﬀects of ideological groups, such as the Christian right. A major challenge for the future is to more clearly identify, deﬁne and measure the inﬂuence of such groups.

NOTES 1. The strong interest of the Chamber in welfare reform is illustrated by the contents of a letter, dated January 9, 1995 from Bruce Josten, a senior vice president of the Chamber to Bill Archer, Chairman of the Committee on Ways and Means. In this letter Josten states that ‘Business has a signiﬁcant stake in the welfare reform issue’ and ‘the Chamber is prepared to help craft legislation to reform the nation’s welfare system, and will play a central role in the upcoming welfare reform debate’. Seven months later, on August 7, 1995 Josten sent a letter to all members of the United States Senate noting that ‘For years the Chamber has been a strong proponent of reforming the nation’s welfare system. In fact, in a survey to construct the Chamber’s 1995–1996 National Business Agenda, our members ranked welfare reform as the second highest prority (behind unfunded mandates) on a list of 64 issues’. 2. It is noteworthy that a major component of welfare reform is funding for job skills training, payments to businesses to hire welfare recipients and indirect subsidization of wages by state and local subsidization of transportation to work and child care services. 3. This ignores the deadweight costs of the existing welfare program. To the extent that such costs of the AFDC program on a per capita basis are exceeded by the new program,

Variables aﬀecting changes in welfare caseloads

193

TANF, support is provided for the public choice rather than the public interest approach. 4. The CEA also question the validity of the waiver variable, suggesting that both waivers and caseloads may be related to a third factor, and Martini and Wiseman (1997) of the Urban Institute suggest that caseload declines free up resources which can then be used to apply for waivers.

10.

Conclusions

This book has examined the relevance of two alternative economic models, the public choice model and the public interest model, in explaining the evolution of social welfare policy embodied in the 1996 Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA). This legislation, it was initially felt, would be minimally aﬀected by the pressures of interest groups and the rent seeking activities which are so much a part of the public choice paradigm. In fact, the legislation turned out to attract not only the traditionally active interest groups with primarily pecuniary or political objectives, but also some very inﬂuential ideological groups, that sought to inject an ideological component into the legislation. After describing and critiquing the two economic models in the ﬁrst three chapters of the dissertation, the focus turned to a description and analysis of the history of social welfare policies in Great Britain and the US in chapter four, and a more detailed historical perspective of PRWORA in chapter ﬁve. This historical perspective was highly instructive in that it made clear certain patterns that have recurred throughout the period under consideration. In particular, both the problems and the proposed solutions have been strikingly similar over decades and even centuries, and moreover, in the search for solutions interest groups have always played a central and critical role. The issues that cause frustration and consternation today, ranging from illegitimacy to the diﬃculties of discriminating between the ‘deserving poor’ and the able-bodied, have changed little. Perhaps more surprising is the similarities in proposed solutions; orphanages were recently proposed for children whose parents cannot aﬀord to support them, and the apprenticeship arrangements of old bear some similarities to the job training eﬀorts of recent programs. Overall, sanctions have been less harsh – public ﬂogging has been replaced by elimination of beneﬁts, for example – and there has been a greater focus on assisting the working poor through positive incentives such as the Earned Income Tax Credit rather than through punitive deterrents.1 A second recurring theme, which is discussed in detail in chapter six, is the political involvement of special interest groups seeking to inﬂuence social legislation to achieve their own goals. This is not a new phenomenon; these groups, which include business interests, political interests and ideologically 194

Conclusions

195

motivated groups, have consistently been involved in inﬂuencing legislation in order to further the interests of their own group. Throughout the period under consideration business interests have consistently been involved in welfare legislation. This category includes a diverse array of enterprises ranging from the manufacturers who opposed changing child labor laws in the nineteenth century to the employers supporting workers compensation in the early 1900s and the Chamber of Commerce whose intimate involvement in the 1996 welfare legislation has been documented. In this group also is the medical profession whose involvement extends from opposition to the Shepherd–Towner Act in the nineteenth century to opposition to Medicaid block grants in the twentieth century. Similarly, political interests have played a role throughout, from local authorities seeking funds and positions enabling them to engage in patronage, to President Johnson’s Great Society, which according to Senator Moynihan, was created by an administration looking for popular policies rather than by popular demand, and to Clinton’s attempted reform, and the ﬁnal passage of PRWORA by a Republican Congress that outmaneuvered Clinton and based their proposals on the outcome of political polls. Finally, a broad range of ideological groups, including both conservatives and liberals has been active and eﬀective not only in providing welfare services and beneﬁts, but, particularly in recent times, also in inﬂuencing welfare policies. This category also comprises a very diverse group of institutions, ranging from the Settlement House workers of the nineteenth century ﬁghting for better conditions in the cities, to the Christian Coalition and Heritage Foundation ﬁghting for moral values to be enshrined in the Welfare Law in the twentieth century. The provision of private charity by philanthropic organizations has been a constant feature of the social welfare landscape. As the locus of such eﬀorts moved from the early monasteries in England to the international organizations, such as those aﬃliated with the major denominations, not only has the scope of their services expanded but there has been a vast increase in political lobbying activities and expenditures. Moreover, ideological groups in recent times have held opposing opinions on appropriate social welfare legislation. One reason for this is the growth of evangelical organizations more interested in inﬂuencing moral values than in providing welfare services. The more traditional groups who provided welfare services were generally liberal organizations and often aﬃliated with the Democratic Party. Groups such as those represented by the Christian Coalition are more concerned with promoting moral values and are closely aﬃliated with the Republican Party. The legislation and the welfare debate of the 1990s was clearly vastly more inﬂuenced by the religious right than

196

The political economy of welfare reform in the United States

by the more liberal religious organizations. The bill that was ultimately passed reﬂected the power of the Christian Coalition and aﬃliated groups, power that was derived in large part from their ability to provide a large block of votes and from their close ties to the Republican Party at a time when that party was politically popular, was given a mandate for change and had control of the Congress. Two further comments need to be made regarding the role of ideology in inﬂuencing policies. The ﬁrst point is to note the importance of certain inﬂuential social critics and other writers whose ideas can have a very powerful impact on public opinion, whether instilling fear of the consequences of adopting or failing to adopt certain policy measures, or engendering sympathy for the groups that might be aﬀected by such policies. The eﬀect of such shifts in public opinion is not lost on legislators. In the nineteenth century, for example, the public was very inﬂuenced by writers such as Thomas Malthus who warned of the danger that population growth, when unrestrained by positive checks such as delayed marriage, would be checked by starvation, sickness, infanticide and other such dire consequences. In the same vein, Alexis de Toqueville wrote about a violent revolution being the consequence of any permanent system of poor relief which would ‘benumb human industry and activity’ (Drescher, 1968). A vastly diﬀerent perspective, but one that was also inﬂuential was that of Charles Dickens who generated a certain amount of sympathy for the cause of poor relief. Writers such as Charles Murray and, at the other end of the ideological spectrum, Michael Harrington, not to mention Senator Moynihan, have been extremely inﬂuential in the twentieth century. Murray, in particular not only inﬂuenced public opinion but indirectly inﬂuenced the language of the 1996 welfare law by adding credibility to the message of the Christian Coalition. The second point to note regarding ideology is that existing economic models do not deal well with the issue. A great deal more eﬀort is required in terms of deﬁnition, measurement and analysis of the eﬀect of ideological interests on economic policy outcomes. The econometric analysis in chapter seven of this dissertation provided evidence that the role of interest groups on economic policies such as welfare reform is substantial and eﬀective. It also showed that it is far easier to measure and analyze the eﬀects of economic interest groups, such as unions or corporate political action committees, than of ideological groups such as the Christian Coalition. Such an endeavor could prove to be highly fruitful in terms of more fully understanding the process and outcomes of economic legislation. The pervasiveness of interest groups in the formulating of economic welfare policies is phenomenal. The growth in the number and size of welfare organizations and the expansion of their scope from local to international

Conclusions

197

has been remarkable. Moreover, the shift from private provision to public provision of welfare has also been monumental. And the larger the government pie, the more proﬁtable rent seeking activities will be and the greater the incentives for interest groups to invest in activities aimed at increasing their share. While the involvement of interest groups in the formulating of economic policy is clearly not new, the extent of their inﬂuence, in terms of expenditures and the number and size of such groups, has reached unprecedented proportions. Whether or not responsibility for social programs continues to be decentralized from the federal to the state level of government and whether the current Administration’s focus on faith-based organizations marks a move from public to private provision of welfare remains to be seen. Regardless of whether either of these notions represents the goals of the government, this dissertation has shown that the actual outcome of any legislation is highly dependent on the inputs of interest groups and the interactions of such groups with those responsible for passing the legislation.

NOTE 1. Although some might say that the Earned Income Tax Credit has an antecedent in the Speenhamland system which provided wage supplements to the working poor in England in the late eighteenth century (see chapter six).

Bibliography Albert, Vicky N. (1988), Welfare Dependency and Welfare Policy: A Statistical Study, New York: Greenwood Press. American Public Welfare Association (APWA) (1998), The Race to Welfare Reform: A Recent History, 12 January, http://www.apwa.org/reform/ timeline.htm. Anderson, G. and R.D. Tollinson (1985), ‘Ideology, interest groups and the repeal of the Corn Laws’, Journal of Institutional and Theoretical Economics, 141, June, 199–212. Anderson, G. and R.D. Tollinson (1988), ‘Legislative monopoly and the size of government’, Southern Economic Journal, 54, 592–645. Aranson, P.H. and P.C. Ordeshook (1972), ‘Spatial strategies for sequential elections’, Decision-Making, Columbus: Merrill. Aranson, P.H. and P.C. Ordeshook (1981), ‘Regulation, redistribution, and public choice’, Public Choice, 37 (1), 69–100. Arrow, Kenneth J. (1951), Social Choice and Individual Values, New York: Wiley & Sons. Arrow, Kenneth J. and T. Scitovsky (1969), Reading in Welfare Economics, Homewood, IL: Richard D. Irwin. Bagehot, Walter (2001), Physics and Politics, Kitchener, Ontario: Batoche Books. Barkume, Anthony (1976), ‘Identiﬁcation of preferences for election outcomes from aggregate voting data’, Public Choice, 27 (1), September. Barnow, Burt S. (1988), Estimating the New Jersey AFDC Caseload, Washington, DC: ICF Inc. Barr, James L. and Otto A. Davis (1966), ‘An elementary political and economic theory of the expenditures of local governments’, Southern Economic Journal, 33 (2), October, 149–65. Barzel, Y. and E. Silberberg (1973), ‘Is the act of voting rational?’, Public Choice, 16, Fall, 51–8. Baumol, W.J. (1959), Business Behavior, Value and Growth, New York: Macmillan. Becker, Gary (1983), ‘A theory of competition among pressure groups for political inﬂuence’, Quarterly Journal of Economics, 98, August, 371–400. Bennett, William and Peter Wehner (1994a), ‘Shifting targets in the welfare trenches’, Washington Times, 25 January. 198

Bibliography

199

Bennett, William and Peter Wehner (1994b), ‘Welfare reform: Is it right? Let’s end welfare for anyone with out of wedlock kids’, The Arizona Republic, Phoenix, Arizona, 29 April. Bennett, William and Peter Wehner (1995), ‘Competing themes in the welfare debate’, Washington Times, 3 August. Bennett, William, Jack Kemp and Vin Weber (1994), ‘Memorandum from Empower America Co-Directors to Congressional Republicans’, 13 April. Bergson, Abram (1938), ‘A reformulation of certain aspects of welfare economics’, Quarterly Jounal of Economics, 52 (2), February, 310–34. Bergstrom, T.C. and R.P. Goodman (1973), ‘Private demands for public goods’, American Economic Review, June, 63 (3), 280–96. Beveridge, William H. (1942), ‘Social Insurance and Allied Services (The Beveridge Report)’, CMND 6404, November, presented to Parliament, London: HMSO. Bjorklund, A. and B. Holmlund (1990), ‘Unemployment policy: lessons from Sweden’, Economic Review, January. Black, D. (1948), ‘On the rationale of group decision making’, Journal of Political Economy, 56, February, 23–34, reprinted in Arrow and Scitovsky (1969), pp. 133–46. Black, D. (1958), The Theory of Committees and Elections, Cambridge: Cambridge University Press. Blair, T. (1998), The Third Way, London: Fabian Society. Blank, Rebecca (1997), ‘What Causes Public Assistance Caseloads to Grow?’, Northwestern University, May. Brazell, Jan F., Irving Lefbergh and Wolfgang Opitz (1989), ‘The impact of population size and the economy on welfare caseloads: the special case of welfare reform’, Applied Demography, 4 (3), Summer, 1–7. Bremner, Robert (1960), American Philanthropy, Chicago: University of Chicago Press. Buchanan, J.M. (1949), ‘The pure theory of government ﬁnance: a suggested approach’, Journal of Political Economy, 57, December, 496–506. Buchanan, J.M. and G. Tullock (1962), The Calculus of Consent, Ann Arbor: University of Michigan Press. Burke, Vee (1995a), Family Cash Welfare: Comparison of House-Passed and Senate Finance Committee Versions of H.R. $., CRS Report for Congress 95–691 EPW, Washington, DC: Library of Congress, Congressional Research Service. Burke, Vee (1995b), Welfare Reform, CRS Issue Brief IB 93034, 6 March, Washington, DC: Congressional Research Service. Burke, Vee (1996), Family Cash Welfare: Comparison of H.R. 4, H.R. 3507/S. 1795, and S.1841/H.R. 3612 (Clinton Bill), Congressional Research Service, 18 June.

200

The political economy of welfare reform in the United States

Burke, Vincent J. and Vee Burke (1974), Nixon’s Good Deed, New York and London: Columbia University Press. Butler, Stuart and Anna Kondratis (1987), Out of the Poverty Trap, New York: The Free Press. Cable News Network (CNN) (1999), ‘Report on CNN.Com’, 26 October. Center on Budget and Policy Priorities (CBPP) (1995), ‘The Initial Impacts of Welfare Reform on the Economic Well-Being of Single Mother Families’, Washington, DC. Children’s Defense Fund (1998), Annual Report, Washington, DC. Clarke, J., M. Langan and F. Williams (2001), ‘Remaking Welfare: the British Welfare regime in the 1980s and 1990s’, in A. Cochrane, J. Clarke and S. Gewirtz, Comparing Welfare States, London: Sage Publications. Clinton, William Jeﬀerson (1991a), Announcement Speech, Little Rock, Arkansas, 3 October. Clinton, William Jeﬀerson (1991b), The New Covenant: Responsibility and Rebuilding the American Community, Georgetown University, Washington, DC, 23 October. Clinton Campaign Economic Plan (1992), Clinton Campaign, Little Rock, Arkansas. Coase, R.H. (1937), ‘The nature of the ﬁrm’, Economica, 4, 386–405. Coase, R.H. (1960), ‘The problem of social cost’, Journal of Law and Economics, 3, 1–44. Cochrane, A.D., J. Clarke and S. Gewirtz (eds) (2001), Comparing Welfare States, London: Sage Publications. Committee on Ways and Means (1996), 1996 Green Book, US House of Representatives, Washington, DC: Government Printing Oﬃce. Congressional Budget Oﬃce (CBO) (1977), Troubled Local Economies and the Distribution of Federal Dollars, Washington, DC. Congressional Budget Oﬃce (CBO) (1993), Forecasting AFDC Caseloads with an Emphasis on Economic Factors, CBO Staﬀ Memorandum, July, Washington, DC: CBO. Congressional Quarterly (1994), ‘Weekly Report’, 24 September, Washington, DC: Congressional Quarterly, Inc. Congressional Quarterly (1995a), Almanac, Washington, DC: Congressional Quarterly, Inc. Congressional Quarterly (1995b), ‘Weekly Report’, 5 August, Washington, DC: Congressional Quarterly, Inc. Congressional Research Service (CRS) (2003), Welfare Reform: With TANF Flexibility, States Vary in How They Implement Work Requirements and Time Limits, IB 93034 25 March. Congressional Research Service (CRS) (2005), Issue Brief for Congress, Welfare Reauthorization: Overview of the Issues, updated 31 March.

Bibliography

201

Council of Economic Advisors (CEA) (1997), ‘Technical Report: Explaining the Decline in Welfare Receipt 1993–1996’, April, Washington, DC: Executive Oﬃce of the President. Council of Economic Advisors (CEA) (1999), ‘Economic Expansion, Welfare Reform, and the Decline in Welfare Caseloads: An Update. Technical Report’, September, Washington, DC: Executive Oﬃce of the President. Craig, F.W.S. (1975), British General Election Manifestos, Basingstoke: Macmillian. Crain, W.M. and R. Tollison (1976), ‘Campaign expenditures and political competition’, Journal of Law and Economics, 19, April, 177–88. Danielson, Albert L. and Paul H. Rubin (1977), ‘An empirical investigation of voting on energy issues’, Public Choice, 31, Fall, 121–8. Danziger, Sheldon H., Gary Sandefur and Daniel H. Weinberg (1994), Confronting Poverty, New York: Russel Sage Foundation. Davis, O.A., M. Hinich and P. Ordeshook (1970), ‘An expository development of a mathematical model of the electoral process’, American Political Science Review, 64, June, 426–48. de Tocqueville, Alexis ([1840] 1969), Democracy in America, New York: Harper and Rowe. de Tocqueville, A. (1957), Democracy in America, New York: Vintage Books. Delors, J. (1985), ‘Preface’, in J. Vandamme (ed.), New Dimensions in European Social Policy, London: Croom Helm, pp. ix–xx. Denzau, Arthur and Kevin Grier (1984), ‘Determinants of local school spending: some consistent estimates’, Public Choice, 44, August, 375–83. Dobelstein, Andrew W. (1980), Politics, Economics and Public Welfare, Englewood Cliﬀs, NJ: Prentice-Hall. Dougan, William R. and Michael C. Munger (1989), ‘The rationality of ideology’, Journal of Law and Economics, XXXII, April, 119–41. Downs, A. (1957), An Economic Theory of Democracy, New York: Harper and Row. Drescher, Seymour (ed.) (1968), Tocqueville and Beaumont on Social Reform, New York: Harper and Row. Drew, Elizabeth (1996), Showdown: The Struggle Between the Gingrich Congress and the Clinton White House, New York: Simon and Schuster. Elman, R. (1966), The Poorhouse State, New York: Pantheon Books. Enelow J.M., and M.J. Hinich (1984), The Spatial Theory of Voting, Cambridge: Cambridge University Press. Fiorina, M.P. (1977a), Congress: Keystone of the Washington Establishment, New Haven: Yale University Press. Fiorina, M.P. (1977b), ‘The case of the vanishing marginals: the bureaucracy did it’, American Political Science Review, 71 (1), 177–81.

202

The political economy of welfare reform in the United States

Fiorina, M. (1980), ‘The decline of collective responsibility in American politics’, Daedalus, Summer, 25–45. Fiorina, M. (1981), Retrospective Voting in American National Elections, New Haven: Yale University Press. Fiorina, M. (1996), Divided Government, Boston, Toronto: Allyn and Bacon. Friedman, Lawrence (1977), ‘The social and political context of the war on poverty: an overview’, in Robert H. Haveman (ed.), A Decade of Federal Anti-Poverty Programs, New York: Academic Press, pp. 21–47. Frum, David (1994), Dead Right, New York: Basic Books. Gabe, Thomas (1992), ‘Demographic Trends Aﬀecting Aid to Families With Dependent Children (AFDC) Caseload Growth’, CRS Report for Congress, Washington, DC: The Congressional Research Service, December. Garasky, Steven (1989), ‘Analyzing the Eﬀect of Massachusetts’ ET Choices Program on the State’s AFDC Caseload’, Technical Analysis Paper No. 39, Washington, DC: Department of Health and Human Services, Oﬃce of Income Security Policy, June. Garasky, Steven (1990), ‘Analyzing the Eﬀect of Massachusetts’ ET Choices Program on the State’s AFDC-Basic Caseload’, Evaluation Review, December, 701–10. General Accounting Oﬃce (GAO) (2002), Welfare Reform With TANF ﬂexibility States Vary in How They Implement Work Requirements and Time Limits, July. Ginsburg, N. (1992), Divisions of Welfare, London: Sage Publications. Ginsburg, N. (2001), ‘Sweden: the social democratic case’, in A.D. Cochrane, J. Clarke and S. Gewirtz (eds), Comparing Welfare States, London: Sage Publications. Gist, J.R. and R.C. Hill (1981), ‘The economics of choice in the allocation of federal grants: an empirical test’, Public Choice, 36 (1), 63–73. Goﬀ, B.L. and K.B. Grier (1993), ‘On the (mis)measurement of legislator ideology and shirking’, Public Choice, 76 (1), 5–20. GOP (1994), ‘GOP accentuates the positive: hopefuls to sign contract’, Congressional Quarterly, 27 (11), 24 September. Grier, Kevin (1993), ‘Introduction: ideology and representation in American politics’, Public Choice, 76, June. Grossman, Jean Baldwin (1985), ‘The Technical Report for the AFDC Forecasting Project for the Social Security Administration/Oﬃce of Family Assistance’, MPR Reference No. 7502–954, Mathematica Policy Research, Washington, DC, February. Hagan, Jan and Irene Lurie (1992), Implementing Jobs: Initial State Choices, Albany: The Nelson A. Rockefeller Institute of Government at the State University of New York.

Bibliography

203

Hantrais, L. (2000), Social Policy in the European Union, New York: St Martin’s Press. Hart, H.A. (1973), ‘Rawls on Liberty and its Priority’, University of Chicago Law Review, 40, Spring, 534–55. Harrington, Michael (1964), The Other America: Poverty in the United States, New York: Macmillan. Haskins, R. (1999), ‘Welfare in a Society of Permanent War’, Ways and Means Committee, House of Representatives, Washington, DC: US Government Printing Oﬃce. Haveman, Robert (ed.) (1977), A Decade of Federal Anti-Poverty Programs, New York: Academic Press. Heclo, H. and H. Madsen (1987), Policy and Politics in Sweden, Philadelphia: Temple University Press, cited in N. Ginsburg (1992), Divisions of Welfare, London: Sage Publications. Himmelfarb, Gertrude (1984), The Idea of Poverty: England in the Industrial Age, New York: Alfred Knopf. Hinich, M.J. (1978), The Mean Versus the median in Spatial Voting Games, Working Paper, Department of Economics, Virginia Polytechnic Institute. Hinich, M.J. and P.C. Ordeshook (1970), ‘Plurality maximisation vs. vote maximisation: a spatial analysis with variable participation’, American Political Science Review, 64, September, 772–91. Hobson, B. and M. Takahashi (1997), ‘The parent worker model: lone mothers in Sweden’, in J. Lewis (ed.), Lone Mothers in European Welfare Regimes, London: Jessica Kingsley. Hochman, H.M. and J.D. Rodgers (1969), ‘Pareto optimal redistribution’, American Economic Review, 59, September, 542–57. Holcombe, R. (1989), ‘The median voter model in public choice theory’, Public Choice, 61 (2), May. Holcombe, Randall and Asgher Zardkoohi (1981), ‘The determinants of federal grants’, Southern Economic Journal, October, 399–420. Hotelling, H. (1929), ‘Stability in competition’, Economic Journal, 39, March, 41–57. The House Wednesday Group (1991), ‘Moving Ahead: Expanding Opportunity in America’, October. Hrebenar, R. (1997), Interest Group Politics in America, Armonk, NY: M.E. Sharpe. Inman, R.P. (1978), ‘Testing political economy’s as if proposition: is the median income voter really decisive?’, Public Choice, 33 (4), 45–65. Isikoﬀ, M. (1992), Uncovering Clinton, A Reporter’s Story, London: Crown Publishers. Joseph, K. and J. Sumption (1979), Equality, London: John Murray.

204

The political economy of welfare reform in the United States

Josten, Bruce (1995), ‘Letter to the Hon. Bill Archer, Chairman of the Ways & Means Committee’, House of Representatives, 9 January. Kalt, Joseph P. and Mark A. Zupan (1984), ‘Capture and ideology in the economic theory of politics’, American Economic Review, June, 279–300. Kammer, Fred S.J., President, Catholic Charities (1996), ‘Testimony Regarding the Mutual Roles of Government and Churches in Providing Basic Human Services for Poor and Vulnerable Groups’, presented to the Senate Labor Committee, Subcommittee on Children and Families, 26 March. Kau, James B. and Paul H. Rubin (1979), ‘Self interest, ideology and logrolling in congressional voting’, Journal of Law and Economics, October, 365–84. Kemp, Jack (1993), ‘Breakdown of the family is a social cost of welfare’, Daily Californian, 26 November. Kincaid, J. (1994), ‘Developments in federal-state relations, 1992–1993’, The Book of the States: 1994–1995, Lexington, KY: Council of State Governments. Kmenta, Jan (1971), Elements of Econometrics, New York: Macmillan Publishing. Labour Party (1974), Britain Will Win With Labour, Labour Party, London. Landes, W.M. and R.A. Posner (1975), ‘The independent judiciary in an interest-group perspective’, Journal of Law and Economics, 18, December, 875–901. Larson, Jan (1992), ‘Dad’s a Deadbeat’, American Demographics, July. Latham, Earl (1952), The Group Basis of Politics, Ithaca: Cornell University Press. Leibfried, S. (1979), ‘Public Assistance in the Federal Republic of Germany’, in M. Partington and J. Jowell (eds), Welfare Law and Policy, London: Frances Pinter. Lesher, R.L. (1995), ‘The welfare trap’, in The Voice of Business, 20 March, US Chamber of Commerce. Lewis, G., S. Gewirtz and J. Clarke (2000), Rethinking Social Policy, London: Sage Publications. Lund, B. (2002), Understanding State Welfare, London: Sage Publications. Lyon, David L., Mark Menchik and Gerard Blais (1976), Predicting the New York City Welfare Caseload, December, Santa Monica, CA: Rand Corporation. Malthus, Thomas (1960), On Population, (ed. Gertrude Himmelfarb), New York. Mandelson, P. and R. Liddle (1996), The Blair Revolution: Can New Labour Deliver?, London: Faber and Faber.

Bibliography

205

Martini, Alberto and Michael Wiseman (1997), ‘Explaining the recent decline in welfare caseloads: is the council of economic advisors right?’, Urban Institute, July, Washington, DC. Maximus (1998), Annual Report, McLean, VA: Maximus Inc. Maximus (1999), Annual Report, McLean, VA: Maximus Inc. Mayhew, D.R. (1974), Congress: The Electoral Connection, New Haven: Yale University Press. McKelvey, Richard D. (1979), ‘General conditions for global intransitivities in formal voting models’, Conometrica, 47 (5), September, 1085–112. Mill, J.S. (1994), in William I. Trattner, From Poor Law to Welfare State, New York and Toronto: The Free Press. Mishan, E.J. (1971), ‘The postwar literature on externalities: an interpretative essay’, Journal of Economic Literature, 9, March, 1–28. Moe, T.M. (1987), ‘An assessment of the positive theory of Congressional dominance’, Legislative Studies Quarterly, November, 475–520, reprinted in Charles K. Rowley (ed.) (1993), Public Choice Theory, vol. II, Aldershot, UK and Brookﬁeld, US: Edward Elgar Publishing. Moﬃtt, Robert (1992), ‘Incentive eﬀects of the U.S. welfare system: a review’, Journal of Economic Literature, 30 (1), March, 1–61. Moynihan, Daniel Patrick (1969), Maximum Feasible Misunderstanding, New York: Macmillan-Free Press. Moynihan, Daniel Patrick (1986), Family and Nation, New York: Harcourt Brace Janovich. Moynihan, Daniel Patrick (1992), ‘How the great society “destroyed the American family” ’, Public Interest, 108, Summer, 53–64. Moynihan, Daniel Patrick (1996), Miles To Go, Cambridge, MA: Harvard University Press. Mueller, Dennis (1989), Public Choice, Cambridge: Cambridge University Press. Mueller, Dennis (1990), ‘Electoral politics, interest groups, and the size of government’, Economic Inquiry, 28 (4), 682–705. Mueller, Dennis (2003), Public Choice III, Cambridge, UK and New York: Cambridge University Press. Murray, Charles (1984), Losing Ground, New York: Basic Books. Murray, Charles (1993), ‘The coming white underclass’, Wall Street Journal, Friday 29 October. Myrdal, A. (1945), Nation and Family: The Swedish Experiment in Democratic family and Population Policy, London: Kegan Paul. Nash, Gary (1976), ‘Urban wealth and poverty in pre-revolutionary America’, Journal of Interdisciplinary History, 6, Spring. Nelson, Douglas and Eugene Silberberg (1985), ‘Ideology and legislator shirking’, Economic Inquiry, 25 (1), 15–25.

206

The political economy of welfare reform in the United States

Niskanen, William A. (1971), Bureaucracy and Representative Government, Chicago: Aldine Atherton. Niskanen, William A. (1975), ‘Bureaucrats and politicians’, Journal of Law and Economics, 18, December, 617–43. Nozick, R. (1974), Anarchy, State, and Utopia, New York: Basic Books. Oates, W.E. (1988), ‘On the nature and measurement of ﬁscal illusion: a survey’, in G. Brennan et al. (eds), Taxation and Fiscal Federalism: Essays in Honour of Russell Mathews, Canberra: Australian National University Press, pp. 65–82. Olson, Mancur (1965), The Logic of Collective Action, Cambridge, MA and London, UK: Harvard University Press. O’Neill, June (1990), Work and Welfare in Massachusetts, Boston: Pioneer Institute for Public Policy Research. Ozawa, Martha (1982), Income Maintenance and Work Incentives, New York: Praeger. Peacock, Alan T. and Charles K. Rowley (1972), ‘Pareto optimality and the political economy of liberalism’, Journal of Political Economy, 80 (3), May/June, 876–90. Peltzman, Sam (1976), ‘Toward a more general theory of regulation’, Journal of Law and Economics, 19 (2), Conference on the Economics of Politics and Regulation, 211–40. Peltzman, Sam (1982), ‘Peltzman on regulation and politics: reply’, Public Choice, 39 (2), 299–300. Peltzman, Sam (1984), ‘Constituent interest and congressional voting’, Journal of Law and Economics, April, 181–210. Peltzman, Sam (1990), ‘How eﬃcient is the voting market?’, Journal of Law and Economics, 33 (1), 27–63. Peters, Charles (1980), How Washington Really Works, Reading MA, Menlo Park CA, London, Amsterdam, Don Mills, Ontario, Sydney: Addison Wesley. Pigou, A.C. (1920), The Economics of Welfare, London: Macmillan, (revised 1924, 1929, 1932). Plotnick, Robert D. and Russell M. Lidman (1987), ‘Forecasting welfare caseloads: a tool to improve budgeting’, Public Budgeting and Finance, Autumn, 70–81. Pommerehne, W. (1978), ‘Institutional approaches to public expenditure: empircial evidence from Swiss municipalities’, Journal of Public Economics, 9, 255–80. Pommerehne, W.W. and B.S. Frey (1976), ‘Approaches to estimating publicexpenditures’, Public Finance Quarterly, 4, 395–407. Pommerehne, Werner W. and Friedrich Schneider (1978), ‘Fiscal illusion, political institutions, and local public spending’, Kyklos, 31 (3), 381–408.

Bibliography

207

Pomper, Gerald, Walter Dean Burnham, Anthony Corrado, Marjorie Ramdon Hershey, Marion R. Just, Scott Keeter, Wilson Carey McWilliams and William G. Mayer (1996), The Election of 1996: Reports and Interpretations, New Jersey: Chatham House. Poole, L. (2000), ‘New approaches to competitive social policy: the changing face of control and Eastern European welfare’, in G. Lewis, S. Gewirtz and J. Clarke (eds), Rethinking Social Policy, London: Sage Publications. Rawls, J.A. (1971), A Theory of Justice, Cambridge, MA: Belknap Press. Reed, Ralph Jr. (1997), ‘Statement by Ralph Reed Jr. Concerning His Resignation from the Christian Coalition’, Christian Coalition Press Release, 23 April, quoted in J. Watson (1997), p. 54. Riker, W.H. and P.C. Ordeshook (1968), ‘A theory of the calculus of voting’, American Political Science Review, 62, March, 25–42. Romer, T. and H. Rosenthal (1978), ‘Political resource allocation, controlled agendas and the status quo’, Public Choice, 33 (4), Winter, 27–43. Romer, T. and H. Rosenthal (1979a), ‘The elusive median voter’, Journal of Public Economics, 12, October, 143–70. Romer, T. and H. Rosenthal (1979b), ‘Bureaucrats versus voters: on the political economy of resource allocation by direct democracy’, Quarterly Journal of Economics, 93, November, 563–87. Romer, T. and H. Rosenthal (1982), ‘Median voters or budget maximisers: evidence for school expenditure referenda’, Economic Inquiry, 20, October, 556–78. Rowley, Charles K. (1986), ‘The relevance of the median voter theorem’, Journal of Institutional and Theoretical Economics, 140, 106–26. Rowley, Charles K. and Alan T. Peacock (1972), ‘Pareto optimality and the political economy of liberalism’, Journal of Political Economy, 80 (3), May/June, 876–90. Rowley, Charles K. and A.T. Peacock (1975), Welfare Economics, London: Martin Robertson. Rowley, C.K., W. Thorbecke and R. Wagner (1995), Trade Protection in the United States, Aldershot, UK and Brookﬁeld, USA: Edward Elgar. Rubenfeld, Daniel L. (1977), ‘Voting in a local school election: a micro analysis’, The Review of Economics and Statistics, 59 (1), 30–42. Salaman, Lester and Alan Abramson (1982), The Federal Budget and the Nonproﬁt Sector, Washington, DC: Urban Institute. Samuelson, P.A. (1947), Foundations of Economic Analysis, Cambridge, MA: Harvard University Press. Samuelson, P.A. (1954), ‘The pure theory of public expenditure’, Review of Economics and Statistics, 36, November, 386–9, reprinted in Arrow and Skitovsky (1969), pp. 179–8.

208

The political economy of welfare reform in the United States

Sanford, T. (1967), Storm Over the States, New York: McGraw-Hill. Scott, H. (1982), Sweden’s ‘Right to be Human’, London: Allison and Busby. Scholl, D. and J.E. Stapleford (1991), ‘Forecasting State Welfare Caseloads: Alternative Methodologies’, unpublished study for Delaware Office of the Budget and University of Delaware. Senate Finance Committee (2002), Summary of Hearings on Issues in TANF Reauthorization: Building Stronger Families, Thursday 16 May. Shepsle, K.A. and B.R. Weingast (1981), ‘Structure-induced equilibrium and legislative choice’, Public Choice, 37 (3), 503–19. Shepsle, Kenneth A. and Barry R. Weingast (1987), ‘The institutional foundations of committee power’, American Political Science Review, 81, 85–104. Silberberg, Eugene (1990), The Structure of Economics: A Mathematical Analysis, London, New York, Tokyo: McGraw-Hill. Silberman, Jonathon I. and Garey C. Durden (1976), ‘Determining legislative preferences on the minimum wage: an economic approach’, Journal of Political Economy, 84, 317. Smith, A. (1759), The Theory of Moral Sentiments, London: A. Miller. Smith, A. (1776), An Inquiry into the Nature and Causes of the Wealth of Nations, reprinted in W.B. Todd (ed.) (1976), Glasgow Edition of the Works and Correspondence of Adam Smith, vol. I, Oxford: Oxford University Press. Smith, T. (1998), ‘When States Lobby’, Unpublished Ph.D. dissertation, State University of New York. Smithies, A. (1941), ‘Optimum location in spatial competition’, Journal of Political Economy, 49, June, 423–39. Stephanopoulos, George (1999), All Too Human, Boston, New York, London: Little, Brown and Company. Stigler, G.J. (1971), ‘The theory of economic regulation’, Bell Journal of Economics and Management Science, 2, Spring, 3–21. Stigler, G.J. (1972), ‘Economic competition and political competition’, Public Choice, 13 (1), 91–106. Stigler, G.J. (1973), ‘General economic conditions and national elections’, American Economic Review, 63 (160). Stigler, G.J. (1974), ‘Free riders and collective action: an appendix to theories of economic regulation’, Bell Journal of Economics, 5 (2), Autumn, 359–65. Stiglitz, Joseph E. (1986), Economics of the Public Sector, New York and London: Princeton University. Thurber, James (1996), Rivals for Power – Presidential-Congressional Relations, Washington, DC: Congressional Quarterly Press.

Bibliography

209

Trattner, William I. (1994), From Poor Law to Welfare State, New York and Toronto: The Free Press. Tullock, G. (1959), ‘Some problems of majority voting’, Journal of Political Economy, 67, December, 571–9, reprinted in Arrow and Scitovsky (1969), pp. 169–78. Tullock, G. (1967), ‘The welfare costs of tariﬀs, monopolies and theft’, Western Economic Journal, 5, June, 224–32. Tullock, G. (1970), Private Wants, Public Means, New York: Basic Books. Tullock, G. (1982), ‘Why so much stability?’, Public Choice, 38 (2), 189–202. US Chamber of Commerce (1997), 1996 How They Voted, Washington, DC: US Chamber of Commerce. US Chamber of Commerce (1999), 1999–2000 National Business Agenda, Washington, DC: US Chamber of Commerce. US General Accounting Oﬃce (1985), An Evaluation of the 1981 AFDC Changes: Final Report, Washington, DC: US GAO, July. Wallis, J.J. (1986), ‘Laws and legislatures’, Mimeo, College Park, University of Maryland, cited in D. Mueller (1989), Public Choice, Cambridge: Cambridge University Press, p. 143. Watson, Justin (1997), The Christian Coalition: Dreams of Restoration, Demands for Recognition, New York: St. Martin’s Press. Weber, M. (1947), The Theory of Social and Economic Organisation, New York, NY: The Free Press. Wedel, Janet M. (1987), ‘An Explanatory Model for Forecasting the New York City Public Assistance Family Caseload’, in Department of Health and Human Services, Social Security Administration, Proceedings: 27th National Workshop on Welfare Research and Statistics, pp. 605–17. Weingast, B. (1981), ‘Regulation, reregulation, and deregulation: the political foundations of agency-clientele relationships’, Law and Contemporary Problems, 44, Winter, 147–77, reprinted in Charles K. Rowley (ed.) (1993), Public Choice Theory, vol. III, Aldershot, UK and Brookﬁeld, US: Edward Elgar Publishing. Weingast, B. and W.J. Marshall (1988), ‘The industrial organization of Congress; or, why legislatures, like ﬁrms, are not organized as markets’, The Journal of Political Economy, 96 (1), 132–63. Weingast, B.R. and M.J. Moran (1983), ‘Bureaucratic discretion or Congressional control? regulatory policymaking by the federal trade commission’, Journal of Political Economy, 91, October, 765–800. Wicksell, K. (1896), ‘A new principle of just taxation’, in R.A. Musgrave and A.T. Peacock (eds) (1967), Classics in the Theory of Public Finance, New York: Macmillan, pp. 72–118.

210

The political economy of welfare reform in the United States

Williamson, O.E. (1964), The Economics of Discretionary Behavior, Englewood Cliﬀs, NJ: Prentice Hall. Wilson, E. (1977), Women and the Welfare State, London: Tavistock. Wilson, J.Q. (1989), Bureaucracy: What Government Agencies Do and Why They Do It, New York: Basic Books. Wiseman, Michael and Alberto Martini (1997), Explaining the Recent Decline in Welfare Caseloads: Is the Council of Economic Advisors Right?, Washington, DC: Urban Institute, July. Woodward, Bob (1994), The Agenda, New York, London, Toronto, Sydney, Tokyo, Singapore: Simon and Schuster. Ziliak, James P., David N. Figlio, Elizabeth E. Davis and Laura S. Connolly (1997), ‘Accounting for the Decline in AFDC Caseloads: Welfare Reform or Economic Growth?’, Discussion Paper No. 1151–97, Institute for Research on Poverty, University of Wisconsin-Madison.

Index Abramson, Alan 110 abstinence education 4 ageing populations 31, 32 see also elderly people Aid to Families with Dependent Children (AFDC) 2, 7, 8, 73, 106, 115, 116, 119–20, 124, 125–6, 129, 130, 132, 141, 142, 154, 169, 171, 178–9, 180 see also welfare caseloads, changes in AFDC and TANF programs 181 AFDC-Basic 169, 179 AFDC-Unemployed Parent (AFDC-UP) 169, 171, 179 historical spending on 166 ratio of recipients to population 172 allocative eﬃciency or redistribution 48–51 America: early years (to 20th century) 96–103 American Revolution 98 Boston Irish Society 97 child labor laws 103 Declaration of Independence 98 Enlightenment, The 98 federal aid 102 Germany Society of New York 97 increasing poverty 98–9 institutions for the poor, growth in 102 outdoor relief, problems with 101 Quakers 97 religious revivals and the Great Awakening 98 Sanitary Commission 102 Scots Charitable Society 97 Settlement House movement 103, 104 Yates Report on poor relief 101–2 America in the early 20th century 103–8 see also American legislation

Children’s Bureau 77, 105 Conference on Dependent Children (1909) 104 Freedman’s Bureau 103–4 New Deal legislation 107–8 public aid, growth in 103–5 years following Great Depression (1929) 106–7 America in mid- to late 20th century 108–17 child support enforcement measures 117 civil rights 108, 111 Earned Income Tax Credit (EITC) 113 Family Assistance Plan 113–15 federal grants 111 Food Stamp Program 112 Great Society program 108–9, 111 Job Opportunities and Basic Skills (JOBS) program 117 New Frontier program 108 Nixon administration 112–16 old age assistance 112–13 poverty issues/programs 108–12 presidential politics 108–9 public works programs 107 social security beneﬁts 112 Supplemental Security Program 113 voting rights 111 American Association for Retired Persons (AARP) 64, 74, 78 American Federation of Labor and Congress of Industrial Organizations (AFL-CIO) 115 American Federation of State, County and Municipal Employees (AFSCME) 76, 150 American legislation 110–12, 131 Americans with Disabilities Act (1990) 117 Area Redevelopment Act (1961) 110 211

212

The political economy of welfare reform in the United States

Balanced Budget Act (1997) 7, 152 Balanced Budget Constitutional Amendment 153 Budget Reconciliation/Welfare Reform Bill 130 Charitable Choice Act 9–11 Charity Aid, Recovery and Empowerment Act (CARE) 9 Deﬁcit Reduction Act (DRA, 2005) 3, 4, 5 Economic Development Act (1965) 110 Economic Opportunity Act (EOA) 110 Family Assistance Plan (1970) 113–15 Family Support Act (1988) 3, 117, 155 Federal Emergency Relief Act (1933) 107 Finance Committee Bill (1995) 8–9 Food Stamp Act (1964) 110 Manpower Development and Training Act (1962) 110 New Deal 107–8 Omnibus Budget Reconciliation Act (OBRA, 1981) 181, 184–5 Personal Responsibility Act 122, 123–4, 140 Personal Responsibility and Work Opportunity Reconciliation Act (1996) 2, 7–8, 12, 30, 52, 70, 80, 194–5 see also main entry Real Welfare Act (1994) 158 Sheppard–Towner Act 77, 105–6, 195 Social Security Act, Amendments to (1962) 110 Social Security Act (1935) 106, 107, 110 Social Security Act: Medicare and Medicaid amendments (1965) 110 State Unemployment Relief Act (Wicks Act) 107 Talent–Fairclough Bill 158, 160 Welfare Reform Act (1994) 119 Work and Responsibility Act (1994): the Clinton Bill 119–21, 123–4 Workforce Investment Act 9

American Medical Association (AMA) 64–5 American Public Human Services Association (APHSA) 147–8, 150 and Miller, Gerald 148 American Public Welfare Association (APWA) 119 Americans for Democratic Action (ADA) 71 Anderson, G. 71 Aranson, P.H. 48–9 Arrow, Kenneth J. 53 Ashcroft, Senator 8 ‘Assessment of the Positive Theory of Congressional Dominance, An’ 84 Bagehot, Walter 100 Barkume, Anthony 59 Barr, James L. 58 Barzel, Y. 61 Bateman, Worth 113–14 Baucus, Senator Max 4–5 Baumol, W.J. 84 Bayh, Senator Birch 78 Becker, Gary 52, 63, 65–7, 74, 76–7 Becker model/theory 66–7 Bennett, William 160, 161 Bentson, Lloyd 137 Beveridge, William H. 20 Bergson, Abram 53 Bergstrom, T.C. 58 Beveridge Report 20 Bjorklund, A. 15 Black, D. 52, 53–4 Blair, Tony 21, 22, 23 Blank, Rebecca 173, 175, 176–9, 180, 181, 182, 184 Bread for the World 155 Bremner, Robert 98 Britain: development of welfare programs in 93–6, 98–9, 194, 197 as basis for poor laws in American colonies 96 Henrician Poor Law 95 laissez-faire 99 Poor Law (1601) 95–6, 99–100 Poor Law Commission 101

Index public relief, concerns about 100–101 Punishment of Sturdy Vagabonds and Beggars Act 95 Speenhamland system (1795) 99, 197 Statute of Labourers 94 Britain (and) 13, 14, 19–28, 31, 32 see also British legislation Conservative Government 20, 21 Labour Government 20–21 National Assistance Board 20 New Deal 21 New Labour 21, 23 welfare state 20–22 see also British welfare state/legislation British legislation 20 Education Act (1944) 20 Family Allowances Act (1945) 20 Industrial Charter (1947) 20 National Assistance Act (1948) 20 National Minimum Wage Act (1998) 23 Social Security Act (1986) 22 British welfare state/legislation 20–22 child support 22 Family Income Supplement/Family Credit 21, 22, 23 Job Creation Program 21 Social Exclusion Oﬃce 23 Work Experience Program 21 Working Families Tax Credit 23 Youth Opportunities Program 21 Buchanan, J.M. 52 Burke, Vee 113, 114, 115, 116, 121, 124, 130 Burke, Vincent J. 113, 114, 115, 116 Burns, Arthur 113, 114 Bush, President 3, 9 and faith-based initiative 9 Bush, President George 117 Butler, Stuart 95, 109–10, 111–12 Calvinism 98 Carter, President Jimmy 92, 116 ‘Case of the Vanishing Marginals, The’ 88 Catholic Charities 150, 154–5, 164 and ‘Contract with the American Family’ 157, 158

213

Catholic Social Justice lobby 154 Chambers of Commerce 148–50, 168, 188, 195 charitable choice see also American legislation and Finance Committee Bill 8–9 opposition to 10 principles 10 provisions of 8–11 rules 9 Charitable Organizations Society 104 Charities Publication Committee 76 child care/welfare 6, 8, 127, 141, 169 funding and legislation 6 and welfare recipients 127 Child Care and Development Fund 6 child labor, restrictions on 77 child support 5 enforcement of 151 and establishment of paternity 5 and non-payers 127 Child Support Enforcement, Oﬃce of 4 Child Support Enforcement program 5 Children’s Bureau 77, 105 Children’s Defense Fund (CDF) 153, 161, 163, 164 and Marion Wright Edelman 153 Children’s Development Fund (CDF) 150 Children’s Health Initiative Program (CHIP) 152 Chiles, Governor 142, 143 Christian Action Network 157 Christian Broadcasting network (CBN) 156–7 Christian Coalition (CC) 134, 156–7, 160, 163, 164, 188–9, 195–6 and Pat Robertson 156 Christian right 192 Christianity/Christian churches (and) 189–90, 191 civic duty to work 23 medieval poor relief 94 monastic orders 93–4 Protestant evangelical 10 welfare 93–4 civil rights activists 75 Clarke, J. 22 Clean Air Initiative 77–8

214

The political economy of welfare reform in the United States

Clinton, Hillary 137 Clinton, President Bill 76, 117, 118–21, 135–46, 155, 161, 163 and New Covenant 118 and scandals 135–6 and the Clinton Bill: Work and Responsibility Act (1994) 119–21, 123–4 Clinton administration 133, 153, 161, 176 Coalition on Human Needs 150 Coalition for Traditional Values 156 Coase, Ronald 46–8 Coase theorem 46–8 examples of 47 and government solution 48 and Pigovian solution 46, 47 Cochrane, A.D. 14, 15, 18 Common Agricultural Policy 32 Compassionate Capital fund 9 Congress: Keystone of the Washington Establishment 87 Contract with America 121–3, 140, 157 Craig, F.W.S. 34 Crain, W.M. 91, 92 Danielson, Albert L. 71 David M. Griﬃth and Associates 151 Davis, Otto A. 58 de Toqueville, Alexis 100–101, 102, 196 deﬁcit reduction budget 137 deﬁnition(s) of caseload 11–12 welfare state (British) 20 Delors, Jacques 29 Denzau, Arthur 58 Dickens, Charles 104, 196 Dole, Senator Bob 119, 127–8, 140, 141–2, 145, 157–8 Dougan, William R. 71–2 Downs, Anthony 52, 54–6, 57, 60, 61, 71 Downs’ model 54–5, 57 Drescher, Seymour 101, 196 Drew, Elizabeth 140 Durden, Garey C. 71 Earned Income Tax Credit (EITC) 180, 194, 197 Economic Theory of Democracy, An 54

Economics of Welfare, The 46 Edgeworth box diagram 37, 41 elderly people 64, 72–3, 112–13, 178 as pressure group 107–8 Elman, Richard 108 employment 7 multiple barriers to 6 Empower America 156, 160, 163, 167 Engler, Governor 141, 142, 144, 145–6, 148, 158–9 Europe and comparative social welfare policy 13–34 European Centre of Enterprises with Public Participation (CEEP) 34 European Commission 32–3 European Social Fund 28, 32–3 European Court of Justice 31, 33 European Economic Community (EEC) 32 European Monetary Union (EMU) 31 European Trade Union Confederation (ETUC) 34 European Union (and) 26, 28–34 Agreement on Social Policy 30, 31 budget 32 Community Action Programme 33 Community Charter of Fundamental Social Rights of Workers 29–30 cooperation on employment 28–9, 34 Economic and Social Committee 31 employee protection measures 34 fears of social dumping 29–30 impact on European social policy 28–32 Maastricht summit (1992) 30 migrants 33 Protocol on Social Policy 30 Single European Act (SEA, 1986) 29 Single European Market (SEM) 30 Social Charter (Council of Europe) 28 social policy requirements 33–4 Treaty of Amsterdam 30–31 externalities 45–6 and marginal external beneﬁt curve (MEB) 45 and marginal social beneﬁt curve (MSB) 45

Index negative 45 positive 45 solution to problem of 45–6 Faith-Based and Community Initiatives, Oﬃce of (OFBCI) 9 fallacy: fractile and multiple 59 Family Research Council 156 Federal Elections Commission 78 Fiorina, M.P. 82, 87–90 ﬁscal implications 7–8 food stamp programs 11, 13, 71, 73, 77, 79, 112, 114, 125, 127, 141, 155, 176 Ford, President Gerald 92, 116 Frey, B.S. 58 Friedman, Lawrence 109 Friedman, Milton 114 funding issues 127–9 Garner, John 106 Gephardt, Richard 124 Germany (and) 14, 23–8 Act of Reuniﬁcation 24 Basic Law 23 immigration controls 27 political forces in 24 recruitment of women/migrant workers 26–7 reuniﬁcation of 25, 27 single parents in 25 Social Assistance: means-test cash beneﬁt 24–5 social insurance 24, 25, 28 Treaty of Monetary, Economic and Social Union 24 unemployment in 27 women’s role in 25 Gingrich, Newt 122, 139, 140, 141, 142, 145, 157, 160 Ginsburg, N. 15–19, 24, 25, 26, 34 Gist, J.R. 85 Goﬀ, B.L. 71 Goodman, R.P. 58 government and bureaucracy (and) 79–92 see also government bureaucracy committee power 82–4 congressional dominance 84 iron triangles 87–90

215

legislative stability 81–4 logrolling 81–2 marginal voting districts, decline of 88–90 modeling legislative process: public choice approach 79–81 presidential role 91–2 veto power/gubernatorial vetoes 91–2 Washington establishment 87–8 government bureaucracy 84–7 and balance of power/objectives 86 budget 85–6, 87 economic eﬃciency of 86 oversupply of output 85 Governors of the nation see also National Governors Association (NGA) pecuniary interest of 144–6 role of 178 Greenspan, Alan 136–7 Grier, Kevin 58, 71 Group Basis of Politics, The 77 Hagan, Jan 117 Hall, Tony 141 Hantrais, L. 28–30, 31–2, 33 Hardin, Cliﬀord 113 Harrington, Michael 108, 109, 196 Hart, H.A. 51 Haskins, R. 121–2 health care programs 73, 137–8 Health and Human Services, Department of 4, 9, 11, 80, 119, 134, 145, 151 Heclo, H. 16 Heritage Foundation 134, 143–4, 146, 154, 156, 158, 160, 163–4, 165, 195 Hill, R.C. 85 Himmelfarb, Gertrude 95, 98, 99–100 Hinich, M.J. 57 Hobson, B. 18 Hochman, H.M. 42 Holcombe, Randall 58, 111 Holmlund, B. 15 Hoover, President Herbert 106–7 Hotelling, H. 54 ‘How They Voted’ 148 Hrebenar, R. 78

216

The political economy of welfare reform in the United States

ideology 70–72, 92, 134, 162, 188–9 as determinant of votes 71 inﬂuence of 71–2 illegitimacy 8, 127, 158–60, 187, 192 see also single parents and anti-illegitimacy groups 160 immigrants 19, 27, 77, 101, 126, 135, 182 Inman, R.P. 58 institutional analysis (and) 132–67 see also Clinton, President Bill American Public Human Services Association (APHSA) 147–8 Chamber of Commerce 148–50 see also Chambers of Commerce Congress 140–44 Governors of the nation see main entry National Governors Association (NGA) 76, 142, 143–4, 146–7, 150 public sector service providers 150–67 welfare reform 143, 160–61 interest groups 63–7, 86, 90, 92, 133, 162, 165–6, 171, 188–9, 196–7 see also voters/voting anti-illegitimacy 160 Becker theory of 65–7 and lobbying 64–5, 116, 171 and TANF 185 iron triangles 87–90 Isikoﬀ, M. 156 job creation 160 Job Opportunities and Basic Skills (JOBS) program 3, 117, 129, 141 Jobs Partnership of Washington County 10 Johnson, President 108, 109–10, 113 Joseph, Keith 21 Josten, Bruce 149, 192 Kalt, Joseph P. 71, 92 Kammer, Rev. Fred 154 Kasich, John 130, 145 Kau, James B. 70, 71, 92 Kemp, Jack 160 Kennedy, President John 108 Kennedy–Johnson era 110, 113

Kerrey, Senator Bob 137 Keynesian economic policies 15, 21 Kincaid, J. 164 Kondratis, Anna 95, 109–10, 111–12 Labor and Human Resources Committee 127 Lagrangian mechanics 36, 38, 40 Landes, W.M. 91 Latham, Earl 77 Lathrop, Julia 105 Leavitt, Governor 145–6, 165 legislation see American legislation; British legislation and Germany Lesher, Richard 149–50 Liddle, R. 22 Liebfried, S. 25 lobbying 64–5, 77, 86, 171 Locke, John 98 Lockheed Martin Corporation 150–51 logrolling 81–2 Lott, Trent 138–9, 141, 147, 160 Lund, B. 20–22, 23, 34 Luntz, Frank 122–3 Lurie, Irene 117 Madsen, H. 16 Major, John 21 Malthus, Thomas 100–101, 102, 196 Mandelson, P. 22 marginal districts, decline of 88–90 marriage, promotion of 4–5 Marshall, W.J. 83 Martini, Alberto 180, 193 Mawyer, Martin 157 Maximus 150–51 and acquisitions 151 and government contracts 152 Mayhew, D.E. 62, 88 McGovern, Senator George 114 McKelvey, Richard D. 82 median voter model, critique of 56–8 median voter theorem 52, 53–58 Black’s theorem and preference curves 53–4 Downs’/Downsian model 54–6, 57 Medicaid/programs 13, 73, 109, 117, 143, 155, 165, 169, 176, 179 block grants for 147, 195 expenditures on 143

Index state historical spending on 166 and welfare reforms 144, 146 Medicare 73, 109, 139, 152 Mill, J.S. 102 Mills, Wilbur 114–15 Mishan, E.J. 43 Moe, T.M. 84 Moﬃtt, Robert 187 Moran, M.J. 85–7 Morris, Dick 138–40 ‘Moving Ahead: Initiatives for Expanding Opportunity in America’ 121 Moynihan, Senator Daniel P. 108–9, 112, 113, 117, 120, 128, 136, 196 Mueller, Dennis 43, 47, 49, 58, 62, 70, 83, 84 Munger, Michael C. 71–2 Murray, Charles 22, 159, 178, 185, 186, 196 Myrdal, A. 15 Myrdal, Gunnar 19 Nash, Gary 117 Nathan, Richard 113 Nathan plan 114 National Governors Association (NGA) 76, 142, 143–4, 146–7, 150 National Urban League (NUL) 155 National Welfare Rights Organization 116 Nelson, Douglas 71 Network 154, 163, 164 New Federalism 116 new welfare law 93–117, 132, 134, 197 genesis of see Personal Responsibility and Work Opportunity Reconciliation Act (1996) in Britain see Britain: development of welfare programs in in New World see America: early years (to 20th century); America in the early 20th century; America in mid- to late 20th century and American legislation Niskanen, William A. 59, 84–6 model 85 Nixon, President Richard 112–14

217

Nixon administration 112–16 Nozick, Robert 51 Nutrition Program for Women, Infants and Children (WIC) 128, 141 Oates, W.E. 63 Olson, Mancur 52, 63–4, 67, 75–5, 133 One-Stop Career system 9 Ordeshook, P.C. 48–9, 60–61 ordinary least squares (OLS) models 188, 190–91, 192 out-of-wedlock births, discouragement of 4–5 overview of social welfare policy issues 1–12 caseload, deﬁnition of 11–12 charitable choice provisions 8–11 child care 6 child support 5 employment 7 ﬁscal implications 7–8 single parents, marriage promotion and abstinence education 3–5 underclass, creation of 1–3 welfare recipients with multiple barriers to employment 6 work incentives, strengthening of 1–3 Packwood, Senator Bob 126–7, 145, 160 Panetta, Leon 137, 138, 140, 141 parenting and social skills training 4 Pareto, Vilfredo 35 Pareto/Paretian 35–43, 45–6, 161, 168 criteria 36, 41–3, 52 frontier 37, 38 optimality 36, 39–42, 45–6, 52, 132, 133 welfare economics 41–2 Passport Denial Program 5 paternity and responsible fatherhood programs 4 and welfare beneﬁts 5 Peacock, Alan T. 41 Peltzman, Sam 52, 62, 69, 71, 87 and model of wealth transfers 69–70 Personal Responsibility and Work Opportunity Reconciliation Act

218

The political economy of welfare reform in the United States

(1996, PL 104–193) 118–31, 132, 134 Clinton Bill, the 119–21, 123–4 Contract With America: Republican politics and ideology 121–3 Personal Responsibility Act 122, 123–4, 140 Republican legislative proposals 123–31 Pierce, President Franklin 102, 105 Pigou, A.C. 45–6, 47 Political Action Committees (PACs) 90, 163, 188, 189, 190–91 Pommerehne, W.W. 58 Poole, L. 24, 25, 27 Poor Law, Henrician 94–5 Poor Law (1601) 95–6, 99–100 Posner, R.A. 91 preference curves 53–4 Problem of Social Cost, The 46 protest groups 77 public choice perspective (and) 52–78 existing theories and welfare programs development 72–8 ideology, inﬂuence of 70–72 interest groups, role of 63–7 median voter model, critique of 56–60 median voter theorem: basic model 53–6 rent seeking 67–70 voters and rational ignorance 60–63 see also rational ignorance and voters/voting public good, characteristics of 63 public goods (and) 43–5 jointness of supply 43–4 marginal social beneﬁt (MSB) 44 non-excludability 43–4 public sector service providers 150–67 ‘Putting People First’ 135 rational choice analysis 84 rational ignorance 55, 62–3 Rawls, John 49–51 and social contract theory 50–51 Reagan, President Ronald 116–17, 169 Rector, Robert 143–4, 146, 156, 158–9, 164, 165–6, 186

Reed, Bruce 119, 140 Reed, Ralph 156–8 ‘Relevance of the Median Voter Theorem, The’ 56 religious freedom 11 organizations 10–11 religious right 11, 156, 158 rent seeking (and) 67–70 monopoly 68 policies sought 68 wealth transfer model 69–70 responsible fatherhood programs 4 Research on Poverty, Institute for (IRP) 153–4 research/studies (on) 4 child poverty reduction programs (congressional research service) 4 outgoing rotation group, current population survey 182–3 poverty in late 19th century to early 20th century 104 social conditions, Pittsburgh survey of 76 voters 62 Richardson, Elliot 115 Right to Life 154 Riker, W.H. 60–61 Rockefeller, Governor 114 Rogers, J.D. 42 roll call votes in Congress 71 and eﬀect of ideology on 71 Romer, T. 56, 58–9, 87 Roosevelt, President Franklin D. 107 Roosevelt, President Theodore 146 Rosenthal, H. 56, 58–9, 87 Rowley, Charles K. 41, 56–7, 60, 90, 91 Rubenfeld, Daniel L. 59 Rubin, Paul H. 70, 71, 92 Salaman, Lester 110 Samuelson, P.A. 42, 53 Sanford, Terry 111 Schneider, Friedrich 58 Schultz, George 113, 114 Scitovsky, T. 43 Scott, H. 16 Settlement House workers 75–6 Sheppard–Towner Act 77, 105–6, 195

Index Shepsle, K.A. 81–3 Silberberg, Eugene 51, 61, 71 Silberman, Jonathon I. 71 single parents 1, 3–5, 6, 8, 13, 17–18, 23, 32, 117, 161, 178, 182, 192 incentives for births to 185 percentage of births to 186 and relationship between welfare and births to 169, 178, 185–8 teenage 126, 128, 129, 154, 159, 160 and working hours legislation 3 Smith, Adam 35, 99–100, 102 Smith, John 22 Smith, T. 142, 143, 147, 164, 165, 167 Smithies, A. 54 social contract theory 50–51 Spencer, Herbert 102 Stephanopoulos, George 135, 136, 137, 138, 139–40 Stigler, George 52, 61, 65, 68–9, 87 Storm Over the States 111 ‘Structure-induced equilibrium and legislative choice’ 82 Supplemental Security Income (SSI) 6, 125, 151 Sweden (and) 13, 14–19, 32 birth rate in 15 caseload growth 19 child care programs 16 claimants/recipients 17–18, 19 New Right 17 refugees and immigrants 19 single parent families 17–18 Social Democratic Party (SAP) 15, 16–17 social welfare policies in 14, 16 State Labour Market Board 15 unemployment insurance beneﬁt (UIB) 19 welfare reforms 16 Takahashi, M. 18 Technical Report: Explaining the Decline in Welfare Receipt 1993–1996 (1997) 171 Temporary Assistance to Needy Families (TANF) 2–3, 4, 5, 7, 73, 106, 129, 132, 169, 171–2, 175, 178, 183, 185

219

and caseload decline 6, 7, 11, 175 caseloads 11–12 and charitable choice provisions 10 eligibility for 8 and unspent funds 147 Texas Department of Human Services 10 Thatcher, Margaret 21, 22 Theory of Justice, A 49 Theory of Moral Sentiments, The 99 Thompson, Governor 142, 145–6 Tollinson, R.D. 71 Tollison, R. 91, 92 traditional public interest model 35–51 see also Pareto/Paretian allocative eﬃciency or redistribution 48–51 see also main entry Coase theorem 46–8 see also main entry Edgeworth box diagram 37, 41 equimarginal principle 39–43 externalities 45–6 see also externalities (and) market failure – public goods 43–5 see also public goods (and) Pareto optimality 36, 39–42, 45–6 production 38–9 pure exchange 36 Trattner, William I. 78, 93, 94, 97, 99, 102, 103–4, 105, 107, 108, 110, 113, 117 Tullock, Gordon 42, 52, 67–8, 81, 82 two stage least squares (2SLS) 188, 191–2 UK see Britain underclass, creation of 1–3 Union of Industrial and Employee’s Confederations of Europe (UNICE) 34 universal citizenship rights 32 unspent funds 146, 147 vetoes/ gubernatorial vetoes 91–2 voters/voting 52–63 see also median voter model, critique of and median voter theorem and ideology 71–2 rational 61–2 and rational ignorance 62–3

220

The political economy of welfare reform in the United States

Wagner, Senator 106 waivers 171, 172, 173, 175, 179–81, 184 Wallis, J.J. 70 Washington establishment 87–8 Watson, Justin 156 Wealth of Nations, The 35, 99 Weber, Max 84, 160 Wehner, Peter 161 Weingast, Barry R. 81–3, 85–7, 92 welfare, ending of entitlement to 130 welfare caseloads, changes in 168–93 Council of Economic Advisors (CEA) study of 171–5 econometric speciﬁcation 172–3 results 173–5 and critique of studies 179–86 econometric speciﬁcation 181–3 proposed alternate model 180–81 results 183–6 econometric models of 171–9 and extensions of basic model 186–92 combined interest group-median voter model 189–90 interest groups and ideology 188–9 relationship between welfare programs and births to unwed mothers 186–8 two stages least squares (2SLS) model 190–92 historical 169–71 Rebecca Blank study of 176–9 econometric speciﬁcation 176 results 176–9 and waivers 171, 172, 173, 175, 179–81, 184 welfare caseloads 11–12 decline in 6, 7, 11 growth of 1–2

variables aﬀecting changes in 168–93 see also welfare caseloads, changes in ‘Welfare Costs of Tariﬀs, Monopolies and Theft, The’ 67 welfare issues/policies 71, 77 welfare legislation 2–12, 52, 145 see also new welfare law losers from 74 welfare programs development of 72–8 funding for 140–41 welfare recipients 5, 7, 73, 127, 128, 135, 150–51, 161–2 level of beneﬁts paid to 181 with multiple barriers to employment 6 as subsidized labor 151 work requirements for 160, 171 welfare rights advocates 75 welfare reform 143, 160–61 ‘Welfare Trap, The’ 149–50 welfare-to-work grant program 7 ‘What Causes Public Assistance Caseloads to Grow?’ 176 Whiteﬁeld, George 98 ‘Why So Much Stability?’ 81 Wicksell, Knut 48 Williamson, O.E. 84, 85 Wilson, J.Q. 157 Wiseman, Michael 180, 193 Witte, Edwin 107 Woodward, Bob 136–7 work incentives 1–3 legislation 2–3 Work Opportunity Tax Credit 149 work requirements 8 Zardkoohi, Asgher 111 Zupan, Mark A. 71, 92