A Well-Tailored Safety Net
A Well-Tailored Safety Net The Only Fair and Sensible Way to Save Social Security
Jed Gra...
37 downloads
478 Views
1MB 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
A Well-Tailored Safety Net
A Well-Tailored Safety Net The Only Fair and Sensible Way to Save Social Security
Jed Graham
PRAEGER
An Imprint of ABC-CLIO, LLC
Copyright 2010 by Jed Graham 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, photocopying, recording, or otherwise, except for the inclusion of brief quotations in a review, without prior permission in writing from the publisher. Library of Congress Cataloging-in-Publication Data Graham, Jed. A well-tailored safety net : the only fair and sensible way to save social security / by Jed Graham. p. cm. Includes bibliographical references and index. ISBN 978–0–313–38169–0 (hbk. : alk. paper)—ISBN 978–0–313–38170-6 (ebook) 1. Social security—United States. I. Title. HD7125.G695 2010 368.4’300973—dc22 2009042103 14 13 12 11 10
1 2 3 4 5
This book is also available on the World Wide Web as an eBook. Visit www.abc-clio.com for details. ABC-CLIO, LLC 130 Cremona Drive, P.O. Box 1911 Santa Barbara, California 93116-1911 This book is printed on acid-free paper Manufactured in the United States of America
For Deborah and Max
Contents
LIST OF FIGURES
ix
LIST OF TABLES
xi
PREFACE
xiii
A NOTE ON THE NUMBERS IN THIS BOOK
xvii
PART I: DEFINING AN URGENT PROBLEM
1
INTRODUCTION
3
HIGHLIGHTS OF A WELL-TAILORED SAFETY NET SOCIAL SECURITY REFORM PROPOSAL
11
CHAPTER 1: A ‘‘VERY FALSE SENSE OF SECURITY ’’
15
CHAPTER 2: ‘‘A POLITICAL FRAUD’’
21
CHAPTER 3: THE PRICE OF DELAY
27
PART II: NO EASY ANSWERS
35
CHAPTER 4: ‘‘A BAD IDEA’’
37
CHAPTER 5: ‘‘IT ’S KIND OF MUDDLED’’
43
CHAPTER 6: COLD COMFORT
49
CHAPTER 7: A NO-BRAINER
53
viii
CONTENTS
CHAPTER 8: A CRACK IN THE FOUNDATION
57
CHAPTER 9: A FIX THAT DOESN’T FIX VERY MUCH
61
PART III: PRINCIPLES FOR REFORM
67
CHAPTER 10: A BAD COMB-OVER
69
CHAPTER 11: THE PROBLEM IS PART OF THE SOLUTION
73
CHAPTER 12: ‘‘THE OBVIOUS THING TO DO’’
79
CHAPTER 13: A BRIDGE TOO FAR
85
CHAPTER 14: THE HILDA & DAVID AND IRENE & BERNIE TESTS
93
PART IV: LAYING OUT A SOLUTION
97
CHAPTER 15: OLD-AGE RISK-SHARING
99
CHAPTER 16: A FAIR AND CONSTRUCTIVE SACRIFICE
111
CHAPTER 17: A HELPING HAND, NOT AN EMPTY PROMISE
125
CHAPTER 18: A WELL-TAILORED ACCOUNT STRUCTURE
131
CHAPTER 19: MEASURING UP
143
CONCLUSION: KEY FINDINGS TO GUIDE SOCIAL SECURITY REFORM
157
APPENDIX 1: DETAILS OF A WELL-TAILORED SAFETY NET SOCIAL SECURITY REFORM PLAN
163
APPENDIX 2: SOLVENCY IMPACT OF A WELL-TAILORED SAFETY NET, BILLIONS OF 2009 DOLLARS
169
NOTES
171
INDEX
191
List of Figures
3.1: THE GROWING BURDEN OF INTEREST ON THE PUBLIC DEBT 3.2: PROJECTED FEDERAL OUTLAYS AND REVENUE 3.3: THE COST OF DOING NOTHING UNTIL THE TRUST FUND IS EXHAUSTED 7.1: FIXING SOCIAL SECURITY FOR 75 YEARS—BUT ONLY ON PAPER 8.1: BUILDING ON A WEAKENED FOUNDATION 11.1: SLOWER GROWTH IS THE NEW NORMAL 16.1: SOCIAL SECURITY ’S DIMINISHING RATE OF RETURN 19.1: INCREASE IN FEDERAL DEBT UNDER VARIOUS REFORM PLANS
29 30 33 55 59 77 112 147
List of Tables
1.1: 3.1: 5.1: 6.1: 7.1: 8.1: 8.2:
Washington piles up debt, as it pretends to save (billions of dollars) The annual cost of fixing Social Security if we start in . . . Benefit cuts under Bush plan in 2075 Benefit cuts under Bennett plan in 2075 The cost of closing Social Security’s financing gap with tax hikes The penalty for retiring early The double-whammy of benefit cuts and early-retirement penalties in 2075 9.1: Reducing cost-of-living adjustments hurts the most vulnerable 9.2: The impact of smaller COLAs and early retirement penalties 10.1: Benefit cuts for an average earner under Diamond-Orszag 11.1: Workers are retiring younger, living longer 11.2: Social Security’s traditional structure is no longer a good fit 13.1: LMS personal account structure 13.2: Benefits under LMS plan in 2055 with Treasury returns 13.3: Benefits under LMS plan in 2055 with 1.5% real returns 13.4: Implied tax hikes under LMS plan—a narrow view 15.1: The regressive impact of a 40-year benefit formula in 2050 15.2: Discouraging early retirement at the expense of income security 15.3: Old-Age Risk-Sharing for low earners in 2035 15.4: One extra work year erases Old-Age Risk-Sharing for low earners 15.5: New minimum benefit could offset full impact on low earners of Old-Age Risk-Sharing and increase in Normal Retirement Age 15.6: Two extra work years erase Old-Age Risk-Sharing for average earners
18 34 46 50 54 58 60 65 65 70 74 74 88 90 90 91 102 104 106 106 107 107
xii
LIST OF TABLES
15.7: Three extra work years erase Old-Age Risk-Sharing for high earners 16.1: Not-so-Progressive Price Indexing 16.2: Bennett plan cuts hit those in the middle 16.3: A less-than-proportional tax hike 16.4: Distribution of sacrifice under LMS plan 16.5: A smaller, more proportional tax hike 16.6: Distribution of sacrifice under a progressive saving offset 16.7: Distribution of sacrifice under a modified progressive saving offset 17.1: Benefit eligibility for workers turning 62 in . . . 17.2: Saving to cope with scaled-back early eligibility 17.3: Employer Social Security payroll tax rates for workers 62 and up in 2022 18.1: The deficit impact of carve-out personal accounts in 2025— an example 18.2: A well-tailored account structure 18.3: Account accumulations for low earner with 40 years of work, 2009 dollars 18.4: Benefit levels at age 92 under two delayed retirement scenarios 18.5: A progressive delayed retirement incentive 19.1: A comparison of benefits for average earner in 2075 19.2: A further comparison of benefits for average earner in 2075 19.3: Benefits to an average earner disabled at age 62 in 2075— a comparison 19.4: A safety net that firms up earlier for lower earners, but protects everyone 19.5: Distribution of sacrifice under A Well-Tailored Safety Net 19.6: The choice A1.1: The impact of a progressive saving offset in 2056 A1.2: Old-Age Risk-Sharing reduction in 2035 and later, retire at 65 A1.3: Benefit eligibility for workers turning 62 in . . . A1.4: A well-tailored account structure A1.5: Initial Old-Age Risk-Sharing reduction for those disabled in 2032 and later A2.1: Solvency impact of A Well-Tailored Safety Net, billions of 2009 dollars
108 113 114 115 116 120 121 122 127 128 128 133 136 137 140 140 148 149 151 152 153 154 164 165 166 167 168 169
Preface
I ambled into the largely empty White House briefing room for the first time early one morning in December 2004 and took a seat in the fifth or sixth row, where I could observe from a distance. Just assigned to cover the White House for Investor’s Business Daily, I was pretty excited, though a little intimidated, by the challenge ahead. But a few minutes later, I got a rude awakening. It turns out all the seats had attached nameplates and my seat was permanently reserved for the Washington Times. I jumped up and decided I’d be just fine standing by the door. If you had told me what the future held on that nervous morning, I would have said you were crazy and had a good, hearty laugh. I didn’t set out to try and save Social Security or challenge the status quo in Washington. My only goal was to bring some fair-minded reporting to the economic issues I’d be writing about. The administration had declared Social Security reform its top domestic priority, and I dived in to learn everything I could. By the end of April 2005, when President George W. Bush announced his plan to erase Social Security’s financing gap by slowing the growth in benefits, I had learned enough to know almost immediately that the plan would at best solve 50 percent of the shortfall. Nevertheless, the plan was attacked as ‘‘an assault on the middle class.’’1 All of Washington would soon declare Social Security reform dead, but as most members of Congress breathed a sigh of relief and my colleagues in the press moved on to the next partisan battleground, I felt that my job was unfinished. My reporting had revealed that there were legitimate reasons not to like major aspects of all the plans from across the political spectrum that had been offered and quickly cast aside, usually without any debate. But if none of these proposals made sense, what did? I felt that the issues at stake were too important to let go of without attempting to find
xiv
PREFACE
an adequate answer to that question, and as I absorbed the best wisdom of all who have wrestled with this longstanding policy puzzle, quite unexpectedly, one idea after another occurred to me. Four years later—an interval certainly extended by the arrival of my amazing son Max—as a new president begins wrestling with Social Security reform, I am ready to present my conclusions. Before I do, I want to express my thanks to those who helped shape my understanding of both the logic and math of Social Security reform and others who also helped make this book possible. Eugene Steuerle’s views on Social Security reform were easily the biggest influence on my own thinking. Steuerle’s argument in favor of back-loading benefits to direct a greater proportion of Social Security’s resources to older retirees planted the seed that would develop into a new reform provision called Old-Age Risk-Sharing at the heart of A Well-Tailored Safety Net. Steuerle, a long-time Urban Institute fellow who served in the Treasury Department under President Ronald Reagan, was gracious enough to engage in an e-mail correspondence over several years and to offer many helpful suggestions as I developed my reform proposal and after reading the first draft of this book. Ed Lorenzen, in between stints as a policy specialist for former Rep. Charlie Stenholm and Majority Leader Steny Hoyer, offered helpful analysis of my earliest efforts to draft a proposal and put me in contact with Steve Robinson, a Social Security expert for Republicans on the Senate Finance Committee, who provided critical insight as I began the intensive process of measuring the financial impact of my policy proposals. Andrew Biggs, former deputy Social Security commissioner and now scholar at the American Enterprise Institute, provided important advice on designing payrolltax relief for employers of older workers. Many other Social Security policy experts whose research I came across also influenced this work, and many are cited within these pages, but here I wish to recognize those whose alternative proposals are subjected to a critical analysis in A Well-Tailored Safety Net. The constructive contributions to this debate from Peter Orszag, Peter Diamond, Robert Ball, Jeffrey Liebman, Maya MacGuineas, Andrew Samwick, Robert Pozen, Sen. Robert Bennett, former Congressmen Jim Kolbe and Charlie Stenholm, among others, served as essential guideposts as I sought to navigate Social Security reform with a proposal that skirted its perils and embraced its promise. I recognize that they were acting in good faith and working within political constraints to make the best of a difficult situation. While I take issue with some of their approaches, they all have at least one big upside: They would substantially reduce an unnecessary burden of debt that would be heaped on coming generations of workers if Washington continues its do-nothing approach. In certain instances, I followed their lead, adopting, for example, a more generous minimum benefit for low career earners along the lines of the Kolbe-Stenholm approach. I also adopted a scaleddown version of a proposal supported by both Bennett and Rep. Paul Ryan to end the tax-free status of employer-provided health coverage.
PREFACE
xv
My work would have been impossible to complete without the extensive resources and analyses made available to the public by the Social Security Administration’s Office of the Chief Actuary, ably led by Stephen Goss. Alice Wade, deputy chief actuary, welcomed all questions and helped me understand why my present value calculations were off by two whiskers. I’m grateful to Robert Hutchinson, my editor at Praeger, for his belief that the merit of an argument is more important than the renown of its author, and for the commitment he and the wonderful team at ABC-CLIO brought to realizing the full potential of this book. I’d also like to thank Pattie Stechschulte, project manager at BeaconPMG, who was a pleasure to work with as I made substantial revisions late in the process. I might have been hard-pressed to publish A Well-Tailored Safety Net without the support of Investor’s Business Daily publisher Bill O’Neil. He encouraged me to write a book about my ideas, even though he only had a vague notion of what I would propose but surely understood that my approach to Social Security reform would depart from IBD’s conservative editorial view. I also want to thank IBD executive editor Chris Gessel, managing editor Susan Warfel and licensing vice president Heather Davis for their support, and I want to recognize my IBD colleague Paul LloydStrongin for generously lending his graphics wizardry to this project. The final product also benefited from the insight of Wall Street Journal economics editor David Wessel, who looked at an initial draft and offered advice on making the presentation more user-friendly and the tone less off-putting. On a personal note, I’d like to thank my wife, Deborah, for having the confidence in me to say way back in 2005, ‘‘If anybody can figure it out, you can;’’ for putting up with wrong turns and too many late nights; and for helping push me across the finish line four years later. Thankfully, I had our beloved cat, the diva/rascal Sammy, to keep me company as I burned the midnight oil. Lastly, I’m grateful that my grandmother Irene—still full of life at 94—and my other wonderful grandparents Bernie, Hilda and David lived long enough to demonstrate for me just how critical it is for us to maintain a robust Social Security safety net that provides comfort and dignity in very old age.
A Note on the Numbers in This Book
Numbers can be used to rig an argument. Because that isn’t my intention, I want to provide a little background about the choices I made, particularly with respect to the benefit levels future retirees would receive under various proposals. Two points are in order. First, in their analyses of reform proposals, the Social Security actuaries display the percentage of benefits future retirees would receive in comparison to the current system. While that is certainly of value, an even more telling comparison is the share of benefits a worker would receive under various proposals relative to one who retired at the Normal Retirement Age and faced no early-retirement penalties. This comparison to the current law Primary Insurance Amount that is unadjusted for age of retirement—what I will refer to simply as the base benefit—reflects my belief that the appropriate benchmark is the level of income security provided by the safety net, rather than the level of income security relative to the current system that does not provide a particularly robust safety net once you subtract early-retirement penalties. Along the same lines, in detailing the impact of various proposals, I display the level of benefits a worker would receive based on 2009 wages. This comparison reflects a different perspective than that held by the Bush administration, which promoted its approach to Social Security reform by assuring workers that most would receive a benefit more generous in inflation-adjusted terms than that received by current retirees. President George W. Bush’s approach emphasized the fact that benefit levels will rise over time while downplaying the fact that workers would, over time, receive much smaller benefits relative to career income. While it is true that wage growth will raise future standards of living, arguably providing an opportunity to make Social Security’s safety net somewhat less generous, I think the share of career
xviii
A NOTE ON THE NUMBERS IN THIS BOOK
wages replaced by Social Security is a more relevant standard than how one’s benefit compares to that of current retirees. One might reasonably argue that I was trying to rig the debate by focusing on Social Security’s income-replacement rate had I offered a proposal that avoided benefit reductions. That is not the case. A Well-Tailored Safety Net emphasizes the importance of replacement rates for those who live to an advanced age and face a growing risk of outliving their savings, rather than for workers turning 62 or 65. So as not to either understate or overstate the size of Social Security’s financing gap, I use nominal numbers (unadjusted for inflation) with respect to Social Security projections within the 10-year budget window; inflation-adjusted numbers for program finances through 2036 and present-value numbers for the 75-year financing gap. For further perspective I provide information on Social Security’s projected gap relative to the size of the economy, or Gross Domestic Product, where appropriate. Present value reflects the level of interest-bearing deposits the government would need right now to close Social Security’s financing gap. Analyses that show Social Security’s actuarial deficit as $5.3 trillion in present value treat the $2.4 trillion Social Security trust fund as money in the bank. Because the trust fund doesn’t provide any resources to the government, I use a figure of $7.7 trillion in present value—the total cost minus tax income found in table IV.B5 of the 2009 Trustees report. To be consistent, when stating how much of Social Security’s deficit might be closed by various reforms, I use the full $7.7 trillion gap. The intent is not to strengthen the case for tax hikes relative to benefit cuts, or vice versa, but to reflect the full reality of the budget challenge related to Social Security in order to produce a more constructive political debate. While part of this challenge involves an unfunded liability from Treasury, i.e. taxpayers, to Social Security, there is no more logical time to address this trust fund debt than concurrent with Social Security reform. Finally, I use the Social Security Administration’s analysis of personal account accumulations to judge what percentage of wages a worker would have to save outside of Social Security to overcome benefit cuts. This analysis can be seen in the Office of the Chief Actuary memorandum of November 17, 2005: ‘‘Estimated Financial Effects of ‘A Nonpartisan Approach to Reforming Social Security.’ ’’ Table B1a shows how big of a lifetime annuity could be purchased by various income groups after a full career of personal account deposits equal to three percent of annual wages. Assuming investment only in Treasury bonds, the analysis shows that one percent of income saved by low earners ($18,900 in 2009) would replace 7.6 percent of their Social Security benefit, compared to 10.3 percent for an average earner ($42,000), 12.4 percent for a high earner ($67,200) and 16.8 percent for those who earn the maximum wage now covered by Social Security taxes and benefits ($106,800). This data reflects average careers of less than 40 years for moderate wage earners. Longer careers would therefore produce better outcomes. The actuaries also assume a 0.3 percentage point annual account fee; however, if investments were limited to Treasuries, that should reduce management and transaction costs, producing slightly higher net returns.
Part I
Defining an Urgent Problem
Introduction
As the monumental house of cards built by Wall Street’s financial wizards came crashing down, bringing the economy and retirement portfolios down with it, Social Security remained a relatively safe port in a raging storm, and thank goodness for that. But Social Security is only as sound and secure as the nation’s finances and economy that serve as its foundation, so for the next generation of retirees, it will hardly be free of risk. The unprecedented surge of federal debt that will be a legacy of the current crisis has only advanced Social Security’s inevitable day of reckoning. The levees are still holding, but the growing tide of red ink—a looming threat just a year ago—is fast becoming a clear and present danger. The current era was supposed to be the calm before the storm—a final chance to get our fiscal house in order before the retirement of 77 million baby boomers places an overwhelming strain on the old-age safety net provided by Social Security, Medicare and, to a lesser extent, Medicaid. Instead, the economic and financial crises are projected to add several trillions in unexpected debt, raising federal borrowing and the associated interest expense to worrisome levels within a decade. Recognizing the urgency of the problem, President Barack Obama, in an interview just before his inauguration, said that the time had come to reform these safety net programs that are collectively known as entitlements because their spending obligations have been written into law and are unconstrained by budgetary limits. ‘‘There are going to be some very difficult choices, and issues of sacrifice and responsibility and duty are going to come in, because what we have done is kicked this can down the road,’’ Obama said. ‘‘We’re now at the end of the road, and we are not in a position to kick it any further.’’1
4
A WELL-TAILORED SAFETY NET
Consider this relatively favorable scenario—at least in regard to the nation’s debt levels—that has tax revenue rising to near-record levels as a share of the economy and health care costs growing a bit more slowly than projected by the Congressional Budget Office (CBO), the official government scorekeeper. Even if all of the tax cuts approved under President George W. Bush—including those for the poor and middle class—were to expire in 2010, the Government Accountability Office (GAO) projects that by 2030, paying the bills for Social Security, Medicare, Medicaid and interest on the debt would take up 85 cents of every dollar collected in taxes.2 Even if the rest of government shrinks to near-record-low levels as a share of the economy, the remaining tax dollars would only cover one-third of the expense.3 By 2040, all of the taxes collected wouldn’t even be enough to cover the interest on the debt, Social Security, Medicare and Medicaid. With debt payments taking 27 cents out of every dollar in tax revenue, there would be nothing left for defense, infrastructure, education, the environment, veterans’ care or anything else. All of those basic needs would have to be charged on Uncle Sam’s credit card. This is the far more optimistic of the two scenarios mapped out by the GAO, and at least in one way it appears too optimistic: It envisions $5.5 trillion in less debt by 2019 than CBO’s analysis of the 2009 White House budget plan.4 That means the spending cuts would need to be deeper or the tax hikes steeper just to return to this dismal and dangerous path. Under the alternative path, with all tax cuts from the Bush presidency extended and no spending restraint, it would take 57 cents of every dollar in taxes just to pay interest on the debt in 2040.5 The key question is whether our elected officials can summon the will to face these challenges head-on with solutions that preserve good public policy. Or will they continue to delay action and put our economy, the safety net and even national security at risk? A failure to act would eventually trigger a different kind of crisis in financial markets—one that sends interest rates soaring and multiplies our problems. The Obama administration is correctly focused on the goal of bringing down the rate of health care cost inflation, which is the primary driver of the explosive growth of Medicare and Medicaid. But while the importance of reining in health care costs trumps all over the long term, the only reasonable conclusion is that in the nearterm, every area of the budget needs to be streamlined for maximum efficiency— and what is essential must be paid for. There are four reasons why. First, even if policy makers are very successful in substantially curbing medical costs—and there’s no cause for complacency in that regard —these health care safety net programs could still double relative to the size of the economy by 2050.6 Second, any unnecessary debt we incur in the intermediate term will increase interest costs and make our large long-term budget gap even bigger. Third, as of this writing, financial markets were pushing up the cost of Treasury debt even as the economy languished. Proactive deficit reduction designed to kick in after the economy is back on its feet could lower the risk of a premature rise in interest rates that would work against recovery. Finally, our elected officials need to remember that good fiscal stewardship is among their highest obligations to the generations
INTRODUCTION
5
of Americans to follow. Their objective can’t just be to avoid a worst-case or even just a really bad scenario—however, there is plenty of cause for alarm that we could get the latter, if not the former. ‘‘In the not-too-distant future, policymakers will need to move well beyond their current positions on both taxes and spending and put everything on the table, in order to achieve deficit reduction that will be necessary—and unavoidable—if we are to avert serious long-term economic damage,’’ said Robert Greenstein, executive director of the liberal Center on Budget and Policy Priorities, in March in reaction to CBO’s projection that the nation would add $9.3 trillion in debt over the next decade under President Obama’s initial budget plan.7 Within this next decade, the very near-term budget debacle will barely dissipate before the longer-term entitlement pressures gather force. While near-term deficits are being fueled by plunging tax revenue, emergency spending and greater dependence on government safety nets like unemployment insurance, those factors are all projected to fade from view by 2012. At the tail end of the decade, CBO projects that tax revenue will be $1.1 trillion higher than in 2008. The only problem is that interest costs are projected to grow by $550 billion over that span, while entitlement spending costs balloon by $1.3 trillion.8 Having provided $72 billion in surplus revenues to the government in 2008, Social Security is projected to be a $101 billion liability for the U.S. Treasury in 2019, as the retirement of the first batch of baby boomers that began last year sets in motion a dramatic transformation in the age makeup of our population.9 Now, about 12.5 percent of our population is over 65. By 2030, that will reach 19 percent and continue to rise slowly thereafter. 10 The whole country will pretty much look like Florida—only older. Today, about 17 percent of Floridians are senior citizens.11 For the past three decades, we’ve had roughly 3.3 workers per Social Security beneficiary, which made the cost of providing income support for each retiree and disabled worker relatively manageable. But this key dependency ratio pointed downward last year as the first batch of boomers exited the work force. By 2010, there are projected to be 3 workers per beneficiary. By 2020, it will be 2.5 to 1. By 2032, just 2.1 to 1.12 At that point, Social Security’s existing revenue stream would only be enough to pay 76 percent of promised benefits.13 Medicare’s expense has already grown well beyond its dedicated funding source— a 2.9 percent payroll tax on all wage income—and its strains are being felt even before the first baby boomers turn 65 and collect benefits. But its challenges will be exacerbated by these same demographic trends that will unavoidably make both safety net programs more of a burden on future workers, somewhat less uplifting for future retirees, or a combination of the two. For Medicare and Medicaid, which provides long-term care for the aged and covers out-of-pocket Medicare costs for millions of seniors in addition to serving the non-elderly population, the most desirable way to rein in the programs’ growth is simply to curb the rate at which health care prices increase, to the extent possible. Such a process, which is closely tied to broader reform of the nation’s health care
6
A WELL-TAILORED SAFETY NET
system, may be one that takes a number of years to play out and show clear signs of success. Social Security reform, on the other hand, presents a fairly straightforward and modest undertaking relative to remaking the health care safety net. Obama, expressing a widely held view, told the Washington Post that Social Security reform is ‘‘easier’’ to do. ‘‘I think that Social Security we can solve.’’14 But that’s not to say it will be easy. Undoubtedly, Obama’s coming attempt to reform Social Security will present one of the stiffest tests of whether he can change the way that Washington works—or, rather, too often fails to work. Social Security is often described as ‘‘the third rail of American politics’’—a reference to the electrified rail on commuter trains that can deliver a lethal jolt to those who touch it. And most lawmakers on both sides of the aisle have done their best to keep their distance. They are quick to criticize any reform idea the other side may offer, but refuse to offer their own. Or they throw out an idea that might sound appealing but that has absolutely no chance for support and would do little to address the underlying financial problems. Or they simply point to other problems that they say require more urgent attention. This is how they have managed to sidestep responsibility for more than a decade. Since 1994, no less than four bipartisan commissions have warned of an urgent need to reform Social Security; both President Bill Clinton and Bush made it their highest priority; and dozens of members of Congress have proposed legislative fixes. But there has been absolutely nothing to show for all this effort—not even a single vote on a major Social Security reform bill. Meanwhile, Social Security’s unpaid future bills have continued to grow as the due date gets ever closer, leaving future retirees and taxpayers to bear the cost of inaction.15 When Bush put Social Security reform on top of his agenda in 2005, most of those Democrats who were even willing to concede that a serious problem exists suggested that tax increases for the wealthy could solve it. Meanwhile, most Republicans rejected any tax hikes out of hand, with those who acknowledged the need for any sacrifice preferring to close future shortfalls with benefit cuts. With very few exceptions, lawmakers from both parties were fixated on ideological solutions that precluded the possibility of finding common ground. As House Majority Leader Steny Hoyer said in a May speech, ‘‘both sides are guilty—Democrats for using Social Security as the ‘third rail’ for political advantage, and Republicans for walking away from the table at the first mention of raising revenues. Neither side has been willing to put forward realistic, effective alternatives.’’16 It is a welcome sign that a Democrat of such high stature is trying to foster a more constructive attitude among lawmakers from both parties, and one of my goals in writing this book is to help hold their feet to the fire. I’ll take apart all of the leading reform ideas to drive home that Social Security reform defies any easy answers offered by members of both political parties, so that after this book, anyone who wants to understand this debate will be able to judge whether lawmakers are giving them the old song-and-dance.
INTRODUCTION
7
But the main purpose of this book is to point the way to common ground, and that process must begin by stating some hard truths about Social Security that one side or the other has been unwilling to acknowledge. While Social Security isn’t the biggest or toughest fiscal challenge we face, it is plenty big, and the cost of delaying reform is unacceptably high. As I will explain in chapter three, simply paying all promised benefits between 2016 and 2036—just before Social Security’s trust fund is exhausted—is projected to raise federal debt levels by $5.4 trillion (in 2009 dollars) under present law.17 The interest on this debt created when the Treasury redeems the special-issue bonds in the trust fund could leave a budget gap even bigger than the official one facing Social Security in 2037 and beyond. This reality, which reveals the perils of the government’s trust-fund accounting, suggests that almost any approach to aligning Social Security’s costs and its resources would be preferable to staying on the present course. However, with national debt levels rising at a dangerous pace, the challenges much bigger and harder to solve outside of Social Security, and the political hurdles to raising taxes always hard to surmount, there’s likely a limit to how much extra revenue Social Security can bring in without limiting funds for other critical programs. The reality is that the benefits that have been promised may be unaffordable—even with significant tax increases. But that doesn’t mean we can afford to simply slash the safety net. Why not? For one of the principal reasons that we face such a daunting challenge: The average retiree will live past 85 and a huge percentage will live well into their nineties. In 2000, there were four million Americans who were at least 85 years old. By 2030, there will be nearly nine million. By 2040, 14 million. By 2050, 19 million.18 Without an effective safety net, far too many of these senior citizens would be denied both comfort and dignity in their twilight years. In short, the government needs to scale back future spending on our retirement safety net, even as rising life expectancies put future retirees at growing risk of depleting their savings, leaving them to scrape by on a Social Security check that is far from generous to begin with. Looked at in that light, it’s not hard to understand why this debate has gone nowhere and aroused such bitter partisanship. The problem isn’t simply one of politics but of policy. Taking into consideration both increasing longevity and the disappearance of defined-benefit pensions that provide a steady income stream through the end of life, the ultimate test for any Social Security reform plan must be whether it ensures a robust safety net in very old age. But every serious proposal on the table fails that test by prescribing significant benefit cuts for retirees whether they are 65 or 95 and, in some cases, adding additional uncertainty from stock-market risk. Such benefit cuts are untenable when Social Security—even if we could afford all of its promises—will fall short of providing a strong and dependable safety net in very old age because many workers face severe early-retirement penalties that will reduce benefit checks by up to 30 percent. This cold-eyed analysis, however, does not suggest that efforts at Social Security reform are futile or misdirected. Rather, understanding exactly what is at stake is
8
A WELL-TAILORED SAFETY NET
critical to balancing the necessary trade-offs, making the most effective use of Social Security’s limited resources, and producing a safety net that is both affordable and effective. The reason I decided to write this book is because I discovered through the most painstaking analysis that there’s really only one approach to saving Social Security that is well-tailored to meet both the nation’s financial challenges and the responsibility of providing a robust safety net for our aging population. Simply put, we face a choice between giving up on the promise of income security in very old age or scaling back the promise of a government-financed early retirement for all. While the correct choice should be obvious, any changes to the retirement safety net must be made with the utmost care because they could place a disproportionate burden on the most vulnerable. Under the proposal detailed in A Well-Tailored Safety Net, the safety net would firm up earlier in retirement for low earners than high earners, but benefit cuts would unwind to provide strong support for retirees of all income levels. There’s no other way we can limit the need for tax hikes and still preserve a Social Security program that meets the needs of both the working class and the broad middle class. The alternatives either cut away critical protections, depend on bigger tax increases, or a combination of both. It is this point that I am making with the book’s subtitle—The Only Fair and Sensible Way to Save Social Security. But there is a limit to how much of Social Security’s shortfall can be closed by scaling back the promise of early retirement. Additional resources are also needed to preserve an effective safety net, and the subtitle is not intended to suggest that there is only one reasonable approach to raising new revenue for Social Security. Indeed, A Well-Tailored Safety Net offers two different versions of reform: one that incorporates a change in the taxation of employer-provided health insurance and a Plan B introduced in chapter 16 that does not. Because of the uncertain path of health care costs, it is the latter proposal that I use to compare the impact of A Well-Tailored Safety Net to that of other serious proposals. What will emerge is a reform plan that goes further than any other to ensure a robust safety net in very old age, yet does considerably more than any other to reduce the nation’s debt while requiring less in the way of tax hikes than other plans that seek to preserve a viable safety net. Beyond simply restructuring benefits, A Well-Tailored Safety Net introduces a complementary set of policies to help workers adapt to less support from Social Security in the earliest years of eligibility and retire with maximum support. Benefits would ramp up as workers wind down their careers; employers would be rewarded for providing older workers with flexible, low-intensity jobs; workers would be automatically enrolled in a new supplemental savings program; and the incentives for delayed retirement would be significantly enhanced. Following a 2005 debate that did more to obscure than to illuminate the serious trade-offs involved in Social Security reform, A Well-Tailored Safety Net attempts to relaunch and transform this debate by proposing commonsense benchmarks for what constitutes a fair, progressive, and fiscally responsible reform of Social Security.
INTRODUCTION
9
Fairness means that sacrifice should be roughly proportional to income, instead of putting the full burden on the highest earners or asking those with little to spare to shoulder an equal share of the sacrifice. But there’s nothing progressive about cutting the safety net for retirees in very old age, when just about everybody will depend upon it. Because Medicare and Medicaid present a far more daunting budget challenge, it’s only responsible that we strive to pay for our Social Security safety net without indebting future generations. But, as I’ll explain, that logical debtreduction target is far more ambitious in scope than just about anyone in Washington has been willing to embrace. An affordable and effective safety net that treats people fairly is really the only logical goal for Social Security reform, and it is likely the only goal that has the potential to bridge the political differences between Republicans and Democrats. But up until now, there haven’t been any Social Security reform proposals that reflect those priorities. That’s not to say the approach I developed can’t be improved upon. I’m quite certain it can—if the very accomplished people who have wrestled with Social Security reform put their minds to it. But the principles that undergird A Well-Tailored Safety Net cannot be compromised if we want to have an effective and enduring Social Security system. I will briefly summarize A Well-Tailored Safety Net’s central provisions after the introduction, and in later chapters I will lay out both the primary proposal and Plan B in detail. But before I do, I want to talk a little about what is at stake and why I have devoted so much time to this problem. The issue was so important to me because, in my mind, Social Security reform offers so much promise, as well as so much peril. The sad truth is that even though Social Security is the single biggest government program—costing close to $700 billion in 2009—a low-wage worker who works a 40-year career can still receive a benefit well below the federal poverty level. While Social Security has done a great deal to combat poverty among senior citizens, about 8 percent of all beneficiaries above 65 had incomes below the poverty level in 2006, including 17 percent of unmarried women, 19 percent of Hispanics and 23 percent of African-Americans. For all women 80 and older, the poverty rate is close to 12 percent.19 Reform shouldn’t just be about saving money, but about making the most of Social Security’s resources. And in the case of those who have worked hard their whole lives but might nevertheless end up in poverty in old age, it means reinforcing the safety net with a new minimum benefit to help lift even the lowest earners who work a full career over the poverty level. My work also was motivated by my dismay that although Social Security is the primary savings vehicle for lower earners, it provides no opportunity for true ownership. Although Social Security can provide the equivalent of a generous inheritance in targeted cases, such as when a deceased worker leaves behind young children, there is a difference between this particular form of insurance and inheritance. While one is humane, the other—the product of hard work—is ennobling. But, too often, Social Security’s overly targeted opportunity for inheritance is neither—such as in the case
10
A WELL-TAILORED SAFETY NET
of a working class single mother who isn’t able to leave any savings to her children past the age of 18 who are trying to make their own way in the world. What is more, the inability to pass along some of the fruits of their labors in the form of a modest inheritance may encourage workers to claim benefits at the first chance they get— even at the cost of income security in very old age. While advocates for low-income workers have viewed a broader right of inheritance tied to personal accounts within Social Security as a sort of Trojan Horse, I will show that A Well-Tailored Safety Net can provide a degree of real ownership without undermining the program’s social insurance function or exposing lower earners’ retirement nest eggs to unwarranted risk. I also think about Social Security in terms of my two sets of grandparents, one working class and one upper-middle class. Both lived to celebrate their 60th anniversaries, and—despite their vastly different financial stations—both came to rely on Social Security. Because of their experience I understand that Social Security reform must preserve a strong safety net in very old age for people of all income levels. The good news is that we can do all of these things—lift up the poor, provide a right of ownership, and preserve the safety net for both lower earners and the broad middle class—while reforming Social Security in a cost-effective way that helps our economy meet the responsibility of caring for our aging population. This book will explain how it can be done.
Highlights of A Well-Tailored Safety Net Social Security Reform Proposal
Nearly half of Social Security’s shortfall would be closed by scaling back the promise of early retirement in a progressive way—to avoid the need for benefit cuts in old age. Old-Age Risk-Sharing: Initial retirement benefits are reduced, but benefit cuts fully unwind over 20 years to ensure a robust safety net in very old age for retirees of all income levels. When fully phased in after 20 years, a low earner would face an upfront 10 percent benefit cut, which could be offset with an enhanced 10 percent delayed retirement credit for working one year past the Normal Retirement Age. An average earner would face a 20 percent up-front cut that could be offset with credits from two extra years of work. High earners would face a 30 percent cut that could be offset with three extra years of work. Retirement age: The Normal Retirement Age would rise an extra year to 68, while the earliest retirement age would gradually climb from 62 to 64. In combination, these measures would reduce the maximum early-retirement penalty to 25 from 30 percent, meaning a retiree in very old age would get a benefit that is at least seven percent bigger than one who retired at 62 under the current system—even if no benefit cuts were necessary. For an average earner, that would mean the difference between annual benefits of $12,900 and $12,050 based on current wages.
12
A WELL-TAILORED SAFETY NET
Roughly half of Social Security’s shortfall would be closed by relying on additional saving and tax increases. Progressive saving offset: Average earners ($42,000 in 2009) would be required to save an extra 1 percent of annual income ($420) in a personal account within Social Security, while those earning 50 percent of the average ($21,000) would have to save an extra 0.5 percent of wages ($105), and those earning at least 150 percent of the average ($63,000) would have to save 1.5 percent of income. For most workers, Treasury-only accounts would largely negate the impact of a progressive saving offset, which will be phased in over 40 years. Above-average earners would have to save more on their own to fully overcome benefit cuts, but the provision expires at life expectancy, meaning that the safety net will be secure in very old age. Tax over the cap: Payroll above $106,800 will be taxed at a rate that reaches 2 percent. Taxing employer-provided health coverage: The growth of payroll-tax-exempt spending on employer-provided health benefits will be limited to the rate of GDP growth. Changes to the safety net must be made with the utmost care, and A Well-Tailored Safety Net introduces a comprehensive set of policies to lift up the floor of support for low earners and help workers adapt to less support from Social Security in the early years of eligibility. Poverty-related minimum benefit: Workers with 40 years of work will get a minimum base benefit equal to 120 percent of the wage-adjusted poverty level before adjustment for other solvency measures. Workers with 30 years of work will receive a benefit equal to 100 percent of the wage-adjusted poverty level, falling to 80 percent for 20-year workers. Phased-in benefit eligibility: Even after the earliest retirement age rises to 64, workers would still have the option of receiving a benefit that ramps up as they wind down their careers—starting at 62. Once the retirement age increase is phased in by 2028, they could still receive 50 percent of their benefit at 62, with the other half ramping up in steady increments through age 68. This would provide workers with a substantial and growing income supplement, while still shielding them from excessive earlyretirement penalties that would deprive them of income security in very old age. Payroll-tax reductions: As an incentive to provide flexible, low-intensity jobs for older workers, employers will be exempted from payroll taxes on the first $10,000 in wages for workers 62 and older after a phase-in period. The exemption would rise with future wage growth.
HIGHLIGHTS OF A WELL-TAILORED SAFETY NET SOCIAL SECURITY REFORM PROPOSAL
13
Supplemental savings: While only those earning at least 150 percent of the average wage ($63,000) would have to save an extra 1.5 percent of income, the full 1.5 percent of payroll will be collected from everyone. The excess amount would be deposited in separate accounts that could provide the foundation for a universal payroll-deduction savings plan that has bipartisan support. Individuals could be afforded pre-retirement access to these separate accounts. A Well-Tailored Safety Net introduces a new approach to personal accounts that involves no up-front borrowing, no risk and takes nothing away from Social Security’s guaranteed benefits, but still provides workers an opportunity to accumulate assets and leave a modest inheritance tied to 1.5 percentage points of their 12.4 percent Social Security tax, as well as new mandated contributions. Add-on accounts: As explained above, all workers would be required to save an additional share of wages in a personal account. With sacrifice roughly proportional to income, average earners will save an extra 1 percent of income; those earning half the average wage will save 0.5 percent of income; and those earning at least 150 percent of the average would save 1.5 percent of income. Account balances related to these add-on deposits would be drawn down in equal (inflation-adjusted) amounts through life expectancy, when the progressive saving offset expires. Delayed Retirement Accounts: Each worker’s account will be complemented by the equivalent of a risk-free Treasury portfolio equal to 1.5 percent of wages, providing an inheritable claim on trust fund bonds to families who would not already receive generous survivor benefits. For workers retiring by 65, this portion of the account will be cashed in for their regular Social Security annuity. Delayed retirement as a route to ownership: Workers who retire after the Normal Retirement Age (NRA) will receive lump-sum payments equal to one half of an enhanced 10 percent delayed retirement credit. Offering the incentive of ownership will encourage workers to delay retirement and, thereby, strengthen incomesecurity in old age. For work between 65 and 68, the default option should be for workers to receive the full increase in the form of bigger benefits; however, to the extent that modest earners may be more likely to extend their careers if the rewards of doing so are enjoyed immediately, it makes sense to provide workers a lumpsum option before the NRA.
Chapter 1
A ‘‘Very False Sense of Security’’
Think of the Social Security Trust Fund as a cookie jar where a couple tries to save for retirement by tucking away a $100 bill every week. But every weekend, they give that $100 to their kids to spend on movies, pizza, new clothes, and video games. Still, the couple doesn’t worry about their inability to save, because each time they remove a $100 bill from the cookie jar, they replace it with a piece of paper that says ‘‘You kids owe us $100 plus interest to be paid out in an annuity upon our retirement.’’ It’s brilliant. The kids are happy, because they get everything they want. The parents, who just can’t say no to their kids, don’t have to, and yet their retirement savings are secure—assuming their kids strike it rich and are happy to foot the bill. This is the same twisted logic that is behind the Social Security Trust Fund. Every year for the past quarter-century, Social Security has collected more in tax revenue than it has paid out in benefits, but Washington has customarily spent all of these extra tax dollars on other government programs and still managed to run a big deficit. Nevertheless, members of Congress insist that there’s no need to worry. Those extra dollars collected by Social Security—and spent by Congress—will ensure the payment of trillions of dollars worth of retirement benefits for aging baby boomers, they assure the public. Don’t feel bad if you can’t follow their logic. It takes members of Congress who can twist themselves into pretzels by taking all sides of an issue to really appreciate the brilliance of trust-fund accounting. Sen. Barbara Boxer, Democrat from California, ran for re-election in 2004 pledging ‘‘to fight to stop the government from using the Social Security Trust Fund to pay for tax breaks for the wealthiest Americans.’’1 She went on CNN and fretted about budget ‘‘deficits as far as the eye can see,’’ charging that President George W. Bush and the Republicans ‘‘have stolen from the Social Security Trust Fund . . . just when the baby boomers are going to start retiring.’’2 But by early 2005, Boxer was singing a different tune as she sought to undermine Bush’s case for revamping Social Security.
16
A WELL-TAILORED SAFETY NET
All is well, she basically declared, and she decried talk of a Social Security funding crisis as ‘‘scare tactics.’’ According to the most conservative estimates, Social Security will be able to pay full benefits for 38 years. In other words, a 37-year-old worker today will get full benefits until he or she is 75 years old if we do nothing to make adjustments to the Trust Fund. A 47-year-old worker today will get full benefits until he or she is 85 years old if nothing is done . . . If someone told you that your family would be in solid shape for the next 38 years, you would probably breathe a sigh of relief.3
In other words, Boxer was saying, the American people should have peace of mind because the same Social Security Trust Fund that the Bush administration had ‘‘stolen from’’ and used ‘‘to pay for tax breaks’’ for the wealthy would take care of them in retirement until 2041. Never mind that it’s now 2009 and the Social Security actuaries are projecting that the trust fund will be depleted in just 28 years.4 Because whether the trust fund exhaustion date is 28, 38, or even 60 years away is completely unrelated to the government’s ability to afford paying for the benefits it has promised. While the Social Security Trust Fund may have political and moral significance, ‘‘it has no economic significance whatsoever,’’ David Walker, the nation’s top accountant until last year, has testified before Congress.5 ‘‘There are no stocks or bonds or real estate in the trust fund. It has nothing of real value to draw down.’’6 All it contains is government IOUs, Walker, who was appointed by President Bill Clinton, has told Congress. ‘‘This is not a real trust fund. Let’s not kid ourselves.’’7 In theory, the trust fund balance represents the funds that the government has saved in years that the taxes paid into Social Security have exceeded the benefits paid out to retirees, spouses, disabled workers, and dependents. That’s happened in every year since 1983, when Social Security was last operating in the red. Back then, with the trust fund nearly empty, Washington addressed the crisis by raising taxes and trimming benefits. And from 1984 through 2008, the government collected about $1.2 trillion more in Social Security taxes than it paid out in benefits.8 But Washington spent the entire $1.2 trillion and still ran budget deficits of $4.3 trillion over that 25-year span, so in no real sense has the nation set aside funds to pay retirement benefits for decades to come.9 Yet, as the government spent those excess Social Security funds, it issued special bonds that were deposited in the trust fund. Rather than providing real resources to pay future benefits, the bonds merely show how much of a debt the government owes to Social Security—a government agency. While these bonds are backed by the full faith and credit of the U.S. government, they have no intrinsic value to taxpayers because they represent both an asset and a liability of equal amounts. In effect, the bonds are IOUs that the government has been writing to itself—IOUs that can only be repaid by future taxpayers. Like the government debt held by the public, these IOUs collect interest—$116 billion in 2008 alone. But this interest only serves to compound the illusion of the
A
‘‘VERY
FALSE SENSE OF SECURITY ’’
17
trust fund. When the government pays interest to holders of public debt—individuals, pension plans, and foreign governments—it is a transaction between two parties and the government recognizes it as a real expense. Those interest payments to public debt holders ($253 billion in 2008) come out of the bottom line and affect how much the government has available to spend without borrowing more money or raising taxes. But when the government pays interest on the IOUs in the trust fund, there are no consequences—that is, until Social Security begins running cash deficits year after year starting in 2016 and the bill starts coming due. In the 2008 budget, the $116 billion in interest that the government paid to Social Security was fully offset in the budget by the $116 billion that Social Security received from the government, with no impact on the bottom line. But, as seen in table 1.1, because of all that accumulated interest, by the end of 2008 the trust fund had grown not just to $1.2 trillion—the amount of excess Social Security taxes collected but spent on other government functions over the years—but to $2.4 trillion. It bears repeating: not only has our government failed to set aside the $2.4 trillion in the trust fund that many of our political leaders say will pay future retirement benefits, but Washington, over the past 25 years, has instead saddled future generations with an additional $4.3 trillion in debt—an amount that is very likely to be matched from 2009 to 2012, based on the White House’s own projections. Far from saving to help defray the cost of retirement benefits for the baby boom generation, the government has been spending well beyond its means. And despite the fact that the government is projected to continue running massive deficits, the trust fund is projected to grow to an eye-popping $4 trillion by the end of 2018.10 So should families ‘‘breathe a sigh of relief ’’ as Boxer suggested? Hardly. There is no huge pile of money that will somehow make it possible to pay for what would otherwise be unaffordable. All the trust fund provides, as ex-Comptroller General Walker has said, is a ‘‘very false sense of security.’’11 Bush argued in 2005 that the trust fund was essentially meaningless, and under his watch, it clearly was: From 2002 through 2008, the trust fund grew by $1.2 trillion even as the government ran deficits of $2.1 trillion. But even under Clinton, who presided over the best of economic times but left office with a faltering economy, the trust fund’s $834 billion increase from 1994 to 2001 was mostly a mirage, as the government ran a cumulative $63 billion surplus during that span. But regardless of who’s to blame, the facts couldn’t be any clearer. So how can Democrats insist that the trust fund will delay Social Security’s financial crisis until 2037? Not to worry, Nancy Pelosi, then leader of the Democratic minority in the House, told George Stephanopoulos on ABC’s This Week in 2006, Democrats have a very simple plan, ‘‘Pay the trust fund back the money that the president has taken from the trust fund,’’ Pelosi said.12 While it is significant that Pelosi acknowledged the duty of Congress to truly set aside resources if it was going to count on the trust fund to pay future retirement benefits, her idea of paying the trust fund back appeared divorced from reality—even before the economy collapsed. Consider what Pelosi was saying. In 2007, the trust fund grew by $190 billion. For that increase to have been meaningful, the
18
A WELL-TAILORED SAFETY NET
Table 1.1 Washington piles up debt, as it pretends to save (billions of dollars)
a
YEARa
Budget deficit/ surplus (includes Social Security surplus)
Cumulative deficits 1984– 2008
Social Security non-interest surplus
Trust Fund interest
Cumulative Trust Fund increase 1984– 2008
1984
(185)
(185)
3
3
6
1985
(212)
(397)
10
3
19
1986
(221)
(618)
11
4
34
1987
(150)
(768)
17
5
56
1988
(155)
(923)
33
8
97
1989
(153)
(1,076)
41
13
151
1990
(221)
(1,297)
45
17
213
1991
(269)
(1,566)
34
22
268
1992
(290)
(1,856)
25
25
319
1993
(255)
(2,111)
19
28
366
1994
(203)
(2,314)
27
31
424
1995
(164)
(2,478)
25
35
484
1996
(107)
(2,585)
32
39
555
1997
(22)
(2,607)
45
44
643
1998
69
(2,538)
58
49
750
1999
126
(2,412)
78
55
884
2000
236
(2,176)
89
64
1,037
2001
128
(2,048)
90
73
1,200
2002
(158)
(2,206)
85
80
1,366
2003
(378)
(2,584)
68
85
1,518
2004
(413)
(2,997)
67
89
1,674
2005
(318)
(3,315)
78
94
1,846
2006
(248)
(3,563)
87
102
2,036
2007
(161)
(3,724)
80
110
2,226
2008
(459)
(4,182)
64
116
2,406
Budget data reflect fiscal year; trust fund data reflect calendar year. Sources: Congressional Budget Office, Historical Budget Data, Table F-1, Revenues, Outlays, Surpluses, Deficits and Debt Held by the Public, 1969 to 2008, http://www.cbo.gov/ftpdocs/99xx/doc9957/Historicaltables09-web.XLS; Social Security Administration, Trust Fund Data, Old-Age, Survivors, and Disability Insurance Trust Funds, 1957–2008, updated January 29, 2009, http://www.ssa.gov /OACT/STATS/table4a3.html.
A
‘‘VERY
FALSE SENSE OF SECURITY ’’
19
government would have actually had to save those funds by running a budget surplus at least reasonably close to $190 billion. And to begin to ‘‘pay the trust fund back’’ a meaningful portion of the $2.4 trillion that the government has failed to save would have required a budget surplus of well above $190 billion. In reality, at the highwater mark of the economic cycle, the government still ran a $161 billion deficit in 2007. While Democrats called for rolling back the Bush tax cuts for the highest earners, that wouldn’t have generated much more than $60 billion in revenue and all of the party’s leading presidential contenders pledged to use those funds to pay for their health care plans. And even if there were no Iraq War—which was then costing $10 billion to $12 billion a month—then instead of a $161 billion deficit in 2007, the budget would have been close to balanced, but there still would have been zero saving for Social Security. Even in relatively good economic times and ahead of the fiscal strain of the baby boomers’ retirement, we weren’t able to set aside anything at all for Social Security, and both parties’ presidential contenders were more interested in spending on their other immediate priorities than saving for Social Security’s future. Of course, since then, deficits have exploded. The Congressional Budget Office (CBO) projects that the government will add $1.6 trillion to the debt in fiscal 2009 and roughly $9 trillion over the course of the decade.13 Now, almost nobody is seriously talking about even balancing the budget, much less paying the trust fund back. When Social Security starts regularly paying out more in benefits than it collects in taxes in 2016, those trillions worth of IOUs in the trust fund won’t do anything to make benefits more affordable. With or without the trust fund, the level of tax hikes, borrowing and spending cuts required for the government to make good on Social Security’s promises is exactly the same. This is not a matter of dispute, but rather a well recognized fact. Here’s what Clinton’s 2000 budget said: Social Security Trust Fund accumulations ‘‘do not consist of real economic assets that can be drawn down in the future to fund benefits . . . The existence of large trust fund balances, therefore, does not, by itself, have any impact on the Government’s ability to pay benefits.’’14 Here’s President Barack Obama’s 2009 budget: ‘‘The holdings of the trust funds are not assets of the Government as a whole that can be drawn down in the future to fund benefits. Instead, they are claims on the Treasury.’’15 Here’s what Social Security’s two public trustees, John Palmer, a Democrat, and Thomas Saving, a Republican,—both Clinton appointees—wrote in March 2007 in the annual report on the program’s financial status: ‘‘While current trust fund reserves provide the authority to cover the first $2 trillion of this funding shortfall before being depleted, Treasury must still come up with this amount in future cash.’’ .16 In an April 2007 Wall Street Journal opinion piece, Rep. Jim Cooper, Democrat of Tennessee, wrote that politicians of both parties—in a sort of conspiracy to prevent bad news on the budget from reaching voters—‘‘[p]ledge to protect Social Security and Medicare ‘trust funds’ without hinting that those trust funds do not exist.’’
20
A WELL-TAILORED SAFETY NET
Cooper continued, ‘‘Future beneficiaries have no real security other than the hope that future workers will pay more payroll taxes.’’17 And even AARP, formerly the American Association of Retired Persons, recognizes that despite the trust fund balances, 2016—the date when Social Security will begin needing extra cash, year after year, to keep paying promised benefits—is a significant turning point. ‘‘After more than 30 years of borrowing from Social Security, the U.S. Treasury will be called upon to transfer cash resources back to the trust fund,’’ John Rother, who directs AARP’s public policy efforts, told Congress in 2005.18 And at that point, as Rother acknowledged, ‘‘the rest of the budget may be in trouble due to unsustainable fiscal policies.’’19 Now, we know for certain that the rest of the budget will be in far more dire trouble than anyone could have imagined just a few years ago. By 2019, even as Social Security needs an extra $101 billion from the Treasury to pay all promised benefits,20 CBO projects that the rest of government will face a deficit of $1.14 trillion under Obama’s 2009 budget plan.21 By then, if nothing is done beforehand, even the most diehard defenders of trust fund accounting may be ready to admit that Social Security faces an immediate crisis. Beneath the headline number that shows a multi-trillion-dollar trust fund, there are actually two separate trust funds—one that finances benefits for disabled beneficiaries until they reach the retirement age and the much bigger one that pays benefits to retirees, spouses and children of deceased parents. In March 2009, CBO moved up its projected date for depletion of Social Security’s Disability Insurance Trust Fund to 2019, after which the payment of full benefits for disabled beneficiaries would be in doubt.22 Beyond 2019, even as the rest of the budget is being stretched to the breaking point, Social Security would need ever-larger infusions of cash. By 2020, Social Security’s annual shortfall will exceed $100 billion (in 2009 dollars), and that will grow to $200 billion a year by 2025 and $300 billion by 2032. The total projected shortfall from 2016 through 2036—just before the trust fund runs out—is $4.2 trillion.23 Those who point to the existence of the trust fund to downplay the urgency of Social Security reform are simply choosing to ignore this $4.2 trillion reality at the expense of those who will be paying taxes, retiring and relying on government services in 2037 and beyond. Those poor suckers would be the ones to bear the full burden of higher taxes, less generous retirement benefits, and a government whose growing debt load erodes its ability to pay for basic services.24
Chapter 2
‘‘A Political Fraud’’
Just five years after Congress reached a bipartisan deal in 1983 to shore up Social Security’s finances, the program’s trustees called together more than 100 policy experts to discuss an entirely different problem: How to handle—and protect—the trillions of dollars that actuaries were now expecting to accumulate in the trust fund.1 ‘‘It does no good to save for the future with one hand but spend it with the other,’’ warned Edward Gramlich, who would later serve as director of the Congressional Budget Office and Federal Reserve governor.2 ‘‘If we don’t insulate the [Social Security] surplus, then we are at least on the verge of committing something that looks to me like a political fraud,’’ cautioned Herman B. Leonard, professor of public management at Harvard University’s Kennedy School of Government.3 Their comments make clear that the predicament we’re in now didn’t catch anyone off guard. Washington has been worrying for two decades that the government wouldn’t really save those trillions of dollars in the trust fund that would be necessary to pay for the benefits promised to baby boomers. In 1990, the late Sen. Daniel Patrick Moynihan, a New York Democrat who had been instrumental in rescuing Social Security seven years earlier, was ready to junk the 1983 agreement. Fed up that the government was continuing to run sizable deficits and spending the extra money flowing in from Social Security, Moynihan proposed rolling back the payroll tax increase Congress had approved. Moynihan was frustrated because instead of saving the extra payroll tax revenue, Washington was using it to pay for other government programs, a practice he blasted as ‘‘thievery.’’ Sen. John Heinz, Republican of Pennsylvania, took issue with Moynihan’s characterization, ‘‘Certainly not. It is not thievery, it is embezzlement.’’4 Without the excess funds from Social Security, Moynihan reasoned that Washington would have to raise income taxes, which he saw as a far more acceptable source of funds for most government programs. Why? Because while income taxes are
22
A WELL-TAILORED SAFETY NET
progressive—with higher earners facing higher tax rates—when it comes to Social Security taxes, the highest earners face the lowest tax rates and the working class faces the maximum rate. ‘‘Because we are looting the Social Security trust fund and using it to fund the ongoing government, we are asking the wage earners to pay more than they should,’’ Sen. Kent Conrad, Democrat of North Dakota, said at the time.5 Though Moynihan’s measure to reduce the payroll-tax rate fell short of the 60 votes needed to overcome a filibuster in October 1990, it was supported by 43 of 56 Democrats, including liberal senators like Tom Harkin of Iowa, Barbara Mikulski of Maryland and John Kerry of Massachusetts.6 These same lawmakers have argued in recent years that the trust fund will help pay Social Security’s bills for decades to come. But Moynihan’s bill was an admission that Washington simply did not have the capacity or discipline to save trillions of dollars on behalf of future retirees—in other words, it could not be trusted. Moynihan’s bill would have turned Social Security strictly into a pay-as-you-go program, meaning that payroll tax rates would have been temporarily reduced to a level that would pay for current benefits but wouldn’t leave a surplus. His bill would have given up the misleading notion that the government was setting aside real resources that would help pay for the benefits of baby boomers. At most, the trust fund reflects the possibility that the nation would have incurred even more debt over the past 25 years if Social Security hadn’t been generating surpluses. But there is certainly plenty of evidence that the extra money from Social Security has facilitated a more easygoing attitude toward fiscal responsibility, reducing pressure on Congress to restrain spending, raise taxes or limit the size of tax cuts. That’s not to say that Washington, over the past quarter-century, has acted without any semblance of fiscal responsibility. After all, both President Bill Clinton and the first President George Bush raised taxes. But it’s worth remembering that the budget deal that Clinton reached with the Republican-led Congress in 1997 set a goal of balancing the budget by 2002—after Clinton left office. Some of the credit for the budget surpluses of the late 1990s belongs to a windfall of tax revenue from an economy with a 4 percent jobless rate that was supercharged by a roaring stock market and the exploding growth of the Internet. When a budget surplus of $70 billion unexpectedly emerged in 1998—the first surplus since 1969—both Democrats and Republicans were quick to celebrate and claim credit. But if Congress hadn’t counted the Social Security surplus as part of the budget, there would have been a $30 billion deficit. 7 As top Clinton administration economists noted in their retrospective analysis of fiscal policy during the 1990s, ‘‘taking Social Security out of the budget . . . would have imposed too much fiscal stringency, and it would have required the Administration to give up talking about its proudest accomplishment, the unified budget surplus and instead report an on-budget deficit.’’8 If not for the extra funds flowing to the Social Security Trust Fund, the $126 billion surplus in 1999 would have been a paltry $1 billion surplus. Just imagine. If the government in 1999 had just barely eked out a surplus after three decades of deficits, is it likely that George W. Bush would have campaigned for president in 2000 on a
‘‘A
POLITICAL FRAUD’’
23
platform of massive tax cuts? Is it likely that Congress would have passed such massive tax cuts the following year as the economy was showing signs of weakness and the good budget news was beginning to evaporate? Perhaps not. But the inclusion of those extra funds that were supposed to be saved for Social Security made the budget surpluses look so large that a huge tax cut was hard to resist. While campaigning in 2000, Bush pledged to protect Social Security’s surpluses. But in 2005, after his re-election, Bush explained that just like any other tax dollar, Social Security tax dollars are available to be spent. His comments suggest that he didn’t attach any special significance to the extra Social Security funds once in office. ‘‘What happens is we take your money, we pay money out for the promises for those people who have retired, and if we’ve got anything left over, we spend it on things other than Social Security,’’ Bush said. ‘‘That’s just the way it works. It’s been working that way for a long period of time. And what’s left are a pile of IOUs, paper.’’9 Democrats have long complained that Bush looted the trust fund, and even members of his own party criticized his fiscal stewardship. But consider that in 2004, Democratic presidential nominee John Kerry made exactly the same promise with regard to fiscal responsibility as Bush—to cut the budget deficit in half by 2009. While Kerry would have rolled back some of the 2001 and 2003 tax cuts, he would have devoted those funds to spending on health care. And once Democrats regained control of Congress in 2006, they set the same fiscal goal as the White House— achieving a balanced budget by 2012. But budget analysts criticized both sides for failing to map out a realistic path to reach that target—even if economic growth remained strong. While Democrats may have greatly different priorities than Republicans when it comes to taxing and spending, there’s been little evidence in recent years that they would do a better job of protecting Social Security’s surpluses. And why should they stick their necks out? According to the current Democratic Party line, no matter how much looting of the trust fund has occurred, it won’t in any way impact the ability of Social Security to pay all promised benefits. And it’s not just Democrats who have feltthat way. As former Rep. Clay Shaw, Republican of Florida, put it at a 2002 hearing on Social Security, ‘‘The surplus is a surplus regardless of whether we spend it or don’t spend it.’’10 The implication of this stance is that it is of little consequence whether or not the government saved the money it was supposed to be tucking away to pay future Social Security benefits. And if members of Congress believed that to be the case, it helps explain why the government has run deficits in all but four of the past 25 years. Because the trust fund, by law, only holds special–issue government debt, the only way to set aside resources has been to use Social Security’s surpluses to reduce the amount of debt that the government owes to other parties. The idea is that if the government reduces its external debt and can therefore avoid spending hundreds of billions on interest payments each year, it would be in a better financial position to make good on Social Security’s promises once the program’s costs overtake its tax revenue. But, as we’ve seen, the government’s deficits throughout the 1980s and most of the 1990s had cast doubt on the merit of the trust fund as a saving vehicle. So policy
24
A WELL-TAILORED SAFETY NET
makers began to look for alternative ways of truly setting aside—and growing—real resources to help finance future benefits and limit the size of the tax increases and benefit cuts needed to keep Social Security solvent. Three possible ways to create real savings within Social Security and pre-fund a portion of future benefits were laid out in a January 1997 report by the bipartisan advisory council that was appointed by the Clinton administration. No approach was backed by more than six members of the 13–member panel, and their three different proposals more or less defined the fault lines in the debate over the next decade. But all three proposals involved investing payroll-tax dollars in the stock market in order to set aside real resources that could be used to fund future retirement benefits.11 One group wanted the government to invest a portion of Social Security’s surpluses in the stock market. Their hope was to earn a better rate of return than the interest earned by the government bonds in the trust fund, thereby reducing Social Security’s future shortfalls. The other two groups both preferred to have individuals invest some of their payroll-tax contributions, but disagreed on where the money should come from. One group wanted to establish accounts by raising the 6.2 percent tax on employee wages devoted to Social Security and directing only the funds from the added tax into an individual investment account that would supplement Social Security checks in retirement. The other group wanted to let individuals also invest a portion of the current 6.2 percent individual payroll tax, reducing the amount of money devoted to Social Security’s traditional structure in the hope that favorable stock market returns would compensate for benefit cuts.12 The return of surpluses at the end of the Clinton era made all three of these options appear more feasible. Because the government was no longer spending the surplus tax dollars flowing in to Social Security, there were real resources that could be used either to deposit in individual accounts or for the government itself to invest in stocks. The White House believed that a deal could be achieved because large projected budget surpluses would make it possible to devote more resources to retirement savings, thereby limiting the sacrifice of benefit cuts and avoiding—or at least deferring—tax increases. Clinton called on Congress to seize the moment to fix Social Security. ‘‘We have been given a gift, and we have to use it,’’ Clinton said in June 1998. ‘‘This is a wonderful moment, but it is a moment of responsibility that we dare not squander.’’13 He said the time was now to save Social Security for the 21st century, calling for legislation by mid-1999. ‘‘It would be unconscionable if we failed to act,’’ Clinton said on another occasion.14 But with no bipartisan consensus on how to set aside real savings or on how to align Social Security’s promises with its resources, the window of opportunity came and went. By 2002, recession, tax cuts and war had ushered in a new era of deficits. By 2004, the deficit hit a whopping $412 billion even with the Social Security surplus, and $567 billion without it. Once again, the government wasn’t just failing to
‘‘A
POLITICAL FRAUD’’
25
set aside resources for the baby boomers’ retirement, but it was piling up a huge debt that will burden future generations. When Bush made his big push for Social Security reform in 2005, there were still a dozen years of projected program surpluses remaining and at least a distant possibility of preserving some of those funds to defray part of the cost of baby boomers’ retirement benefits. But the unexpected tide of red ink stemming from the current economic and financial crises has washed away those hopes and brought an early end to the two-decade-long debate over what to do with the Social Security surplus. With the spike in unemployment taking a toll on payroll tax revenues, the surpluses are now projected to be quite modest before Social Security starts running permanent cash-flow deficits in 2016. There have been real challenges over the past quarter-century—the Cold War, recessions, financial crises and 9/11. But the best that can be said is that Washington dealt with some of the real challenges of the present while making the very serious challenges of the future even more difficult to overcome. What the $2.4 trillion in the trust fund really signifies is a failure of government: A failure to account honestly for the promises it makes; a failure to rein in promises it can’t keep; a failure to reach a consensus; and a failure to face reality. The pertinent question is what this failure means for Social Security reform. It would be possible for Washington to try and perpetuate the mythology of the trust fund and simply limit our efforts to closing $5.3 trillion of Social Security’s cumulative $7.7 trillion, 75-year deficit (in present value)—while ignoring the fact that the $2.4 trillion trust fund doesn’t contain any assets to help the government pay future benefits.15 Indeed, virtually every reform proposal offered by politicians on both sides of the aisle has focused on the $5.3 trillion target, treating the trust fund like money in the bank, rather than as a debt that must be settled. But if this is the type of reform that is embraced, then the administration and Congress will have forfeited a significant part of the opportunity presented by Social Security reform to help address the nation’s fiscal challenges. As I’ll explain in chapter three, ongoing interest payments on the debt incurred as Treasury makes good on the bonds in the trust fund could, by itself, create a budget gap as big—or even bigger—than the official one facing Social Security. Beyond this measurable fiscal impact is the less-predictable political impact of a reform effort whose promise to keep Social Security solvent depends on cashing in trillions worth of IOUs that one agency of government has written to another. The political danger lies in the near certainty that tax increases will be a part of Social Security reform. The question is whether it will be possible to sell a tax increase to save Social Security if Social Security isn’t really being saved, but instead will require additional trillions in taxpayer revenue—even as the government’s finances are stretched to the breaking point. In other words, politicians may have a difficult time making the case for ‘‘sacrifice and responsibility and duty,’’ as President Barack Obama put it, if they aren’t prepared to embrace honest accounting.16
Chapter 3
The Price of Delay
In April 2005, President George W. Bush visited the Bureau of the Public Debt in Parkersburg, West Virginia, to look inside the four-drawer filing cabinet that houses the Social Security Trust Fund and drive home the point it doesn’t have any real resources to pay for the promises that politicians have made. ‘‘There is no ‘trust fund’—just IOUs that I saw firsthand, that future generations will pay for either in higher taxes, or reduced benefits, or cuts to other critical government programs,’’ Bush said.1 On the Senate floor that same day, Sen. Jim DeMint, Republican of South Carolina, said, ‘‘I am afraid while the trust fund is a nice idea, it is no more real than Santa Claus or the Easter Bunny.’’2 DeMint’s colleague Sen. Debbie Stabenow, Democrat of Michigan, was incredulous: I am shocked to hear him say the people of America who have paid into the Social Security Trust Fund, the baby boomers, do not have a secured obligation by all of us . . . He is actually saying that for the folks who have paid in as baby boomers that we are not obligated to pay those benefits? I want to make it clear that we Democrats believe with all our hearts and souls we have a responsibility to pay and we will pay those obligations.3
Democrats talk about making good on the trust fund as a matter of principle, and it is a reasonable and important principle to make sure that people who have worked hard their whole lives have a significant degree of income security in old age. But members of Congress don’t only have an obligation to baby boomers; they have just as much of an obligation to the younger generations that will be paying taxes and relying on the government in decades to come. That is the obligation that our elected leaders have shirked for more than a decade as they’ve failed to come to grips with once long-term financial challenges that are now just around the corner and only made more difficult by a legacy of debt produced by circumstance, shortsightedness and divisive politics.
28
A WELL-TAILORED SAFETY NET
Democrats congratulated themselves in 2005 for ‘‘standing sentry’’ to defend Social Security from Republican plans to slash promised benefits or divert payroll taxes into personal investment accounts.4 And, for reasons I will detail in coming chapters, there was good reason to oppose those efforts. But the only real way to protect Social Security is by offering realistic proposals that would align its resources and its promises. Anything else is simply an embrace of the status quo, which is really a vote to pass on trillions in debt to younger generations and a vote to weaken Social Security. For it’s not possible to save Social Security by running up federal debt costs and making it even more difficult to afford a robust safety net. Already in 2008, interest payments on the public debt cost $253 billion. That’s about 50 percent of what the government collectively spent on all of the discretionary programs budgeted on an annual basis by Congress, including such areas as homeland security, education, job training, housing assistance, veterans’ health care, science, workplace safety, transportation, the environment and international aid. By 2019, interest on the debt could cost $806 billion a year, or 112 percent of what we’ll spend on all those discretionary programs, according to a Congressional Budget Office (CBO) analysis of President Barack Obama’s fiscal 2010 budget blueprint.5 The simplest way to think about the budget is in terms of the size of the economy, or Gross Domestic Product (GDP). This provides a view that isn’t distracted, for example, by the massive dollar amounts expended by the federal government, but it reveals when fundamental changes are taking place because an area of the budget —or the overall budget—rises or falls substantially in proportion to GDP. As seen in table 3.1, interest costs are projected to grow more than twice as fast as the economy, reaching 3.8 percent of GDP in 2019, while non-defense discretionary spending is projected to fall modestly as a share of GDP. Already, well before our entitlement spending problems reach crisis stage, our choice is becoming clear: either we plug the hole in our budget, or we won’t be able to afford basic necessities. And that’s why fiscal responsibility should be a critically important value for both liberals and conservatives; as federal debt spirals, the growing burden of interest payments means that spending cuts will need to be deeper and tax increases will need to be steeper than would have been the case if our elected officials hadn’t spent quite as far beyond their means. The $625 billion cost of Social Security in 2008 equaled about 4.3 percent of GDP, making it, along with defense spending in a time of war, the largest government programs. But as the retirement of 77 million baby boomers continues apace over the next two decades, CBO projects that Social Security will grow nearly 40 percent faster than the economy, reaching 6.2 percent of GDP in 2030. Over the same span, due to an aging population and rising health costs, Medicare and Medicaid are expected to grow even faster, from a combined 4.8 percent of GDP in 2008 to 8.0 percent in 2030. Thus, the total cost of the three programs would balloon from 9.1 percent of GDP today to 14.2 percent of GDP by 2030.6 Now consider that the annual tax revenue paid to the federal government has averaged just over 18 percent of GDP in the past four decades. If Social Security, Medicare and Medicaid cost a combined 14.2 percent of GDP in 2030, and interest
THE PRICE OF DELAY
29
Figure 3.1 The growing burden of interest on the public debt
on the federal debt costs 3.8 percent of GDP—CBO’s projection for 2019 under the Obama budget plan—and federal revenue is near its average of 18.3 percent of GDP, then the government would have almost nothing to spare for defense, homeland security, education, unemployment insurance or any other domestic program. Of course the government isn’t about to stop funding all those critical areas. So consider another troubling scenario. This reflects CBO’s long-term budget outlook from 2007, the most recent available at this writing, but updates it by factoring in CBO’s projection that public debt would rise to 82.4 percent of GDP in 2019 under Obama’s 2009 budget plan. Assume that all of the Bush tax cuts expire in 2011— including those for the middle class and poor—and taxes rise to 21.4 percent of GDP—the highest in U.S. history—by 2030. Assume spending on all government programs outside of the big three entitlements shrinks to a record low 7.7 percent of GDP, down from 10.3 percent in the last pre-recession year of 2007.
30
A WELL-TAILORED SAFETY NET
Even then—if nothing is done to scale back the commitments to Social Security, Medicare and Medicaid—the government would be running a deficit equal to 4.5 percent of GDP in 2030. That’s the equivalent of a $630-billion deficit in 2009. While that might not look frightening by the current standard of the present recession, it’s important to understand that tax revenues would be 6 percentage points higher as a share of GDP than at present under the CBO scenario—21.4 percent versus 15.5 percent today. As seen in Figure 3.2, by 2040, even if taxes rose still higher—to 22.4 percent of GDP—the deficit would reach 6.7 percent of GDP—equivalent to a $930-billion deficit today—and federal debt would be on the verge of spiraling out of control. No matter how pro-actively Washington responds, it appears certain that coming generations will have to wrestle with the most difficult financial challenges brought on by the retirement of the baby boomers, rising health care costs and increasing life
Figure 3.2 Projected federal outlays and revenue
THE PRICE OF DELAY
31
spans. But a failure to make the necessary hard choices—and make them soon—will shackle coming generations with massive debts that will eat away the foundations of government. ‘‘Not only haven’t we come close to paying for the government transfers we are scheduled to receive, but we plan to pay for them by dwindling almost to oblivion the rest of government that would serve our children and grandchildren,’’ says Eugene Steuerle, a tax policy official in the Reagan-era Treasury Department and a longtime Urban Institute fellow.7 Back in June 1998, President Bill Clinton, the first baby boomer elected president, called for immediate action ‘‘so that we baby boomers do not bankrupt our children in their ability to raise our grandchildren.’’8 And when he invited lawmakers from both parties to a White House summit on Social Security that December, Clinton reminded them that reforming Social Security wasn’t just about saving money, but also about providing additional help for those most in need of Social Security’s support. He noted that about 20 percent of elderly single women were living in poverty. ‘‘Those who think we can wait should never forget that fact,’’ Clinton said.9 Yet the common refrain among Democrats in recent years has been that Medicare is a much bigger problem and that addressing Social Security’s problems can wait. But it is precisely because the costs of Medicare and Medicaid threaten to overwhelm the government that Social Security reform is so urgent. While some will make the argument that we can act later to fix Social Security if efforts to rein in health care costs aren’t as successful as hoped, we can no longer afford such complacency when our federal debt burden is expected to approach dangerous levels by the end of this decade. While slowing health care inflation is the most essential prescription for our longterm budget problems, it is hardly a cure-all. In fact, the CBO calculates that demographic factors will be the biggest driver of entitlement spending increases for the next four decades.10 Even under the positive scenario that sees health care cost inflation slow by 1.5 percentage points a year, White House budget director Peter Orszag wrote in a May 2009 Wall Street Journal op-ed that Medicare and Medicaid would still double in size relative to the economy, reaching 10 percent of GDP by 2050— or more than half of average federal tax revenues in recent decades.11 Thus, even under this successful outcome—which is hardly assured—budget strains could become severe. At the least, there’s no reason to think that the government will have extra money to burn, and that means a failure to align Social Security’s costs with its revenues would only serve to drive up federal debt levels. Another excuse some offer for delaying Social Security reform is the possibility that Social Security’s projections are too gloomy and that the problem won’t turn out to be as bad as feared. Social Security actuaries project a wide range of possible financial paths for the program, though policy makers focus on the intermediatecost projections that I also use throughout the book. But some in Congress have sought to undermine the case for reform by focusing on a set of low-cost projections that measure the potential impact of stronger-than-expected wage growth, higher levels of immigration, more births and slower increases in longevity.
32
A WELL-TAILORED SAFETY NET
If all of those more optimistic projections come to pass, then Social Security’s shortfalls would be more manageable. But even under the best imaginable circumstances, the actuaries project that Social Security’s deficits would start just four years later, in 2020, and surpass $100 billion a year by 2027. On the other hand, under the actuaries’ high-cost assumptions, Social Security’s shortfall could plausibly reach $300 billion a year by 2025.12 There’s no reason that Social Security reform can’t accommodate the potential for better-than-projected outcomes by delaying or even reversing benefit reductions (or tax increases) if the program is in surplus. But this possibility is certainly no excuse for inaction when the potential cost of delay is so high. It’s also worth noting that Clinton administration economists saw a ‘‘very low probability’’ that all of the demographic and economic inputs would turn out better than expected. ‘‘The distribution of outcomes seemed weighted toward a worse, not a better, financial situation,’’ wrote Douglas Elmendorf, Jeffrey Liebman and David Wilcox.13 The third argument some make to downplay the urgency of reforming Social Security is that the trust fund will permit the payment of all benefits through 2036. To the extent that there is a near-term financing issue, it is Treasury’s problem, not Social Security’s, they say. Thus, the question isn’t whether we have resources available to pay all benefits, but how to address the cost of redeeming trust fund bonds through tax increases outside of Social Security reform. There are several reasons why this argument is counterproductive, if not disingenuous. Consider first that a number of Democratic policy makers have proposed devoting estate tax revenues to Social Security. While this would leave a hole in the rest of the budget, they argue that it makes sense because Congress would be less likely to abolish the estate tax if it is dedicated to a popular program like Social Security. The takeaway is that the political hurdles to tax increases are more surmountable in the context of saving Social Security than in the context of deficit reduction. It follows, therefore, that Congress is much less likely to raise taxes simply to deal with deficit spending that is indirectly related to Social Security. In other words, given the bigger budget strains outside of Social Security, it can be expected that trustfund financed benefits would simply be paid for with debt. To understand why this is a bad idea, consider what would happen under present law, if the government issues new debt to pay the $4.2 trillion in unfunded benefits from 2016 through 2036—just before the trust fund runs out. The additional interest costs would add another $1.2 trillion in debt by the end of 2036, for a total of $5.4 trillion.14 In this present-law scenario, the ongoing interest payments on this debt from 2037 and beyond would soak up a bigger portion of the budget than it would take to close Social Security’s official shortfall with additional tax revenue. While Social Security, by itself, faces a budget gap that averages 1.3 percent of GDP from 2037 to 2083, as shown in figure 3.3, the combined cost of paying to fix Social Security and covering the interest on the trust-fund-related debt would average 2.9 percent of GDP. 15 That is a luxury we can’t afford when we can reasonably expect the government to be swamped by soaring Medicare and Medicaid costs.
THE PRICE OF DELAY
33
Figure 3.3 The cost of doing nothing until the trust fund is exhausted
Based on the intermediate-cost projections that serve as a guidepost for policy makers, the benefits paid out by Social Security will exceed the money the program takes in from taxes by $100 billion in 2020. At that point, Social Security would only have enough tax revenue to cover 90 percent of benefits. Closing that shortfall with a payroll-tax increase would require raising the rate by 1.5 percentage points (to 13.9 percent), the equivalent of a $630 tax increase for an average earner making $42,000 in 2009.16 By 2028, Social Security faces a funding gap of $250 billion and could only afford 80 percent of benefits without taking money from the rest of the budget. If we closed that shortfall by raising the payroll tax rate, we’d have to raise it by 3.3 percentage points (to 15.7 percent), the equivalent of a $1,380 tax increase for today’s $42,000-earner.16 Even if these tax increases were split between worker and
34
A WELL-TAILORED SAFETY NET
Table 3.1 The annual cost of fixing Social Security if we start in . . . Year tax hike is imposed
Size of tax hike as % of GDP needed to close 75-year financing gap
Annual tax hike (based on 2009 GDP)
2009
0.98%
$138 billion
2016
1.09%
$154 billion
2037
1.85%
$261 billion
Source: Author’s calculations based on Social Security Administration’s cost, revenue, economic and interest rate projections in The 2009 OASDI Trustees Report.
employer, the net result would be about the same, since the employer could be expected to compensate for the heavier tax burden by paying less generous wages. The question that must be answered is who will bear the sacrifice. Almost nobody would suggest that it is fair to cut benefits for people who have already retired. Therefore, all the sacrifice has to fall on new retirees and taxpayers, and only modest changes would be fair to those just a few years from retirement, so benefit adjustments must be phased in gradually. All this means that time is not on our side. Every year we wait, we exempt more new retirees from sharing any part of the sacrifice, and we heap more of the burden on future taxpayers. Consider that if we had acted by the beginning of 2009, we could have closed Social Security’s full cash shortfall by imposing a tax increase over the next 75 years equal to just under 1 percent of GDP each year, or $138 billion in 2009. As seen in table 3.1, by waiting until 2016, the cost of the annual tax increase needed to close Social Security’s shortfall through 2083 would rise 11 percent to 1.09 percent of GDP. And by waiting until 2037, the necessary tax increase would rise by almost 90 percent, reaching 1.85 percent of GDP. That means waiting until 2037 would raise the extra tax bill by $123 billion a year based on 2009 GDP levels. Whether this extra $123 billion annual cost is paid in taxes, reduced benefits or more debt heaped on younger generations, this shows why the price of reforming Social Security is already so high that it would be irresponsible to let it continue to escalate. When we know that the country will face the most daunting fiscal challenges, it would be immoral for our political leaders to delay action and, in effect, make a choice to saddle the next generation with trillions in additional debt or much higher taxes than they would otherwise face. The truth is that if we don’t act soon and ask most baby boomers to share in the sacrifice, the chances of maintaining a robust safety net will be much slimmer. Because we can reasonably expect the government’s finances to get steadily worse, it would be reckless to pass up what may be our last chance to do what the past three bipartisan Social Security commissions all recommended. They all believed that we should truly begin setting aside some extra savings now in order to limit the tax increases and benefit cuts faced by future workers and retirees. While putting aside real savings in the near-term would require trade-offs, if we fail to seize this opportunity, all we’ll be able to do is try and contain the bleeding.
Part II
No Easy Answers
Chapter 4
‘‘A Bad Idea’’
Thank goodness for Republicans like Sen. Jim DeMint, who had the courage to forthrightly declare that the trust fund is no more real than the Easter Bunny. So what kind of sacrifice did he call for to close Social Security’s $4.2-trillion cash deficit between now and 2036? ‘‘Let’s not cut benefits. We don’t want to cut benefits. Let’s not raise taxes,’’ DeMint said. ‘‘Let’s save Social Security with real savings.’’1 And where would those ‘‘real savings’’ come from? Real borrowing would be the only sure source of funds under legislation co-sponsored by DeMint that would permit workers to deposit pretty much their full 6.2 percent Social Security tax in a personal account and invest it in stocks and corporate bonds.2 Under the 2005 proposal from fellow Republicans Sen. John Sununu and Rep. Paul Ryan, instead of just a $4.2 trillion hole, Social Security would create a $13.1 trillion financing gap between 2009 and 2036, at which time it would still be running a deficit of more than $700 billion a year. Add in the interest owed on this additional debt and the Ryan-Sununu plan could rack up $17.6 trillion in debt by 2036.3 Forget the Easter Bunny. DeMint’s idea for saving Social Security was to pull a 6foot-tall, talking rabbit out of a hat. Under Ryan-Sununu, all workers younger than 55 would have the option of depositing their payroll taxes in an investment account, rather than allowing the government to spend those funds. For workers just entering the workforce, the money saved up in those accounts would entirely replace the benefits from Social Security, assuming the stock market performs reasonably well by historical standards. But if it doesn’t, no problem: under this plan, the government would guarantee that individuals who participate in personal accounts will get at least as much as Social Security promises. If it sounds too good to be true, that’s because it is. The system is now designed so that the taxes of current workers pay the benefits of current retirees. Because the Ryan-Sununu proposal would divert hundreds of billions of dollars a year into
38
A WELL-TAILORED SAFETY NET
personal accounts, the government would have to find a huge new pile of money to pay benefits promised to 50 million beneficiaries, whose numbers are rapidly growing. On top of that, the government would still have to find the funds to close Social Security’s impending cash-flow shortfall that will, by itself, amount to $4.2 trillion by 2036. The Ryan-Sununu plan assumed that the government could cover a big part of this massive hole, in part, by phasing in federal spending cuts that would approach 1.5 percent of GDP after eight years, or close to 8 percent of the entire federal budget in a typical year. Based on projected GDP in fiscal 2009, such a spending cut would equal $200 billion.4 But apparently even the authors of the proposal had their doubts that Congress would be willing to go along with such spending cuts. As the Social Security actuary’s analysis of the proposal reveals, the transfers of Social Security contributions into personal accounts ‘‘would, however, not be contingent on achieving these reductions in actual Federal spending.’’5 When they issued their plan back in 2005, Ryan and Sununu conveniently ignored the fact that the government was already running deficits of several hundred billion dollars a year before fiscal challenges mounted with the retirement of the baby boomers, so even saving $200 billion wouldn’t leave Washington with a lot of money to burn. It also seemed to ignore the much bigger budget challenge associated with Medicare and Medicaid. ‘‘The Ryan-Sununu plan purports to solve the Social Security shortfall by shifting massive sums from the rest of the budget even though the rest of the budget has no money to spare,’’ concluded the liberal Center on Budget and Policy Priorities. ‘‘By counting on massive transfers from general revenues to Social Security, the Ryan-Sununu plan essentially wishes away the entire Social Security deficit.’’6 Of course, Republicans had little success reining in government spending when they controlled the White House and both houses of Congress, so there’s little reason to think they’ll be able to shrink the size of government with Democrats calling the shots and looking to invest more resources in education, health care and other areas. And now, in the current era of trillion-dollar deficits, it’s much harder to make an argument that all budget savings should be offset with new government borrowing to establish personal investment accounts within Social Security. The spending cuts assumed by Ryan-Sununu would have only closed part of the hole created by diverting hundreds of billions of dollars each year from Social Security into individual investment accounts. However, they also assumed that the Social Security funds invested in personal accounts would stimulate the economy to such a degree that corporations would generate much greater profits and, therefore, a whole lot more in tax revenue for the government. But the huge problem with this assumption is that the extra corporate taxes won’t materialize unless Congress gets its fiscal house in order, which would be much more difficult to do if RyanSununu became law. According to basic economic theory, the diversion of Social Security taxes into personal accounts would do nothing to stimulate faster economic growth if it is financed
‘‘A
BAD IDEA’’
39
by an increase in government debt. In that case, the additional amount of money available for investment in the corporate sector would be exactly offset by the amount of private investment dollars that would be needed to purchase new government debt. Thus, there would be no net increase in the pool of investment dollars available for the private sector and no net increase in national savings. Only real increases in national savings can bring down the cost of borrowing for corporations, making it easier for them to expand and fueling the economy. Ryan and Sununu assumed that investment in personal accounts would stimulate the economy to such a degree that corporate tax payments would increase by 1 percent of GDP a year within 25 years. Add in the saving from unspecified cuts in government programs, and they assume that 2.5 percent of GDP a year would be freed up for personal accounts by 2035. Abracadabra. Allakhazam. There’s $550 billion.7 Quite a trick. But if these savings don’t magically materialize, the government would have to borrow an extra $12.2 trillion by 2036 on top of the trillions in borrowing that would be required to pay all promised Social Security benefits. Would the so-called ‘‘conservative’’ brain trusts behind this plan still support it if the government savings aren’t obtainable? Apparently, they didn’t see that as a problem. ‘‘Another proposed option is to fund the entire transition [to personal accounts] through long-term borrowing,’’ Stephen Moore and Peter Ferrara, who is credited with developing the Ryan-Sununu approach, wrote in a 2005 paper. ‘‘We would not be opposed to this.’’8 But others—including the Bush White House—dismissed their entire approach as reckless. In a January 2005 memo to conservative Republicans, Peter Wehner, deputy to White House political strategist Karl Rove, wrote that trying to ‘‘solve the Social Security problem with Personal Retirement Accounts alone’’—and no benefit changes—is ‘‘a bad idea.’’9 ‘‘If we duck our duty, it can have serious short-term economic consequences,’’ Wehner wrote. Borrowing trillions to finance a transition to personal accounts, while doing nothing to curtail Social Security’s unaffordable promises ‘‘could easily cause an economic chain-reaction: the markets go south, interest rates go up, and the economy stalls out.’’10 The White House memo suggested that if Washington went on a borrowing binge while failing to exercise any semblance of fiscal discipline by reining in its future promises, then that could jolt the confidence of investors and trigger an increase in interest rates. Instead of lower interest rates and the flood of new corporate tax revenues that Ryan and Sununu predicted, interest rates would be pushed higher, making it more expensive for businesses to borrow money and dampening economic growth. In other words, Ryan-Sununu’s free lunch would result in a miserable case of heartburn. Federal Reserve Chairman Alan Greenspan explained to Congress in 2005 that the case for a debt-financed privatization of Social Security hinges on the notion that it is possible to bring forward Social Security’s future ‘‘contingent’’ liabilities without putting upward pressure on interest rates because financial markets already discount
40
A WELL-TAILORED SAFETY NET
those liabilities in the price of U.S. government debt. ‘‘But we don’t know that. And if we were to go forward in a large way and we were wrong, it would be creating more difficulties than I would imagine.’’11 Why would self-described conservatives advocate plans that would make the government go on a huge borrowing spree when the nation was already at risk of taking on too much debt? One reason seemed to be cold political calculus. They preferred to talk about what they considered to be the feel-good aspect of Social Security reform—giving people the opportunity to invest their payroll taxes to seek a better return—rather than ask for shared sacrifice. During the height of the Social Security debate in 2005, when attention was shifting to the White House’s ideas for curtailing Social Security’s promises, Stephen Moore and Larry Hunter of the Free Enterprise Fund laid out their thinking in a memo to fellow conservatives: ‘‘We fear that in recent weeks President Bush and the Republicans in Congress have been suckered into a debate about shoring up the finances of Social Security, and have put on the table a series of unattractive options that voters will ultimately reject.’’12 Privatization advocates aren’t exactly wrong when they charge that Social Security provides a low rate of return on contributions for many beneficiaries, particularly higher earners, but they’re only telling part of the story. Social Security’s low rate of return reflects its social insurance function that provides higher returns to the disabled and their families, as well those who live to a very old age and collect benefits for more years than an average retiree. According to Social Security’s actuaries, roughly 3.5 percentage points of the 12.4 percent payroll tax will be needed to provide benefits for disabled workers and their families and for the children of parents who die before claiming benefits.13 The paying of benefits to spouses with inconsistent work histories is another factor that lowers the rate of return for singles and twoearner couples. Furthermore, the low rate of return for higher earners is partly a reflection of the redistributive nature of Social Security, which replaces more than 50 percent of wages for a low earner but less than 30 percent for high earners.14 All of these factors make it of questionable value to compare the rate of return on Social Security contributions to market-based investment returns. However, there is some value in comparing how Social Security returns will fare across generations, and it is true that the rate of return on payroll tax contributions is expected to become far stingier in the future, as benefits are reduced, taxes are raised, or a combination of both. Nevertheless, the idea that individuals would get a much better return on their Social Security contributions under Ryan-Sununu is highly misleading, even apart from the Bush White House’s concerns that their proposal would hurt the economy and the stock market. The argument that a system of personal accounts would generate much better returns on contributions than the current Social Security system is based, in large part, on the difference in how the two systems are financed. Unlike the current pay-as-you-go system, in which the taxes of current workers go right out the door to pay the benefits of current retirees, a system of personal account would allow an individual’s payroll tax contributions to accumulate and earn interest. But while
‘‘A
BAD IDEA’’
41
personal accounts sound attractive, in theory, the cost of transitioning from a pay-asyou-go system would wipe out any gains achieved by investing one’s payroll tax contributions in non-risky assets. As the Congressional Budget Office explained in 2004, ‘‘a pay-as-you-go retirement system such as Social Security cannot move to . . . a system of private retirement accounts—without putting an extra burden on some generation or generations.’’ CBO concluded, ‘‘In other words, to raise the rate of return for future generations . . . some generations must receive rates of return even lower than they would have gotten under the pay-as-you-go system.’’15 In general, personal account plans are treated, at least implicitly, as a loan from the government to an individual. This is logical because the diversion of payroll taxes into the accounts would, in theory, force the government to borrow an equal sum to pay benefits to current retirees. Upon retirement, account holders would pay off these loans—with interest—as their account balances are offset by reduced Social Security benefits. If the government doesn’t charge individuals an interest rate that is equal to its own borrowing cost, then it would be subsidizing the accounts and, in effect, artificially boosting the rate of return. Under the personal account plan proposed by President George W. Bush, individuals would have been charged a 3 percent interest rate (after inflation)16, reflecting the projected long-term Treasury rate, and a roughly 0.3 percent account fee.17 Earning a better return than that provided by Social Security would have required inflation-adjusted investment returns in excess of 3.3 percent, a reality often overlooked by those who talk about the ease of outperforming the roughly 1 percent real returns now provided by Social Security to higher earners. Under this scenario, an individual investing in long-term Treasuries yielding 3 percent would fare worse than under a system with no personal accounts. Under Ryan-Sununu, individuals would have a clear incentive to shun safe government bonds and take on risk because the plan included a guarantee that all individuals would get at least as much as Social Security promised, no matter how the stock market performed. But this guarantee, while designed to alleviate concerns about stock market risk, only served to add to concerns that the plan would take away resources from other government programs. The Social Security actuaries estimated that the cost of providing this guarantee would have been the equivalent of adding $2 trillion to the current public debt, along with future interest payments.18 In reality, without an ability to deliver on the budget cuts they sought, proponents of Ryan-Sununu offered nothing more than a sleight of hand. Workers now in their 40s might get a better return if we financed the baby boomers’ Social Security benefits with trillions in government borrowing. But the next generation, which would inherit a mountain of debt, would be much worse off once you factor in the tax increases they would face to keep the government functioning. The logic of Ryan-Sununu is simple: If we didn’t have to pay the benefits already earned by those who have retired or are nearing retirement, then we could afford a higher level of benefits for those now in their twenties, thirties and forties. But the question we need to ask isn’t ‘‘What kind of Social Security system would we create
42
A WELL-TAILORED SAFETY NET
if we were starting from scratch and cost were not an issue?’’ We have to deal with the reality of paying for the system that is in place, the reality that the government has been running deficits that have now reached a once-unimaginable scale, and the reality that Medicare and Medicaid also will place unsustainable burdens on the government in coming years. If they continue to offer happy talk and ignore this reality, ideologues on the right will only show that theirs is an ideology that is divorced from the reality of the challenges the country faces. The real question we need to answer is this: what is the best way to change our Social Security system to meet the real challenges we face while still providing an effective safety net? Back during Clinton years, most supporters of Social Security privatization at least paid some attention to the challenge of paying for their plans, arguing that budget surpluses could help finance the transition to personal accounts. But even after the budget surpluses vanished amid recession, war and tax cuts; even after we dug a deeper hole for future generations by amassing trillions in additional debt; and even as the hour of the baby boomers’ retirement grew ever closer, the most ardent supporters of privatizing Social Security remained so fixated on ideology that they didn’t adjust their ideas to fit the reality of the challenges the country faces. One would hope that the current economic and budget crises have put an end to the era of happy talk. While that remains to be seen, Ryan, at least, has changed his tune. In 2008, he offered a new approach to privatization that includes both additional tax revenue and major benefit cuts. Discussion of this new proposal is included in chapter six.
Chapter 5
‘‘It’s Kind of Muddled’’
Unlike President Bill Clinton, who also put Social Security reform high on his agenda but never ended up putting an actual reform plan on the table, President George W. Bush did firmly grab hold of the ‘‘third rail of American politics’’ by proposing major reductions in promised benefits. Yet while the Bush White House rejected as irresponsible the Ryan-Sununu approach of borrowing to pay all of Social Security’s promises and borrowing much more to fill the hole left by diverting payroll taxes into personal accounts, Bush’s own plan would also have required borrowing on a very large scale. Although the president proposed a very gradual change in benefits that would slowly but steadily erode income security in old age for all but the lowest earners, it would only modestly lower the cost of Social Security in the next few decades.1 Meanwhile, his plan included personal accounts that were roughly two-thirds as big as those in Ryan-Sununu. Although the White House never subjected its plan to the scrutiny of Social Security’s actuaries, it’s safe to say that Bush’s plan would add at least a couple of trillion dollars to the $4.2 trillion cash deficit the program faces between 2016 and 2036.2 At an April 2005 prime-time news conference during which Bush unveiled his proposal, he said it would ‘‘solve most of the funding challenges facing Social Security.’’ He added, ‘‘A variety of options are available to solve the rest of the problem, and I will work with Congress on any good-faith proposal that does not raise the payroll tax rate or harm our economy.’’3 The White House later clarified his remarks, saying that the president’s reforms would not erase most of Social Security’s over the coming 75 years—the traditional target for Social Security solvency; rather, it would erase most of the annual shortfall in the 75th year.4 While there’s reason to suspect that Social Security’s finances will grow gradually worse over time as life expectancy increases, how Bush’s reforms would reduce Social Security’s shortfall 75 years from now, while important, doesn’t count nearly as much
44
A WELL-TAILORED SAFETY NET
as their impact in the next several decades. After all, the budget strain from entitlement spending has long been expected to grow serious by 2030. Now, the trillions in borrowing resulting from the economic and financial crises has only heightened the need for a reform of Social Security that eases the burden of debt the government will assume in coming decades. Yet Bush’s proposal called for fairly modest sacrifice in the first decade or two, even as the government would have to go on a massive borrowing spree to pay the Social Security benefits of retirees so that payroll-tax dollars can be diverted to personal accounts. Because borrowing large sums to establish personal accounts—though not without risk—primarily has the effect of changing the timing of Social Security’s cash flows rather than leading to a permanent increase in debt, it’s reasonable to look at what his proposal would do to close Social Security’s shortfall irrespective of personal accounts. Although Bush made a point of traveling to Parkersburg, West Virginia, to drive home the point that the trust fund was a mere accounting device not backed by real resources, his proposal would solve less than half of Social Security’s full 75-year shortfall of $7.7 trillion in present value. And even excluding the $2.4 trillion in promises represented by the trust fund, Bush’s proposal would only close about two-thirds of the gap.5 As Bush’s own Treasury Department concluded in a 2008 analysis, ‘‘it is not a complete plan.’’6 Yet after the Democrats gained control of Congress in 2006 and Bush said he wanted bipartisan talks on reforming Social Security, he tried to allay fears among Republicans that he would accept a tax increase by declaring, ‘‘I’ve showed that you can solve Social Security without raising taxes.’’7 In fact, he had shown nothing of the kind. It would be much closer to the truth to say that Bush, by offering a proposal that erased less than half of Social Security’s deficit, yet nevertheless was attacked by Democrats as an assault on the middle class, had demonstrated that it was impossible to save Social Security without raising taxes. Bush might have understood that if he had asked the advice of Thomas Saving, a Republican economist and one of Social Security’s two public trustees, as well as a member of Bush’s 2001 Social Security commission. Just before the trustees met with Bush in the Oval Office in March 2005, Saving had held a news conference to announce his own reform proposal called ‘‘Social Security Reform Without Illusion.’’8 As Saving later explained, his own reform plan that included personal accounts ‘‘bit the bullet and recognized that reform will cost something and requires some taxation.’’9 Personal accounts were the lightning rod that distracted attention away from the real heavy lifting of Social Security reform—deciding how to restructure benefits and increase taxes to make the program sustainable over the long term. But the uproar over the president’s plan to reduce the level of promised benefits—which fell far short of erasing Social Security’s shortfall—helps explain why Congress, after wrestling with the problem for more than a decade, is no closer to finding an answer. Critics denounced the Bush plan as ‘‘a gut punch to the middle class’’ that would ‘‘drastically reduce Social Security benefits.’’10 And the outcry came not just because it was Bush’s plan, but because his proposal would gradually erode Social Security’s
‘‘IT ’S
KIND OF MUDDLED’’
45
primary function, which is to replace a good percentage of one’s work income in retirement. In general, wages tend to grow a bit faster than inflation each year, enabling each generation to enjoy a higher standard of living than the prior generation. And because Social Security benefits are based on wages, each generation of retirees can maintain a standard of living in retirement that reflects those wage gains. But say Social Security’s formula is changed so that benefits only keep up with inflation, as a number of policy specialists have recommended. Because wages are projected to grow about 40 percent faster than inflation11, this one change could wipe out most of Social Security’s deficit over the next 75 years—all of it, if you ignore the hole in the trust fund.12 But over time, as wage gains outpaced inflation, Social Security would replace a smaller and smaller share of a retiree’s wage income. While those in the workforce would continue to enjoy ever higher standards of living, retirees would have to make an increasingly abrupt adjustment to a much more modest lifestyle unless they had accumulated enough wealth to do with much less support from Social Security. By 2075, linking benefits to inflation instead of wage growth could result in a nearly 50 percent cut in promised benefits. Social Security now replaces about 41 percent of wage income for someone with average career earnings ($42,000 in 2009) who retires at the expected retirement age (currently 66). But if benefits are pegged to inflation, Social Security would eventually replace just 21 percent of wages.13 Bush’s plan adopted this idea of linking benefits to inflation rather than wages, but it included a progressive twist to shield moderate earners from the full impact and low earners from any benefit cuts at all. Under this plan, which was first suggested by Robert Pozen, a Democrat who served on Bush’s 2001 Social Security commission, those whose wages are in the bottom 30 percent of the income scale (roughly $23,000 and lower in 2009) would still see their benefits keep up with wage growth. But those who earn the maximum income subject to Social Security taxes ($106,800 in 2009) over the course of their careers would have their benefits pegged to inflation; and those in between would see benefits rise faster than inflation, but slower than wage growth.14 The impact of this change would only be felt very modestly by those retiring in the next decade or two, but the benefit cuts would grow steeper over time. Under Bush’s plan, as depicted in table 5.1, career-average earners would face a 28 percent cut in promised benefits by 2075. Instead of replacing 41 percent of wage income, Social Security would replace 29.5 percent. For those at the top of the earnings scale, Social Security would replace just 14 percent of wage income, down from 27 percent. Keep in mind that financial planners say that retirees can live comfortably on 70 percent to 80 percent of their pre-retirement income. But because most people save less than they ideally should, the reality is that Social Security provides at least half of the income for more than 6 of every 10 seniors and at least 90 percent of the income for 4 of every 10 seniors 80 and older.15 But over time, under the Bush plan, Social Security would no longer be able to provide the mainstay of retirement income for most seniors.
46
A WELL-TAILORED SAFETY NET
TABLE 5.1 Benefit cuts under Bush plan in 2075 Income level (2009 wages)
Social Security income replacement rate
Social Security Annual benefit
$18,900
55.3%
$10,450
55.3%
$10,450
$42,000
41%
$17,220
29.6%
$12,420
$67,200
34%
$22,850
19.8%
$13,300
$106,800
27.3%
$29,150
13.8%
$14,740
Bush plan Bush plan income replacement rate annual benefit
Note: Benefit levels reflect retirement at age 67 and don’t include any potential impact from personal accounts. Source: Adapted from Social Security Administration, Office of the Chief Actuary, ‘‘Estimated Financial Effects of a Comprehensive Social Security Reform Proposal Including Progressive Price Indexing,’’ Table B2, OASDI Benefit Before Offset, February 10, 2005.
So why did President Bush think this idea made sense? This is how he explained his approach to reform at a February 2005 Social Security forum in Tampa, Florida, according to an official White House transcript: [A]ll which is on the table begins to address the big cost drivers. For example, how benefits are calculate, for example, is on the table; whether or not benefits rise based upon wage increases or price increases. There’s a series of parts of the formula that are being considered. And when you couple that, those different cost drivers, affecting those— changing those with personal accounts, the idea is to get what has been promised more likely to be—or closer delivered to what has been promised. Does that make any sense to you? It’s kind of muddled. Look, there’s a series of things that cause the—like, for example, benefits are calculated based upon the increase of wages, as opposed to the increase of prices. Some have suggested that we calculate— the benefits will rise based upon inflation, as opposed to wage increases. There is a reform that would help solve the red if that were put into effect. In other words, how fast benefits grow, how fast the promised benefits grow, if those—if that growth is affected, it will help on the red. Okay, better? I’ll keep working on it. (Laughter.)16
The president did keep working on it, and when he officially unveiled his plan a couple of months later, he spelled it out this way: ‘‘As a matter of fairness, I propose that future generations receive benefits equal to or greater than the benefits today’s seniors get.’’17 But while his explanation was more succinct, and his aim of distributing the burden of sacrifice tied to Social Security reform in a progressive fashion a sensible one, the logic of his idea was still muddled. What is fair about putting the brunt of the benefit cuts on future generations while at the same time passing on a heavy burden of debt? It will be of small comfort for workers retiring in 2075 to know that their benefit checks, adjusted for inflation, are at least as big as those paid out to retirees at the beginning of the twenty-first century. What will matter to them is that Social
‘‘IT ’S
KIND OF MUDDLED’’
47
Security will be of much less help to them in trying to maintain the standard of living they enjoyed before retirement. What will be of great concern is that health care costs reliably grow much faster than inflation, and even faster than wage growth, and so the premiums future retirees will have to pay for Medicare could take a much bigger bite out of their benefit checks. And what could worry these retirees sick is that they will live into their late eighties, on average, and a huge percentage will live well into their nineties, putting them at growing risk of depleting their savings and leaving them to scrape by on an insufficient Social Security check. My point isn’t that Bush’s proposal was overly harsh, but rather that it is a poor fit for meeting the challenge of an aging population. It reflects a view that Social Security’s role, over time, should transition from providing a mainstay of retirement income to providing a minimal level of income support that keeps seniors out of poverty, but not a great deal more unless investment returns are favorable. But this ignores both the reality that many Americans currently save relatively little outside of Social Security and the increasing risk retirees will face of outliving their savings. Of course, it’s conceivable that workers could offset a portion of the benefit reductions if they invest some of their payroll taxes in the stock market and their investments pan out. But the stock market is no sure thing, and it’s not inconceivable that they would fare even worse. Economic growth—one of the drivers of stock prices—is expected to be considerably slower in coming decades as the baby boomers retire and the American work force expands at sluggish pace. And the government’s precarious financial position is a real threat to future stock market performance. In a 1999 column in Fortune magazine—before he served as chairman of the Council of Economic Advisers under President Bush—Harvard professor Greg Mankiw made a case that President Bill Clinton’s idea of investing part of the Social Security Trust Fund in the stock market was ‘‘a bad bet.’’18 His reasoning, ‘‘[A] problem pointed out in the fine print in most mutual fund ads: ‘Past performance cannot guarantee future results.’ ’’ ‘‘We live in the world’s richest country, at the end of the most prosperous century ever; it should come as no surprise that the market has done so well. The future may give us a similarly lucky draw, but let’s not count on it.’’19 While Bush’s personal account plan wouldn’t have begun ramping up until 2009, the recent 50 percent-plus plunge in major market indexes underscored the risk in making optimistic stock returns central to the retirement safety net. Bush deserves some credit for offering a specific idea for reducing unaffordable benefit promises, and for trying to preserve a robust safety net for those who depend on it most. His call for strengthened anti-poverty protections for very low earners also made a lot of sense.20 But ultimately his efforts were doomed to failure because his proposal failed on the two most important fronts: it could seriously undermine income security in very old age, yet it wouldn’t even come close to addressing Social Security’s financial challenges.
Chapter 6
Cold Comfort
So if President George W. Bush’s plan only chipped away at Social Security’s shortfall, yet he and most Republicans are adamant about not raising taxes, that begs a question: how badly would you have to gut benefits to erase Social Security’s deficit without raising taxes? Thanks to one intrepid Republican, we have a fairly good idea. In June 2005, with Social Security reform going nowhere as Democrats stonewalled over the president’s plan for personal accounts, Utah Sen. Robert Bennett tried to break the logjam by offering a plan to restore solvency that didn’t include such accounts. After telling Bush of his proposal, Bennett got a word of encouragement. ‘‘He just said, ‘I like your bill,’ ’’ Bennett recounted.1 If so, Bush is one of the few. Bennett’s bill has gained exactly zero co-sponsors in the Senate.2 And while it prescribes some very tough medicine, it would only erase 65 percent of Social Security’s shortfall over the next 75 years, or 96 percent of the program’s official deficit that assumes the rest of the budget will make good on the trust fund.3 Bennett included the president’s proposal to break the link between benefits and wages, except for the lowest earners. On top of Bush’s benefit reduction, Bennett included a provision that would gradually reduce benefits over time based on increases in life expectancy. ‘‘Absent any adjustment for changes in life expectancy beyond the age of retirement, longer lifetimes in retirement would mean increasingly greater dollar amounts of lifetime Social Security retirement benefits in future decades,’’4 Bennett explained. Under this provision, known as longevity indexing, anyone retiring in 2075—including very low wage earners—would face an additional 15 percent benefit reduction. 5 So not only does Bennett’s proposal leave out the strengthened poverty protections recommended by Bush and his 2001 Social Security commission, but he would slowly unwind the safety net for those most in need of Social Security’s help. Instead of a safety net, Social Security would become a lifeline.
50
A WELL-TAILORED SAFETY NET
Social Security now replaces 55 percent of wage income for a worker with low career earnings ($18,900 in 2009) who retires at the Normal Retirement Age. But by 2075, under Bennett’s plan, Social Security would replace 47 percent of wage income for a low-wage earner. As seen in table 6.1, for low earners retiring today, that would be like cutting annual benefits from $10,450 to $8,925. For average earners ($42,000 in 20009), Social Security would replace just 25 percent of wage income by 2075, down from 41 percent currently. Based on 2009 wage levels, that would be the difference between $17,220 and $10,600 in annual benefits. Bennett’s plan would go a big step further than Bush’s in taking a program that is intended to provide a large degree of income security and turning it into a modest income supplement for the vast majority of retirees. As increasing longevity puts individuals at ever-greater risk of outliving their savings, Social Security would provide cold comfort. Bennett said his plan ‘‘gradually and sensibly reduces’’ Social Security’s promises to ‘‘substantially reduce the mountain of unfunded debt’’ the program would otherwise leave to future generations.6 But his proposal is designed with a singular focus on cutting Social Security’s expenses and largely ignores the critical issue of providing income security for our seniors, millions of whom will be living well into their nineties. While Bennett tried to exercise the maximum care to minimize sacrifice for those who most need Social Security’s protection, his proposal reflects how difficult it is to achieve that goal without bringing in new revenue to Social Security. To understand why, just consider how Bennett’s proposal would impact disabled beneficiaries. In introducing his proposal in the Senate, Bennett said, ‘‘As I discussed this issue with many of my colleagues and others, it was clear that there was broad consensus that the disabled should be held harmless,’’7 yet that is far from the case under his plan. Bennett’s proposal doesn’t prescribe any cuts in Social Security’s disability insurance program, which now covers more than seven million disabled workers and their family members.8 Under current law, disability benefits are paid out of the Disability Insurance Trust Fund through the Normal Retirement Age, now 66 and soon to rise to 67, but the transition from disability to retirement benefits is a seamless one with
TABLE 6.1 Benefit cuts under Bennett plan in 2075 Income level (2009 wages)
Social Security income replacement rate
Social Security Annual benefit
Bennett plan income replacement rate
$18,900
55.3%
$10,450
47.2%
$8,925
$42,000
41%
$17,220
25.2%
$10,600
$67,200
34%
$22,850
16.9%
$11,360
$106,800
27.3%
$29,150
11.8%
$12,625
Bennett plan annual benefit
Note: Benefit levels reflect retirement at age 67; no suspension of progressive price indexing before 2072. Source: Adapted from Social Security Administration’s February 12, 2009, analysis of Sen. Bennett’s ‘‘Social Security Solvency Act of 2009’’ to reflect The 2009 OASDI Trustees Report.
COLD COMFORT
51
no change in base benefits. But under Bennett’s plan, workers disabled at 62 would receive their full disability benefit through age 67, after which they would face the same severe benefit cuts for the rest of their lives as someone who elects to retire at 67.9 Those disabled when they are 60 would face only slightly more modest benefit cuts. This helps explain why the only way we can ensure that the most vulnerable are protected—including those who have their careers cut short by disability—is to devote new resources to reinforce the safety net. Bennett said he also supports measures that would encourage workers to save more on their own. And, although it wasn’t part of his bill, he wants people to be able to invest a portion of their Social Security taxes in order to seek a better return than Social Security can provide. But many people might not be able to afford to substantially increase their savings, and others won’t have the foresight to do so. Favorable stock market returns might help to some extent. But even if personal accounts could be included in a Social Security solution without piling trillions in debt on future generations, who can say for sure that the stock market will perform consistently well? Simply hoping that stocks go up and people save more on their own doesn’t address the fundamental challenge of providing income security for an aging population. All it does is provide an excuse to ignore the challenge. It is worth noting, however, that Bennett is one of the foremost supporters in Congress of a measure that could bring in substantial revenue to Social Security, even though it isn’t part of his reform proposal. Bennett was the first Republican to sign on to a universal health care reform plan drafted by Sen. Ron Wyden, Democrat of Oregon, that would end the tax-free status of employer-provided health coverage. This tax exemption, which now costs the government about $250 billion a year— including about $75 billion in Social Security payroll taxes—is widely seen as regressive because the biggest tax breaks go to high earners with the most comprehensive and expensive health plans.10 At the same time, modest wage earners who can’t obtain coverage through the workplace aren’t in a position to benefit from such subsidies. Eliminating or simply curbing the growth of this subsidy for employerprovided care is a constructive idea that I will examine further in chapter 16. But laying this idea atop Bennett’s Social Security plan would amount to building on a faulty foundation. As I’ll detail in chapter eight, in addition to the steep benefit cuts that Bennett prescribes for future retirees, whether age 62 or 92, those seniors would also face penalties for early retirement that would reduce benefits by as much as 30 percent each year. Although Bennett’s plan for scaling back Social Security’s promises didn’t incorporate this approach of taxing employer-provided health coverage, Rep. Paul Ryan combined all of these pieces in a 2008 proposal that also features large personal investment accounts. As noted at the end of chapter four, which focused on the Ryan-Sununu privatization plan, Ryan had, by 2008, left behind his feel-no-pain approach to Social Security reform and begun advocating major benefit reductions and tax increases. His current approach adopts largely the same combination of progressive price indexing and longevity indexing proposed by Bennett to reduce Social Security’s benefits to match the level of resources.11 In contrast to Ryan-
52
A WELL-TAILORED SAFETY NET
Sununu’s guarantee of all benefit promises with no tax increases, these sharp benefit cuts acknowledge the demographic reality facing Social Security: The return workers get on their payroll tax contributions is on a downward slope. If benefits are reduced to match present tax revenue, the Social Security actuaries estimate that a single male with average earnings retiring in 2050 would see a 1.42 percent real rate of return, while one who retired in 2002 will end up getting a 2.04 percent real return.12 Ryan’s new approach continues to provide a guarantee that—no matter how the stock market performs—individuals won’t fare any worse under his personal account plan than they would under traditional Social Security; however, the new guarantee would only match Social Security’s reduced rate of return after Ryan’s benefit cuts. Nevertheless, Social Security’s actuaries figure this guarantee will have a cost of $1.5 trillion (in present value)—even if stocks perform well over the long haul.13 Perhaps the starkest change from his 2005 approach is Ryan’s proposal to end the tax-free status of employer-provided health coverage, which is part of his sweeping plan to address the country’s health care, Social Security and budget challenges.14 Additional payroll taxes generated by treating these health care benefits as wage income would be applied to Social Security under Ryan’s 2008 reform plan.15 A range of economists associated with both political parties have made the case that changing the preferential tax treatment of health care could help rein in runaway health care costs. Overall, this could be a progressive policy, yielding higher wage growth and providing a financing source for subsidies to make health care more broadly affordable. Yet, while it’s important to keep the big picture in mind, opponents might be expected to focus narrowly on what ending the payroll-tax exclusion means in the context of Social Security reform. Although the income tax exemption for employer-provided health coverage does much more to reduce the tax bill of higher earners facing a 35 percent tax rate than lower earners facing a 10 percent rate, subjecting health benefits to Social Security payroll taxes, by itself, would be regressive in nature. A $40,000-earner and a $140,000-earner who each had a $12,000 family policy would both face a $1,488 tax increase, shared by worker and employee. In their 2008 analysis of his plan, Social Security’s actuaries estimated that Ryan’s plan to apply payroll taxes to health benefits would increase Social Security’s tax revenue by $5.2 trillion (in present value) over 75 years, while increasing benefits by $1.2 trillion, for a net cash-flow gain of $4 trillion.16 Add to that more than $4 trillion in benefit cuts and Ryan’s measures would erase Social Security’s $7.7 trillion gap and then some.17 Yet despite the fact that Ryan has proposed some of the toughest medicine to address Social Security’s shortcomings that has been offered, once you factor in borrowing to privatize Social Security to a significant degree, the plan would still raise federal debt levels by $6.1 trillion in present value, only marginally better than the hole Social Security would face without any changes.18 Presumably, Ryan thinks that future retirees will reap the rewards of healthy stock market gains and an economy that benefits from greater investment. But putting a disproportionate burden of sacrifice on moderate earners doesn’t appear to offer a viable path to an ideological end in which perhaps few believe as strongly as he does.
Chapter 7
A No-Brainer
Facing pressure to offer their own Social Security fix in 2005, some Democrats pitched a plan that sounded like a no-brainer—just roll back President George W. Bush’s tax cuts for the rich. ‘‘If you just didn’t do the tax cut for the top 1 percent of Americans, you could pay Social Security through the entire century,’’ claimed Sen. John Kerry, the 2004 Democratic presidential nominee from Massachusetts.1 The first problem with Kerry’s statement was that it wasn’t true. A 2008 analysis from the Center on Budget and Policy Priorities found that the tax cuts for the highest 1 percent of all earners (those earning above $400,000) would equal 0.6 percent of GDP over the next 75 years2—which is now slightly less than the Social Security Trust Fund’s shortfall of almost 0.7 percent of GDP due to the deterioration in Social Security’s finances over the past year.3 But the bigger reason this comparison is misleading is that it treats the illusory $2.4 trillion in the trust fund as money in the bank, rather than as a debt to be settled. As seen in table 7.1, the real difference between the cost of Social Security’s tax revenue and its promised benefits over the 75-year period is 0.98 percent of GDP. So even if you raise taxes by 0.67 percent of GDP to close the trust-fund shortfall, you’d still have to raise taxes substantially higher to make good on the $2.4 trillion in government IOUs currently in the trust fund to pay all promised benefits. Nevertheless, even though Democrats never proposed it, say we had repealed much of the Bush tax cuts and brought in an extra $138 billion in tax revenue for Social Security every year, starting in 2009. Remember, Social Security won’t even need much extra cash until it starts running a permanent deficit in 2016. So what would happen? Potentially, more of the same. Washington would feel somewhat freer to spend more outside of Social Security than it would otherwise, meaning there might be little in real saving but the illusion of the trust fund would be magnified. Instead of merely $4.1 trillion, the trust fund would grow to $6.6 trillion by 20194. While on paper it might appear that Social Security has been saved for 75
54
A WELL-TAILORED SAFETY NET
Table 7.1 The cost of closing Social Security’s financing gap with tax hikes Financing gap
Annual tax hike as % of GDP
Annual tax hike 2009–2083 (based on 2009 GDP)
$5.3 trillion trust-fund gap
0.67%
$94.5 billion
$2.4 trillion debt to trust fund
0.31%
$43.5 billion
Combined $7.7 trillion gap
0.98%
$138 billion
Source: Author’s calculations based on The 2009 OASDI Trustees Report.
years, unless we are really saving—instead of just putting more resources at the disposal of politicians—then we aren’t doing nearly enough to make Social Security more affordable over the long-run. By 2025, even with $193 billion in extra tax revenue, Social Security would again be facing a small cash-flow shortfall, and as seen in figure 7.1, those shortfalls would continue as far as the eye can see. Dear ChrisA much bigger problem with Kerry’s idea is that the budget outlook will be exceptionally bleak even if all of the Bush tax cuts are reversed—even for those in the bottom half of the income scale. The Government Accountability Office projects that by 2040, even if all of the Bush tax cuts expire in 2011 and federal tax revenue returns to near-record levels as a share of the economy, Washington will be able to cover the interest on the debt, Social Security and about 80 percent of the cost of Medicare and Medicaid. National defense and everything else would have to be paid for with debt.5 Even before the fiscal outlook deteriorated with the surge in debt tied to the economic and financial crises, the liberal Center on Budget stressed that its analysis shouldn’t be misused to argue that repeal of a portion of the Bush tax cuts could avert the need for benefit cuts. ‘‘[H]ard choices will be needed regarding both the program’s revenues and its benefits to make Social Security solvent over the long term,’’ Center on Budget policy experts concluded in a 2008 analysis.6 A 2002 paper by the Center’s director Robert Greenstein, Richard Kogan and Peter Orszag, now the Obama administration’s top budget official, asserted that it ‘‘would be irresponsible’’ to use a repeal of the high-earner tax cuts to avoid any changes within Social Security.7 Indeed, it’s not even clear whether Kerry supported his own call to unwind the tax cuts for the highest earners and devote those funds to Social Security; during his 2004 presidential run, Kerry proposed spending those same funds—and more—to extend health coverage to the uninsured.8 And 2008 nomination rivals, Sen. Hillary Clinton, Sen. Barack Obama, and former vice presidential nominee John Edwards all proposed devoting the revenue raised from letting the tax cuts on high earners expire to their plans for universal health insurance. But Obama’s plan to use that tax revenue to pay for his health care agenda ran smack into a deep recession with budget deficits so massive in scale that he committed to paying for health care reform with a new revenue stream, apparently reserving the soon-to-expire tax cuts for deficit reduction.
A NO-BRAINER
55
Figure 7.1 Fixing Social Security for 75 years—but only on paper
So where else might Democrats turn for funds to close Social Security’s shortfall and avoid any benefit cuts? One idea is to apply the 12.4 percent payroll tax to income above $106,800—the current payroll-tax ceiling. But even that wouldn’t be quite enough to close the 75-year deficit and pay for the $2.4 trillion in IOUs in the trust fund.9 It would require a roughly 14.4 percent tax hike starting in 2009 to erase the full gap, even if we don’t pay a penny in extra benefits to those facing the massive tax hike. Yet as with the Kerry proposal, if the extra revenue wasn’t truly saved for the future, then Social Security would again face a small cash shortfall in 2025 and moderate deficits in each of the next 60 years and beyond. What’s more, if you combine a repeal of the Bush tax cuts and a 14.4 percentage point tax hike on work income to pay all of Social Security’s promises, then the top tax rate on work income would jump from 37.9 percent to 56.9 percent, including Medicare taxes, but excluding state taxes.
56
A WELL-TAILORED SAFETY NET
Democratic presidential candidate John Edwards said he would rule out benefit cuts and bring in more revenue for Social Security by raising taxes only on those making over $200,000 a year.10 But to bring in enough money to close Social Security’s shortfall and make good on the $2.4 trillion in the trust fund would require a tax increase of roughly 25 percentage points on earnings over $200,000.11 In combination with the phasing out of the Bush tax cuts for higher earners, such a tax increase would raise the top rate on work income to roughly 67.5 percent. Beyond the counterproductive economics of devoting the revenue from such a large tax increase to Social Security, the evolution of President Barack Obama’s positions during the presidential campaign suggest that it is politically unrealistic to save Social Security solely with tax increases on upper income groups. During the primaries, Obama initially raised the idea of applying the 12.4 percent payroll tax on all work income above the payroll-tax cap, now $106,800.12 But after Hillary Clinton blasted his idea as ‘‘a trillion-dollar tax increase on America’s hard-working families,’’13 Obama retreated. Obama later moved toward the Edwards position of applying the 12.4 percent payroll tax on those earning more than $200,000, while shielding sub-$200,000earners from tax increases. ‘‘Among the options that are available, the best one is to lift the cap on the payroll tax, potentially exempting folks in the middle—middleclass folks—but making sure that the wealthy are paying more of their fair share,’’ Obama said.14 But in the general election, Obama backed off of this hazily defined stance. Eventually, he settled on a tax increase of between 2 and 4 percent on those earning more than $250,000 to be imposed no earlier than 2017.15 While more politically palatable, Obama’s ultimate plan might erase just 10 percent of Social Security’s shortfall.16 The point is not that additional revenue shouldn’t be brought in to Social Security. It should be. But we can’t go on pretending that we are saving for the future when we are spending beyond our means. And the budget challenges facing the country are so overwhelming that if we fill the financial hole in Social Security with a very large tax increase on the wealthy, we could limit the resources available to the rest of government that provides for our defense, health care, children and the poor. Alternatively, if we keep raising taxes to continue to try and fund all of these programs, the tax burden could become so heavy that it would sap the vitality from our economy and make it harder to take care of everyone.
Chapter 8
A Crack in the Foundation
There are many important tests for Social Security reform. Does it provide the disabled real income security? Does it keep the vulnerable from falling into poverty? Is the plan financially responsible? But the ultimate test for Social Security reform is whether it provides a strong and dependable safety net in very old age for workers across the income spectrum. On that key question, the Bush and Bennett plans fail badly. But it turns out that even if we could pay all promised benefits, the safety net we have today will fall short. Forget for a second that Social Security is likely to pay out more in benefits than it collects in tax revenue in 2016 and beyond, collecting less than $8 in taxes for every $10 it has promised in benefits by 2030. The reality is that the safety net provided by Social Security isn’t only unaffordable, it’s also poorly designed for the challenges of caring for an aging population. It’s weakest where it should be strongest—in very old age—because it’s stronger than it needs to be for workers turning 62. As currently structured, Social Security rarely replaces the percentage of income it’s intended to replace because most people face steep penalties for claiming benefits several years before the Normal Retirement Age.1 And that’s why even if benefit cuts weren’t necessary, as seniors live longer and longer the safety net in old age will grow increasingly threadbare. For those who claim Social Security at 62—as more than half of all workers now do2—the monthly benefit checks they get for the rest of their lives will be 25 percent smaller than if they had waited until they turned 66—the so-called Normal Retirement Age at which few people actually retire. And when the Normal Retirement Age rises to 67 in coming years, those who claim benefits at 62 will get a Social Security check that is 30 percent smaller than those who retire five years later.3 Social Security is designed to replace about 41 percent of the income for an average-wage earner, but for one who retires at 62, it will only replace about 29 percent of wage income, just a fraction of the 70 to 80 percent of pre-retirement income
58
A WELL-TAILORED SAFETY NET
Table 8.1 The penalty for retiring early Income level (2009 wages)
Income replacement rate (retire at 67)
Annual benefit (retire at 67)
Income replacement rate (retire at 62)
Annual benefit (retire at 62)
$18,900
55.3%
$10,450
38.7%
$7,320
$42,000
41%
$17,220
28.7%
$12,050
$67,200
34%
$22,850
23.8%
$15,990
$106,800
27.3%
$29,150
19.1%
$20,410
Note: Benefit levels reflect Normal Retirement Age of 67, which won’t take effect until 2022 under current law. Source: Author’s calculations.
that financial planners recommend. As seen in table 8.1, for someone who earned $42,000 a year, those extra years of work would be the difference between receiving $17,220 from Social Security each year and getting $12,050. And that’s before any benefit cuts to erase the program’s shortfalls and even before subtracting Medicare premiums, which come out of one’s Social Security check and now cost about $1,500 a year, including premiums for the new prescription drug program. Think of Social Security as a 12-story structure that was erected in 1940 and gradually built higher. By 1961, when a three-story addition was built, the structure had reached 18 stories. By 1983, a crack had opened in the foundation of what was then a 19-floor structure, but the engineers didn’t see any danger in building still higher. But now that crack is widening and some are questioning the wisdom of adding a 22nd, 23rd and 24th floor. This, more or less, is the unsturdy state of Social Security even before looming financial pressures begin to put enormous additional stress on a weakened foundation. Design changes that politicians deemed expedient in earlier eras make little sense in the context of the stresses that will come to bear both on Social Security’s own finances and the retirement income of those who live to a very old age. Back in 1961, when President John Kennedy signed legislation moving the earliest age for collecting retirement benefits from 65 to 624, the average retiree lived to 79.55 and there were fewer than one million Americans living past the age of 85.6 By 2050, the average retiree will live to 85.47 and there will be 19 million seniors living past 85.8 Then in 1983, faced with the imminent depletion of the Social Security Trust Fund, politicians cobbled together an agreement that trimmed benefits and raised taxes in the near-term, but eventually raised the Normal Retirement Age to 67 to begin addressing the long-term funding shortfall.9 Starting in 2000, the retirement age began rising two months a year until it reached 66 in 2005. It’s scheduled to resume climbing in 2017. The combined effect of lowering the earliest eligibility age and raising the Normal Retirement Age has been to subject those retiring early to severe benefit cuts. As
A CRACK IN THE FOUNDATION
59
Figure 8.1 Building on a weakened foundation
depicted in figure 8.1, this large and growing early-retirement penalty is the crack in the foundation that is gradually undermining Social Security’s promise of incomesecurity in very old age. Relatively few retirees have felt the impact of raising the retirement age to 66— which increased the penalty for claiming benefits at 62 to 25 percent from 20 percent. This means that statistics showing that 8 percent of all beneficiaries above 65 had incomes below the poverty level in 2006 didn’t yet reflect the impact of bigger early-retirement penalties. Nor did the 12 percent poverty rate for women 80 and older yet reflect the weakening safety net.10 But poverty statistics only reflect the most glaring failure of the safety net. Even under today’s Social Security system, those across the income spectrum who retire
60
A WELL-TAILORED SAFETY NET
Table 8.2 The double-whammy of benefit cuts and early-retirement penalties in 2075 Income level (2009 wages)
Social Security replacement Rate (retire at 67)
Social Security annual benefit (retire at 67)
Bennett plan replacement rate (retire at 62)
Bennett plan annual benefit (retire at 62)
$18,900
55.3%
$10,450
33.1%
$6,250
$42,000
41%
$17,220
17.7%
$7,420
$67,200
34%
$22,850
11.8%
$7,950
$106,800
27.3%
$29,150
8.3%
$8,840
Note: Benefit levels reflect no suspension of progressive price indexing before 2072. Source: Adapted from Social Security Administration, Office of the Chief Actuary’s February 12, 2009, analysis of Sen. Bennett’s ‘‘Social Security Solvency Act of 2009’’ to reflect The 2009 OASDI Trustees Report.
early and live long enough to deplete their savings will have to scrape by on an insufficient Social Security check, and that’s why Democrats fight so hard against any benefit cuts. As seen in table 8.2, under Sen. Robert Bennett’s plan, by 2075 Social Security would replace just 17.7 percent of wage income for an average earner retiring at 62. Based on today’s wages, that amounts to just $7,420. Subtract 2009–level Medicare premiums of about $1,500 and that would leave about $5,900 a year to live on —a far cry from income security. There are three ways to approach the necessary restructuring of Social Security. A few Republican engineers think the building should be torn down altogether, even if there’s only enough money to replace it with a bare-bones structure that can’t provide protection from the elements. Still other Democratic engineers are determined to keep on building higher, putting ever-greater stress on a foundation that they regard as untouchable. But the reality is that we can’t afford to pay promised benefits in full, and most retirees who live long enough will discover—too late—that early retirement is a luxury they can’t afford to accept. The only sensible choice is to repair the crack in the foundation—by raising the earliest retirement age—while modernizing the entire structure.
Chapter 9
A Fix That Doesn’t Fix Very Much
Perhaps the closest thing there is to a liberal framework for Social Security reform is a three-part solution that was promoted as ‘‘The Ideal, Pain-free (for Almost Everyone) Way to Strengthen Social Security’’ by Nancy J. Altman in her 2005 book, The Battle for Social Security: From FDR’s Vision to Bush’s Gamble (John Wiley, 2005). The three core elements of this approach, which reflect the work of former Social Security Commissioner Robert Ball, involve shifting estate tax revenues from the general budget to Social Security, gradually raising taxes on the highest 6 percent of wage earners and investing part of the Social Security trust fund in the stock market.1 While a closer look at the fine print of the proposal reveals that it is neither ideal, nor pain-free for anyone, let’s first examine the three main elements of this proposal that purports to fix Social Security without any heavy lifting. As a starting point, it’s important to understand that this three-part solution would not only fall short of closing Social Security’s full financing gap of $7.7 trillion in present value, it wouldn’t even come close to erasing the official $5.3 trillion present value gap that treats the trust fund as money in the bank. Rather, Ball and Altman set their sights, at least initially, on a different target— close actuarial balance. This would be achieved if Social Security is projected to be able to pay 95 percent of its 75-year bill that isn’t paid for by the trust fund.2 As measured by the Social Security actuaries in 2005, these three provisions would reduce the 75-year gap by 1.47 percent of taxable payroll—meaning that the savings are equal to the sum that would be generated from a 1.47 percent tax on all of the wages exposed to Social Security’s 12.4 percent payroll tax throughout the 75-year period.3 That means the Ball plan would erase about three-fourths of Social Security’s official trust-fund shortfall of 1.9 percent of taxable payroll and about 54 percent of the full Social Security-related financing gap of 2.7 percent of payroll. Yet even this greatly overstates the budget savings that would be generated. Dedicating estate-tax revenues to Social Security is just a shifting of funds that doesn’t do
62
A WELL-TAILORED SAFETY NET
anything to help the government’s overall budget deficit and doesn’t constitute real savings. In other words, it is a fix that doesn’t fix anything, but in Ball’s plan, it would wipe out nearly one-fifth of Social Security’s shortfall.4 The actuaries estimated that another 15 percent of Social Security’s shortfall would be erased by investing part of the trust fund in equities—if returns average an inflation-adjusted 6.5 percent a year. But if stocks only perform as well as Treasuries, then there would be no budget savings from investing in equities and this three-part plan would erase less than 25 percent of Social Security’s total shortfall. Ball’s plan implicitly acknowledges that there is a limit to how much new revenue can be raised for Social Security by imposing tax increases on higher earners, but it still attempts to get around the need for broad-based sacrifice by coming up with other potential revenue streams—the estate tax and investment gains. Yet because these wouldn’t go very far toward solving Social Security’s problems, his 2005 plan went much further than close actuarial balance, prescribing both a broad-based payroll tax hike and a reduction in annual cost-of-living adjustments that would hit not just future retirees but present retirees as well.5 Ball, who died last year at the age of 93, first went to work for the Social Security Administration in 1939 and was deeply involved in the policy changes and debates through the rest of the century. He was appointed commissioner by President John Kennedy and served in that capacity through 1973. He was part of the 1983 Greenspan Commission that crafted the last big fix for Social Security. He was a member of President Bill Clinton’s advisory council that tried to prescribe another fix a decade later. And he remained on the case, offering a reform plan in 2005 that he continued to tinker with in his final years. Ball was a committed public servant and his dedication was admirable. But in every instance he advocated essentially the same idea: Let’s patch things up for a while by bringing in more revenue. ‘‘Benefits are already being cut because of Congress’ decision in 1983 to gradually increase the retirement age’’—a policy Ball argued against. ‘‘And benefits are being cut every year by the rapidly rising increases in Medicare’’ premiums, he said. ‘‘We don’t need—and can’t justify—any more cuts.’’6 Ball’s assertion was more right than wrong, but the failings of his plan show that Social Security’s challenges have become too deep to fix with patchwork. While Ball believed so strongly that the safety-net protections now provided by Social Security are the bare minimum that should be acceptable, his reform proposal would make the safety net even weaker where it needs to be strongest by cutting cost-of-living increases. And although Social Security’s inadequate poverty protections are widely acknowledged, Ball’s proposal doesn’t address them. Meanwhile, the overall tax burden imposed by his plan could be nearly as great as if politicians did absolutely nothing to constrain the program’s costs. The estate-tax provision is grounded in the misguided notion that it is either possible or desirable to improve Social Security’s financial picture by shifting resources from the rest of the budget. Ball and others have defended this proposal by arguing that Republicans are dead-set on killing the estate tax. But any realistic Social Security solution requiring an across-the-board sacrifice would drive home to both the
A FIX THAT DOESN’T FIX VERY MUCH
63
public and lawmakers that there is no free lunch, and future tax–cut proposals would be judged in that context. Ball’s support for investing the trust fund in the stock market has helped keep alive an idea that was proposed by Clinton but derailed by Federal Reserve Chairman Alan Greenspan’s warning that giving the political system a large stake in the capital markets could lead to ‘‘diminished economic efficiency and lower overall standards of living than would be achieved otherwise,’’7 But if there ever was a time for investing part of the trust fund, it was at end of the Clinton era, when the budget was running large surpluses. At that point, the government had extra funds it could have invested in the stock market, truly setting aside resources for future retirees to ease the burden on tomorrow’s workers. But that time has come and gone. Because the Social Security Trust Fund only holds IOUs that the government has written to itself, there are no real resources to invest. So now that the government is running substantial deficits, Ball’s plan would require a combination of additional borrowing and tax increases to finance government investment in stocks. And while there are legitimate public policy concerns about government investing in capital markets on a large scale, borrowing or raising taxes to finance government investment makes the idea look even more dubious. The problem is that Ball’s proposal would promise a certain level of benefits based on a prediction of favorable stock returns without knowing if those predictions will be borne out. And if the stock market doesn’t perform well, who will bear the cost? Future taxpayers, of course. Whenever the size of the trust fund declines in relation to Social Security’s annual cost, Ball’s 2005 plan would raise payroll taxes to make sure the trust fund doesn’t become depleted. If stocks perform as well as hoped, his plan would have raised the payroll tax rate by 1 percentage point in 2023. But if stocks only performed as well as Treasuries, his plan would call for a 1.8 percentage point payroll tax hike in 2021, according to the Social Security actuaries.8 This fallback plan for a bigger payroll-tax hike illustrates why depending on investment returns in order to minimize any near-term sacrifice amounts to a gamble that risks lumping even more of Social Security’s bills on the generations of workers who follow the baby boomers. The third main element of Ball’s proposal would gradually raise the income ceiling subjected to Social Security payroll taxes. In 2009, the payroll-tax ceiling is $106,800. But over the next several decades under his plan, the payroll-tax ceiling would continue to rise faster than currently scheduled, eventually reaching the rough equivalent of $195,000 in today’s terms.9 That means high-earning workers and their employers would face a combined 12.4 percent tax on the $88,200 in income between the new ceiling and the old ceiling, which amounts to a tax hike of up to $10,900 each year. Ball deploys several arguments to justify this tax increase. He notes that income gains among the highest earners have outstripped average wage gains, so Social Security taxes are expected to apply to 83 percent of national wage income, down from 90 percent in 1983.10 Over time, under Ball’s change, Social Security taxes would again apply to 90 percent of wages and salaries, and only the very-highest earners would
64
A WELL-TAILORED SAFETY NET
have income in excess of the new payroll-tax ceiling. But it’s safe to say that much of the decline in the percentage of work income subject to payroll taxes is the result of huge income gains among the very-highest earners—those earning above $195,000. So Ball’s proposal, which would disproportionately hit the broad uppermiddle class, is off target. To maintain a semblance of fairness, the higher earners facing the tax increase would receive bigger benefit checks under Ball’s plan. Eventually, those who earn $195,000 a year over the course of their careers would get roughly $42,400 a year in Social Security benefits, up from $29,150 without Ball’s tax increase.11 But when the government will already be overwhelmed by the promises it has made, it defies logic to make even bigger promises to those who will depend least upon Social Security. An analysis by the Social Security actuaries in 2008 shows that the cost of paying bigger benefits to the highest earners could reach $455 billion in present value over 75 years, or about 6 percent of Social Security’s total shortfall.12 Instead of getting more mileage out of Social Security’s resources by directing them where they are most needed, Ball’s plan would create a less-efficient safety net, while arguably skimping on the poor and the very old. The last major piece of Ball’s plan, and the only provision that would constrain the cost of Social Security, would trim the annual Cost Of Living Adjustment (COLA) that is meant to insure benefit checks keep up with inflation, erasing about 17 percent of Social Security’s shortfall. Thus, if the hoped-for investment gains don’t materialize, the other 83 percent of the shortfall would have to be covered by raising taxes and taking money from the rest of the budget that won’t have anything to spare. The idea of reducing annual cost-of-living adjustments reflects the Congressional Budget Office’s estimate that the government’s inflation measure tends to be overstated by 0.3 percentage point per year.13 But even if that is the case, it doesn’t necessarily follow that the cost-of-living adjustments for Social Security beneficiaries are too generous. Bureau of Labor Statistics (BLS) research shows that prices for the elderly increased 0.3 percentage point faster than prices for the general public from 1982 to 2007, primarily due to rapid increases in health care and housing costs.14 If these BLS numbers are an accurate gauge of inflation for the elderly, then cost-of-living increases haven’t been too generous. Ball’s plan to reduce cost-of-living increases wouldn’t just affect future retirees. It would also hit those who have already retired. But what is most troubling about his plan is that it would have the biggest impact on those who live to a very old age and are most at risk of outliving their savings. As seen in table 9.1, the longer one collects Social Security benefits, the bigger the benefit reductions would be under this provision. At age 80, a worker who retired at 62 would face a 5.3 percent benefit cut, while a worker disabled at 45 would face a 10 percent cut. The safety net would be weakest where it needs to be strongest. While these reductions are not large compared to those called for in the Sen. Robert Bennett’s plan, they would further weaken a safety net that is already fraying. As seen in table 9.2, under Ball’s plan, a worker who retired at 62 and faced a 30 percent
A FIX THAT DOESN’T FIX VERY MUCH
65
Table 9.1 Reducing cost-of-living adjustments hurts the most vulnerable Age
Benefit cut for worker who retires at age 62
Benefit cut for worker disabled at 45
70
2.4%
7.2%
80
5.3%
10.0%
90
8.1%
12.6%
Source: Benefit cuts reflect use of the ‘‘superlative’’ (or ‘‘chained’’) Consumer Price Index for COLAs, which the Social Security Administration’s Office of the Chief Actuary estimates would reduce the COLA by 0.3 percentage point, ‘‘Long Range Solvency Provisions: Provisions Affecting Cost of Living Adjustment,’’ A3, http://www.ssa.gov/OACT/solvency/provisions/cola.html.
Table 9.2 The impact of smaller COLAs and early retirement penalties Age
Social Security annual benefit for average earner (retire at 67)
Ball plan annual benefit for average earner (retire at 62)
67
$17,220
$12,050
70
$17,220
$11,770
80
$17,220
$11,420
90
$17,220
$11,080
Note: Benefit levels are based on 2009 average wage of $42,000 and reflect the 0.3 percentage point reduction in annual COLAs estimated by the Social Security Administration’s Office of the Chief Actuary. If the Consumer Price Index does overstate the inflation experienced by retirees, then both Social Security benefits and Ball plan benefits would be somewhat higher. Source: Author’s calculations.
early-retirement penalty would face additional benefit cuts that would grow larger in every year of retirement due to smaller COLAs. Scaling back support for the oldest retirees while paying the full benefit promised to those who retire at 62 could make early retirement look deceptively attractive. But as retirees spent down their savings and faced higher health care costs, too many would be left to scrape by on an inadequate benefit check. As discussed further in chapter 11, encouraging early retirement at the cost of income security in old age isn’t only the wrong prescription for retirees, it would be bad medicine for an economy that will come to rely more heavily on the productive capacity of older workers.
Part III
Principles for Reform
Chapter 10
A Bad Comb-Over
Picture a middle-aged man with a rapidly receding hairline who has a tough time facing up to reality. As he loses hair on the top of his scalp, he lets the hair grow a bit longer on the sides and combs it over to cover up his bald spots. At first, it looks almost natural. But as he loses more and more hair on top, he lets the sides grow still longer as he tries to conceal his scalp. Eventually even this proves fruitless. The bald patches become more and more obvious, no matter how long he lets his hair grow, and no matter how hard he tries to cover them up. This is essentially the approach to Social Security reform devised by two liberal– leaning economists, Massachusetts Institute of Technology professor Peter Diamond and Peter Orszag, then a scholar from the Brookings Institution and now White House budget director. Their plan would require increasingly large tax hikes, but those extra resources would still be insufficient to patch over Social Security’s shortcomings, which would grow more obvious over time as the Diamond-Orszag plan prescribes growing benefit cuts. As seen in table 10.1, benefit cuts would reach about 8 percent for an average earner retiring in about 35 years, but 15 percent for one retiring in 55 years. And these significant benefit cuts would be applied on top of penalties for early retirement of up to 30 percent. A safety net that is hardly robust to begin with would unwind a little bit more each year. What are the chances that workers would make up for such benefit reductions with greater personal saving? Under Diamond-Orszag, they would be somewhat lower, because the tax increases in the plan could make it increasingly difficult to save outside of Social Security. In addition to substantial tax increases on those earning above the current maximum subject to Social Security taxes1, the payroll-tax rate for all workers would gradually increase over time. By 2045, the payroll tax rate would be 1 percentage point higher than it is now. By 2065, the rate would be roughly 2 percentage points higher.2
70
A WELL-TAILORED SAFETY NET
Table 10.1 Benefit cuts for an average earner under Diamond-Orszag Year worker turns 62
Diamond-Orszag benefit cut
Social Security benefit (retire at 67)
Diamond-Orszag benefit (retire at 67)
Diamond-Orszag benefit (retire at 62)
2045
8%
$17,220
$15,840
$11,090
2055
12%
$17,220
$15,150
$10,610
2065
15%
$17,220
$14,640
$10,250
Note: Benefit levels based on the 2009 average wage of $42,000. Source: Author’s calculations adapted from Congressional Budget Office, ‘‘Long-Term Analysis of the Diamond-Orszag Social Security Plan,’’ December 2004, p. 26–27.
By phasing in both benefit cuts and payroll-tax increases very slowly over the next several decades, Diamond and Orszag make a choice to save the real sacrifice for later generations of workers. In other words, future workers may have a more difficult time accumulating personal savings because of a steadily rising tax rate, even as they need to save increasing amounts to overcome bigger and bigger benefit cuts. One of the important choices before policy makers—how to share the pain of Social Security reform across generations—was constructively outlined in two alternative policy options. Social Security’s actuaries calculate that the budget savings over 75 years from raising the payroll tax rate 1.8 percentage points to 14.2 percent in 2009 would be slightly greater than the savings from raising the payroll tax rate by 3.2 percentage points to 15.6 percent with half of the tax hike coming in 2021 and the other half in 2051.3 While neither of these two policy options would come close to erasing Social Security’s financial hole of $7.7 trillion in present value over the next 75 years, and the latter option would begin to have a bigger impact on Social Security’s financing gap starting in the 76th year, the actuaries’ analysis effectively demonstrates that near-term sacrifice can help lower the burden of sacrifice on future workers. Yet Diamond-Orszag opted for a tax increase that comes much more slowly than in either of the above options. They justified these steadily increasing payroll-tax rates and benefit cuts, in part, as a way to offset the additional cost Social Security will face because each new group of retirees will be expected to live longer in retirement, on average, than the prior group and will, therefore, collect benefits for a longer period of time. Presumably, they believe this approach would treat each generation fairly. But would it? The logic of the Diamond-Orszag approach implicitly rests on the notion that there is a connection between increased longevity and the ability of workers to pay higher taxes. In theory, because wages tend to grow faster than inflation, future workers will be able to afford a higher payroll-tax rate and still have wage gains left over to attain a higher living standard than today’s workers. While that might seem logical in an economics textbook, there are multiple problems with basing Social Security reform on such a presumption.
A BAD COMB-OVER
71
First, consider that because Diamond and Orszag phase in their broad-based tax hikes and benefit cuts so gradually, their plan falls quite a bit short of closing Social Security’s full financing hole of $7.7 trillion in present value. In effect, their plan relies on $1.6 trillion worth of IOUs in the trust fund to limit the burden of sacrifice for those retiring in the next couple of decades.4 Later generations are left to bear the brunt of the tax increases and benefit cuts—and the extra government debt taken on to pay the unfunded benefits awarded to baby boomers. Next, it’s necessary to consider that taxes outside of Social Security may rise over time to deal with the far more difficult financing challenges faced by Medicare and Medicaid, as well as the legacy of the government’s rapidly escalating debt levels. It’s also worth considering that income tax brackets only rise with the rate of inflation, so higher real wages over time will push workers into higher tax brackets; hence effective tax rates are already on track to rise outside of Social Security. Likewise, on the lower end of the income spectrum, the Earned Income Tax Credit is pegged to inflation, so higher real wages would, over time, yield smaller subsidies. Finally, over time, workers facing a higher tax rate would have to set aside a greater share of their career income under Diamond-Orszag to overcome reductions in Social Security’s income-replacement rate that would gradually phase in over a number of decades. On top of that, workers who can expect to live longer lives, on average, than today’s seniors would arguably need to save a greater portion of their income to address the growing risk that they will outlive their savings. All of these considerations suggest that a proposal that is generationally fair wouldn’t force later generations to pay a significantly higher payroll tax rate than we are willing to levy on ourselves, after some delay during this time of economic strife and perhaps a short transition period. Leaving aside the question of fairness, there is the question of whether the Diamond-Orszag plan would produce an effective safety net. Yet, that is not at all certain, despite all of the extra revenue they would devote to Social Security—and perhaps partly because of it. As seen in table 10.1, the benefit cuts for an average earner are projected to reach 15 percent for workers retiring six decades from now. In combination with a 30 percent early-retirement penalty for workers retiring at 62, Diamond-Orszag could result in a 40.5 percent cut in annual retirement income that would apply even when these retirees turn 90 and beyond. Based on 2009 wage levels, that would be the difference between annual benefits of $17,220 and $10,250. While workers might decide on their own to delay claiming Social Security benefits until after age 62, Diamond-Orszag continues to hold out the promise of early retirement and, to some extent, might make it harder to resist. That’s because the tax increases in their plan might only reinforce the desire of workers to begin getting back what they paid in as soon as possible. While there is no definitive correlation between increased longevity and the ability to retire at a later age, the improved health outcomes that lead to longer lives certainly suggest some linkage. But because Diamond-Orszag rejects any increase in the age at which benefits become available, their plan relies solely on tax hikes to try to ensure that
72
A WELL-TAILORED SAFETY NET
workers have adequate retirement income. Yet despite significant tax hikes on average workers and bigger tax hikes on the highest earners5, their plan reveals the futility of this approach. It is essentially the plan of a government that is too set in its ways to accept reality, like a man with a bad comb-over who just won’t give up his futile attempts to conceal the fact that he’s balding, no matter how unfortunate the results.
Chapter 11
The Problem is Part of the Solution
In 1950, a decade after Social Security’s introduction, the earliest retirement age was 65 and average workers didn’t claim benefits until they were 68.1 At that point, they were only expected to live for another decade.2 Now, 50 percent of workers claim Social Security benefits at 62 and they can be expected to live, on average, to 83. By 2085, the average life expectancy for 65-year-olds will rise to 87, meaning that those who retire early could claim benefits for a quarter century—even if they don’t outlive their peers. The task force appointed by President Franklin D. Roosevelt, in its 1935 report, said the purpose of Social Security would be to support seniors ‘‘beyond the productive period.’’ 3 But, as seen in table 11.1, it has gradually ‘‘morphed into a middle-age retirement system’’ in which the average person already spends close to one-third of their adult lives, Eugene Steuerle, then a senior fellow at the Urban Institute, told Congress in 2005.4 Today’s retirees not only receive Social Security for a longer period, but they also receive bigger monthly benefit checks that keep pace with the wage growth from one generation to the next. Up until now, Social Security has been able to afford to pay bigger checks for a longer period to each new group of retirees because of a combination of payroll-tax increases and favorable demographics. But, as seen in table 11.2, because those retirement checks are financed by the payroll taxes of current workers, and the 77 million baby boomers about to retire chose not to have as many children as their parents, the good luck is about to run out. By 2020, it would take the equivalent of a payroll-tax increase of 1.5 percentage points to pay all promised benefits. By 2030, it would take a payroll-tax hike of 3.6 percentage points. And by 2083, it would take a 4.2-percentage-point hike.5 Though I have argued the Ball and Diamond-Orszag plans are less-than optimal, they are very constructive contributions to the debate because they make this clear: Even with major tax increases and large infusions from the rest of the budget, Social Security’s promises will still have to be cut to make ends meet.
74
A WELL-TAILORED SAFETY NET
Table 11.1 Workers are retiring younger, living longer Year
Average age workers claim Social Security retirement benefits
Average life expectancy at age 65
1950
68.5
79
1960
66.3
79.5
1970
64.1
80.3
1980
63.6
81.3
1990
63.5
82.2
2000
63.8
82.5
2007
63.4
83.0
Source: Social Security Administration, ‘‘Annual Statistical Supplement, 2008,’’ Table 6.B5: Number, average age, and percentage distribution, by sex and age, selected years 1940–2007, http://www.ssa.gov/ policy/docs/statcomps/supplement/2008/6b.html#table6.b5; Office of the Chief Actuary, ‘‘Unisex Life Expectancies at Birth and Age 65,’’ June 2008, http://www.ssa.gov/OACT/NOTES/ran2/index.html.
Diamond and Orszag justify the payroll-tax increases and benefit reductions they prescribe, in part, as a way to offset increases in longevity. In other words, because future retirees will collect benefits for more years, they’ll have to pay extra and still get smaller checks than they’ve been promised. But the tax increases would grow steeper and benefit cuts would grow deeper with each new group of retirees, making it more difficult for workers to save outside of Social Security, even as the safety net gradually unwinds.
Table 11.2 Social Security’s traditional structure is no longer a good fit Year
Cost of scheduled benefits as a percentage of taxable payroll
Workers per Social Security beneficiary
1970
8.1%
3.7
1980
10.7%
3.2
1990
10.7%
3.4
2000
10.4%
3.4
2010
10.8%
3.2
2020
14.5%
2.6
2030
16.8%
2.2
2040
17.0%
2.1
Source: Social Security Administration, The 2009 OASDI Trustees Report, Supplemental Single-Year Tables, Table IV.B1: Annual Income Rates, Cost Rates, and Balances, Calendar Years, 1970–2085, http:// www.ssa.gov/OACT/TR/2009/lr4b1.html; Table IV.B2: Covered Workers and Beneficiaries, Calendar Years, 1945–2085, http://www.ssa.gov/OACT/TR/2009/IV_LRest.html#34542age point3.
THE PROBLEM IS PART OF THE SOLUTION
75
Sen. Robert Bennett’s plan, which proposes no tax increases, adjusts for longevity with benefit reductions alone. The basic idea is that as expected longevity increases, retirees will get a smaller annual benefit spread out over a greater number of years. If life expectancy in retirement rises 15 percent, retirees will get a Social Security check that is 15 percent smaller. The problem with this kind of approach is that as longevity increases and retirees live to very old ages, they will be more likely to deplete their savings and will depend more on Social Security, not less. An even bigger problem with these kinds of longevity provisions is that they would be placed on top of a structure that is already becoming unsound. Significant benefit cuts on top of an up to 30 percent penalty for early retirement would turn Social Security’s promise of income security in old age into an empty promise. Policy makers have a simple choice: either maintain Social Security’s outdated structure that isn’t built to withstand the coming demographic and fiscal pressures, or adapt it to accommodate the gains in life expectancy among our senior population and, in doing so, get more mileage out of Social Security’s limited resources. While additional resources are necessary to preserve an effective safety net, raising taxes and transferring money from the rest of the budget is mostly a zero-sum game. Locking in new revenue streams to finance Social Security can only be done by denying those resources for health care, education and all other government functions. As former Federal Reserve vice chair and Clinton White House budget director Alice Rivlin put it, proposals to expand financing for Social Security must weigh ‘‘the future consumption of the elderly against that of other groups (the young, the middle aged, the sick, the poor) or against other activities (national defense, scientific research).’’6 The government’s resources may be stretched beyond the breaking point by the more daunting financial challenges presented by Medicare and Medicaid. And taxes —and debt levels—can only go so high without hurting the economy and shrinking the pie that must be divided between the elderly and everyone else. Therefore, it makes sense to limit the need for tax increases by expending Social Security’s resources where they are most needed. And the truth is that those resources aren’t being used very efficiently. Consider that the average worker didn’t retire until past 68 in 1950, yet today, when the life expectancy of a retiree is four years longer, 36 percent of retirement benefits are paid to people 69 and younger7—retirees who perhaps shouldn’t be considered old. Back in the mid-1970s, 29 percent of people aged 62 were in fair or poor health, according to the long-running National Health Interview Survey. By the mid-1990s, a similar percentage didn’t report fair or poor health until they were past 70.8 A study examining the status of men applying early for Social Security benefits at the beginning of the 1990s found that only 10 percent were both in poor health and lacked pension income outside of Social Security.9 These kinds of findings support the case for scaling back early eligibility for Social Security to reflect increases in life expectancy, though they also suggest a need to do so with great care. The answer to Social Security reform isn’t to devote more resources than we can afford to salvage a safety net that is inefficiently designed and has a growing hole in
76
A WELL-TAILORED SAFETY NET
it. Social Security reform should be about redesigning the safety net so that the support is strongest where it needs to be the strongest, and fewer resources are expended where less support is required. The less people depend on Social Security in their early and mid-60s, when most are able to keep working, at least part-time, the more resources the program will have to provide a very solid safety net for seniors late in life—which is the ultimate test of Social Security reform. What’s more, the time is right for such a redesign: The near-term catalyst for Social Security’s funding problem—the retirement of the baby boomers—can also be part of the solution. Within a decade, the growth of the U.S. labor force is projected to slow to less than half of its average pace in recent decades as more and more baby boomers exit the work force. As seen in figure 11.1, the economy’s growth potential is expected to decline because there simply won’t be enough new workers to continue the average pace of economic growth we have experienced in recent decades. In 2001, the Government Accountability Office warned that the approaching retirement of the baby boomers could create ‘‘possibly acute occupational labor shortages’’ that could impede economic growth, and it called for policies ‘‘to extend the work life of older Americans.’’10 For now, the severe recession and jump in unemployment have disrupted the projected employment trends. In the near-term, higher joblessness may lead more workers to claim benefits earlier than they had anticipated; on the other hand, the hit to retirement savings from falling stock prices may lead many to extend their careers. But the longer-term picture still holds. The downside is that there will be fewer workers to shoulder the burden of benefits for retirees. The upside is that this also means that our workforce will be stretched thin and older workers will be in high demand. In this context, it makes little sense to spend precious resources trying to hold out an empty promise of early retirement for all and encouraging workers to retire at an early age. Not only would individuals who leave the workforce too early come to regret that decision if they live long enough to deplete their savings or if they face steeper health care costs than they counted on, but depriving the economy of the productive capacity of older workers would make it more difficult for the nation to meet the challenge of taking care of our aging population. The point isn’t to keep older workers on the job because the economy can’t do without them. The real payoff is that workers who extend their careers will be able to afford a higher standard of living in retirement. Early retirement would mean big penalties and smaller checks from Social Security even if there wasn’t a serious funding shortfall. In a society in which most people save far too little, additional work would go a long way toward providing an acceptable level of income security beyond the age of 80, when the average retiree depends on Social Security for more than 70 percent of income.11 This will become increasingly important as life spans continue to lengthen. We can’t reasonably expect workers with moderate earnings who retire at 62 and have a good chance of living into their 90s to save enough over a 40-year career to live comfortably for 30 years in retirement, particularly if they are faced with higher payroll taxes, significant benefit cuts and severe early-retirement
THE PROBLEM IS PART OF THE SOLUTION
77
Figure 11.1 Slower growth is the new normal
penalties. As Alicia Munnell and Steven Sass of Boston College’s Center for Retirement Research wrote: ‘‘[T]he outlook for retirement income for future cohorts of retirees is dismal. People are not going to be able to continue to retire at 62 and 63 and maintain their pre-retirement living standards over an increasingly long period of retirement. Working longer is an obvious solution.’’12 There’s also another important payoff from longer careers: If workers stay on the job a few years longer, Social Security will bring in more revenue, which will enable it to pay out a higher level of benefits than it could otherwise afford. An Urban Institute analysis found that if all workers extend their careers by an average of five years, Social Security would bring in enough payroll taxes to reduce the projected shortfall in 2045 by $70 billion, or 20 percent of the projected $348 billion deficit, reducing the need for tax hikes and benefit cuts to achieve solvency.13 And even this greatly understates just how much additional work would increase the nation’s resources to
78
A WELL-TAILORED SAFETY NET
meet the challenge of an aging population. Including additional federal and state income taxes, the Urban Institute found that a five-year increase in work would bring in more than $375 billion. Yet it should be noted that this particular study focused on the financial impact of longer careers that are, at least implicitly, related to voluntary decisions, as opposed to an increase in the retirement age. The Social Security actuaries have found that a mandatory increase in the earliest eligibility age for benefits might actually be a slight negative for Social Security’s finances because it would lead more workers to claim disability benefits. Clearly, there are very serious issues to weigh in redesigning the safety net. It is likely that a scarcity of workers will be most pronounced among high-skilled and college-educated workers, so it’s necessary to take into account that workers with only modest skills and minimal savings will be the ones most affected by any change in Social Security’s retirement age. It’s also important to consider that workers with strenuous jobs might not be able to work as many hours or perform all of the functions they did earlier in their careers. What’s more, recent research confirms that life expectancy is increasing much more slowly for lower-income groups.14 These concerns can’t be ignored, but they don’t change the fact that Social Security’s present structure is ill-suited to meet either the economic challenges we face as a nation or the needs of future retirees, who will be living longer and who will depend even more on a robust safety net in old age. A one-dimensional reform of Social Security that relies only on cutting benefits and raising payroll taxes, and doesn’t adjust the earliest retirement age or at least tilt the incentives in favor of delayed retirement, will waste precious resources and do an increasingly poor job of protecting retirees in very old age. And even major tax increases can’t keep the safety net, in its current form, from gradually unwinding. In effect, proposals such as DiamondOrszag that seek to preserve Social Security’s outdated structure are merely buying time and delaying the inevitable, but at substantial cost to taxpayers, the economy and—most importantly—to retirees. Only by scaling back Social Security’s promise of an early retirement, can we avoid the need for benefit cuts in very old age, when people can least afford them.
Chapter 12
‘‘The Obvious Thing To Do’’
A worker who retires this year at the Normal Retirement Age after having earned the average wage over a 40-year career will receive a Social Security benefit of about $17,220 a year, adjusted for future inflation. Because the average person who retires at 66 will live to 83 and collect benefits for 17 years, someone with career-average earnings would be expected to collect close to $340,000 in benefits that will be financed by taxing the wages of those in workforce. Now imagine for a second that workers retiring today, instead of paying the benefits of old retirees, had been able to deposit their payroll taxes into a personal account composed of ultra-safe investments that offered a rate of return equal to that Social Security provides on one’s payroll tax contributions. That same 66-year-old would by now have accumulated about $230,000 in savings, which along with future interest payments would be able to pay the same $17,220 a year through age 83. Instead of relying on payroll taxes from younger workers to meet the needs of older consumers, retirees could rely on their own savings, allowing today’s workers to save and invest their own payroll-tax contributions. Those savings would expand the pool of new capital available to invest in the private sector by hundreds of billions of dollars. And because of this dramatic increase in savings available for investment, businesses would be able to borrow at lower interest rates and have more resources to expand their operations and devote to developing the technologies of the future. Now I’m not actually suggesting we transform Social Security in this dramatic fashion. For one thing, Social Security provides critical protections that would be difficult to replicate in an individual savings program—guaranteed, inflation-protected benefits for life, survivor benefits and insurance against disability, for example. And it’s probably not feasible to shift primarily to a savings-based system without borrowing trillions to keep paying the benefits of retirees and those nearing retirement, because doing so would require extremely large tax increases.1 The point I’m making is that there is an economic cost to simply taking the potential savings of younger
80
A WELL-TAILORED SAFETY NET
workers and giving them to older retirees. Conversely, there is a potential economic benefit to be gained by introducing some savings component to Social Security—as long as it does not come at the cost of increased debt—and that presents an important opportunity that must be seized. We have every reason to expect that we will have an economy that will struggle to provide for the income and health care needs of the elderly, while still leaving enough resources to provide for children, the poor, national security, and everything else. Therefore, in reforming Social Security, we need to make sure we do it in a way that expands the economic pie to the extent possible. ‘‘[F]ocusing too narrowly on the Social Security funding question—in isolation from the more fundamental economic challenge of an aging population—risks muddling the problem and perhaps picking a wrong answer,’’ then-Federal Reserve Vice Chair Alice Rivlin warned in a 1999 speech.2 Trying to figure out today what taxes to raise to pay for Social Security’s promises in the future, she explained, was a less urgent matter than doing everything we can to make sure that the nation will have the resources to afford the promises it makes. ‘‘If we can find ways to make the future workforce more productive, both they and future retirees will benefit . . . If we want to have a bigger GDP to divide among the elderly and the non-elderly in the future, the obvious thing to do is to save more now.’’3 Therefore, she said, the merits of any proposal to fund Social Security’s future obligations ‘‘depend on how much it contributes, on balance, to national saving— the more it does, the more it helps to ease the burden of paying all future claimants, elderly and non-elderly alike.’’4 Looked at in this light, the Diamond-Orszag approach of gradually raising taxes on workers—not to save for the future, but to write bigger checks to the elderly than we could otherwise afford—is the wrong answer for the economic challenges we face. The rising tax increases would only shrink the pool of savings available for investment in the economy and take us further in the direction of a nation that consumes but does not save. Social Security, as currently structured, ‘‘does not contribute to national savings— indeed, it may reduce it to the extent that it discourages private saving,’’ wrote Martin Baily, who chaired the Council of Economic Advisers under President Bill Clinton, and Jacob Kirkegaard inU.S. Pension Reform: Lessons from Other Countries.5 Rather than a solution to our need for savings, funding Social Security with a steadily increasing tax rate would be a consequence of a lack of savings. Still, it must be said that Diamond-Orszag is among the most fiscally responsible of the reform plans that have been offered and would have a much more positive impact on national savings than a number of the other recent proposals that eschew tax hikes and include only very gradual benefit cuts or none at all. When Rivlin made her speech, the government was projected to run large budget surpluses for the foreseeable future. At the time, it appeared that the Clinton administration’s plan to use much of those surpluses to pay down the federal debt would give a big boost to national savings and free up capital for investment in the private economy. But because those surpluses vanished and were replaced by large budget
‘‘THE
OBVIOUS THING TO DO’’
81
deficits, the challenge of raising national savings has become both more difficult and more urgent. Unfortunately, there is no longer any ready pile of money that we can use to pay off the government’s debt or set aside for future retirees. So how do we start saving for the future while putting Social Security on a sustainable financial path? The reality is that every reform plan that tries to preserve a viable safety net without massive borrowing would require tax increases. We can’t avoid tax hikes, but at least we can make sure that any broad-based payroll-tax increases are truly saved for future retirees—unlike the surplus revenues that have flowed to Social Security over the past quarter-century. And there are only two ways to save: either the government can invest the funds in stocks and other marketable assets, or we can deposit these funds into personal accounts. The question over who should do the investing—individuals or the government— has been debated since at least the mid-1990s, when the Clinton Social Security advisory council split over the issue, with seven members favoring individual accounts and six preferring government investment.6 One might have guessed that this question would have been settled by the March 1999 vote in which the Senate rejected government investment of Social Security assets by a 99–0 margin.7 Yet because the idea of individual accounts within Social Security—and particularly the idea of financing accounts using current resources devoted to Social Security—has been such a lightning rod, this debate has lived on. It is well past time to put it to rest. Opponents of government investment on a large scale come from across the political spectrum. Some argue that politicians would use the government’s power as a shareholder to meddle in corporate decisions. Others worry that the government would bend over backward to help the stock market. But if there ever was a good case for the government to invest Social Security tax dollars, it has only grown weaker over time as the nation’s financial position has deteriorated. Now that the government is running massive deficits, we have no surplus funds to invest. That means the only way the government can buy stocks or other investment assets is by raising taxes or borrowing money. And now that our budget challenges have grown so overwhelming, we can’t afford to make promises based on a hope of favorable market returns. All of these arguments against government investment of Social Security funds are reason enough to discard the idea. But perhaps the best reason to reject it is because of what would be lost by not allowing individuals to accumulate real savings within Social Security. Although it is unfortunate that we have to ask for sacrifice even from modest-wage earners, there is no realistic way to avoid it if we want to preserve a robust safety net. That’s why even Ball’s 2005 plan and Diamond-Orszag, which impose significant tax hikes on high earners as well as benefit cuts, also call for an increase in the payroll-tax rate. But if we have to ask those who already have little ability to accumulate savings to contribute even more of their income to keep the safety net robust, then we should at least make sure that the sacrifice is modest and that the taxes are really extra savings that go into a personal account, rather than into a government account.
82
A WELL-TAILORED SAFETY NET
Consider that among households ranking among the bottom 20 percent in annual income, the median family owned just $1,700 in financial assets in 2007, according to the Federal Reserve’s most recent Survey of Consumer Finances.8 Among the fifth of households whose incomes place them above the lowest-earning 20 percent, the median family owned just $7,000 in financial assets. And among all non-white households, the median family owned just $9,000 in financial assets. Social development experts believe that accumulating assets leads to important psychological and social effects that can’t be achieved when the government simply provides individuals with an income stream to sustain them. Among the benefits are more long-range planning, improved financial literacy and increased social and political participation.9 Although Democrats express support for personal accounts outside of Social Security, raising the combined 15.3 percent payroll-tax rate for Social Security and Medicare without adding a personal savings component to Social Security would make it even harder for workers already struggling to make ends meet to find additional dollars to save. What’s more, unless the government gets its fiscal house in order, it won’t be able to afford the savings incentives that a number of Democratic lawmakers have proposed.10 The function of individual accounts within Social Security would be to accumulate savings that could offset a significant part of the reduction in promised benefits that is necessary to bring Social Security’s obligations in line with its resources. And because future retirees, in effect, would pay a portion of their own benefits, the burden of paying for Social Security would be somewhat more manageable for future workers, who would be able to truly save a portion of their own payroll taxes. Beyond the potential economic benefit from any increase in national savings, having individuals finance some of their own benefits also could have a positive impact on retirement income. As Baily and Kirkegaard explain, workers are more likely to view contributions to individual accounts as saving rather than taxes. ‘‘And when they see that by working more and contributing more they can add to the retirement income they will receive later, this perception reduces the work disincentive intrinsic to many public pension plans and alleviates the labor market challenge.’’11 Of course, the impact on retirement income may be greater if Social Security reform doesn’t simply alter appearances but tangibly improves the incentives for delayed retirement and reduces the disincentives for continued work. Toward that end, as I will explain further in chapter 18, the degree of ownership provided by personal accounts can be designed to reduce the risk that individuals might forfeit some of their contributions if they delay retirement. On the other hand, a failure to provide such ownership while raising the payrolltax rate could encourage workers to claim benefits as early as possible to begin to get back what they have paid in, even if they have to accept significant penalties for retiring early and sacrifice a degree of income security in old age. This is the exact opposite of what we need to do, because our best chance of saving Social Security without abdicating other critical responsibilities of government is encouraging workers to delay retirement and increasing the resources available for investment in the
‘‘THE
OBVIOUS THING TO DO’’
83
economy’s productive capacity so that the nation will be able to afford the promises it makes. Democrats—and a majority of Americans, according to polls12—sensibly rejected President George W. Bush’s personal accounts because they came with potentially high costs—a risk that poor market returns will unwind the safety net and a need for trillions in borrowing to finance current benefits as workers begin to save their own contributions. But it would be a mistake to conclude that the public is opposed to adding a personal savings component to Social Security. If lawmakers put on the table an increase in the payroll tax—and both Diamond-Orszag and the 2005 Ball plan demonstrate that a broad-based tax increase is needed—then the public is likely to reexamine the relative benefits of individual saving versus government saving. To rebut the Democrats’ argument that risk is incompatible with Social Security’s role as the bedrock of income security in old age, account supporters make the case that benefits are already at risk because the government has made promises it can’t afford. But Social Security reform should seek to limit risk to the retirement safety net, rather than simply trading one risk for another. The concern about investment risk is a legitimate one, and there are other serious concerns that also must be effectively addressed. Among the biggest is that individuals will draw down their personal account balances and be left with an inadequate retirement income. I will address these issues in chapter 18 and lay out a new approach to personal accounts that includes only their most constructive and progressive features while avoiding their potential perils.
Chapter 13
A Bridge Too Far
Because we want a robust Social Security safety net that we can afford without neglecting other critical national priorities, then two elements need to be part of any reform proposal: We need to set aside real savings in the near-term, increasing investment in the productive capacity of the economy and providing real resources to help pay future retirement benefits and ease the burden younger workers will face to take care of an aging population. And we need to encourage older workers to stay in the workforce so that they won’t face steep penalties for retiring early at the cost of income security in old age and the economy will be better able to meet the challenge of providing a robust safety net. At present, there is really only one plan on the table that incorporates both of these essential features—real saving and delayed retirement. The Nonpartisan Social Security Reform Plan was developed by Harvard University professor and current deputy White House budget director Jeffrey Liebman, New America Foundation’s fiscal policy program director Maya MacGuineas and Dartmouth University professor Andrew Samwick.1 Having respectively served as advisors to President Bill Clinton, Sen. John McCain and President George W. Bush, the three policy wonks from across the political spectrum got together in 2005 in an effort to turn the Social Security debate in a more constructive direction. Together, they hashed out a reform plan that they believed could serve as the basis for a bipartisan compromise. Being a true compromise, they all had to accept provisions that they saw as less than ideal, but all saw their proposal as superior to the prospect of a protracted stalemate—and they were surely right about that. Under the Liebman-MacGuineas-Samwick (LMS) compromise, in return for Republicans agreeing to bring a great deal more revenue into Social Security, Democrats would agree to devote all that new revenue and more to finance a system of personal accounts that would supplement a significantly reduced guaranteed benefit. In addition, their plan would raise the expected retirement age an extra year to 68 and
86
A WELL-TAILORED SAFETY NET
gradually lift the earliest eligibility age from 62 to 65 in order to shield retirees from severe early-retirement penalties on top of the benefit reductions they propose. Their proposal calls for an additional 1.5 percent tax on all wages subject to Social Security taxes, and—like the Ball plan—it calls for raising the Social Security tax base to cover 90 percent of all wages. In 2009, only the first $106,800 in wages was subject to Social Security taxes, but the LMS proposal would raise that payroll-tax ceiling over a decade to today’s equivalent of about $195,000.2 Those at the top of the new scale would face the full 12.4 percent tax on an extra $88,200 in income, raising their annual tax bill by $10,940, though it would be split between worker and employer. Unlike the Ball plan, those facing a tax on income over the current Social Security tax ceiling wouldn’t get additional benefits to compensate them.3 To understand the LMS compromise, it’s important to consider that if the tax revenue raised by their plan were directed straight to the government and paid out to beneficiaries, then it would be enough to close almost the entire 75-year shortfall without any benefit cuts, including the portion of the deficit represented by the $2.4 trillion trust fund.4 Add in the proposal’s moderate-scale benefit cuts, and it’s evident that it prescribes somewhat more sacrifice than is absolutely required to erase the 75-year deficit. There are two primary reasons why the authors went this route. First, their goal is to store up real savings that would pay a portion of future benefits and help ease the burden on coming generations to provide a safety net for an aging population. Because Social Security’s shortfalls are projected to grow over time as longevity increases and the ratio of workers to retirees decreases, the lack of such saving, or pre-funding of future benefits, would leave a relatively bigger hole in Social Security that would have to filled with bigger benefit cuts or tax increases down the road. Second, this increase in savings in the near term would increase the capital available for investment in the economy’s productive capacity, theoretically leading to stronger economic growth and higher wages—the best recipe for meeting the nation’s daunting budget challenges on the horizon. Because the history of Social Security’s trust fund over the past quarter-century has provided abundant reason to doubt the government’s ability to store up real savings, two of the LMS plan’s authors, Samwick and MacGuineas, insisted that all of the saving be done in personal investment accounts. Under the LMS plan, every worker would have 3 percent of annual wages deposited into a personal account each year. Half of these funds would come from the 1.5 percentage point increase in the payroll tax rate (to 13.9 percent). For some perspective, for average earners who contribute 1.5 percent of income into a personal account over the course of a full career, these accounts could provide roughly 15 percent of their Social Security income, assuming investment returns equal to that provide by long-term government bonds. Instead of limiting account deposits to 1.5 percent of wages, the LMS plan matches the add-on contribution with another 1.5 percent of wages that is financed, in part, by diverting some of the revenues that now flow to Social Security into the accounts. In the shorthand of Social Security reform, the account structure combines a 1.5 percent add-on deposit with a 1.5 percent carve-out deposit—meaning that it is carved out of Social Security’s revenue stream. And it is this carve-out portion of the
A BRIDGE TOO FAR
87
accounts that has generated the fierce and uncompromising opposition of today’s Democratic Party. Although the political rhetoric grows more heated when the debate moves from personal accounts financed by additional payroll taxes to personal accounts financed by carving revenues out of Social Security, the issues are largely the same. But as more resources are devoted to personal accounts instead of the traditional Social Security system, both the potential opportunities (assuming the accounts aren’t financed primarily with borrowing) and risks of accounts are magnified. Let’s first look at the opportunities offered by the LMS plan. One simple measure of the benefit of setting aside savings in advance in personal accounts can be seen in a comparison with the Diamond-Orszag plan. Although both of these plans prescribe benefits cuts that appear roughly similar in scale (assuming the LMS accounts only perform as well as government bonds) and both impose comparable tax increases on the highest earners, the Diamond-Orszag payroll tax increase on all workers would reach 2 percent in 2065 versus 1.5 percent in the LMS plan. A side-benefit of building up saving within personal accounts is that workers may focus less on the marginal work disincentive created by their higher tax burden and more on the connection between additional work and increases in retirement income. ‘‘Workers will be less likely to perceive their required contributions as a pure tax if they see those contributions being directed to an account that they own,’’ the plan’s authors reasonably argue.5 A constructive perception of a Social Security program with a higher tax burden would be encouraged by the broader opportunity workers who die before retirement would have to leave an inheritance under the LMS plan. While that opportunity is generally limited under the current Social Security system to workers who have lower-earning spouses or children under 18, the LMS account plan would allow all workers to pass on their account accumulations. With annual deposits equal to 3 percent of annual wages, someone earning just 50 percent of the average wage (about $21,000 in 2009) over the next 40 years would accumulate about $49,000 (in 2009 dollars) in their personal account if returns merely equal the return on safe government bonds. That’s a little more than three years’ worth of Social Security benefits. An average wage earner (about $42,000 in 2009) would accumulate $98,000. These are not incredible sums, but they provide moderate earners with ownership of financial assets that they can pass on if their lives are cut short. So what is it that Democrats find so objectionable about personal accounts? To really understand, it helps to take a look under the hood of the LMS plan. Remember, half of each account is financed by a 1.5 percent payroll tax increase, and the other half—the carve-out portion—comes from revenue that would otherwise flow to Social Security. Because it is the carve-out approach that Bush proposed and that Democrats directed all of their fire at, it’s important to understand just how different the LMS carve-out is from the Bush approach. As seen in table 13.1, two-thirds of the carve-out in the LMS plan would be paid for with a new tax on earnings over the payroll-tax ceiling. In other words, the funds raised by taxing higher earners are redistributed into the accounts of all workers.
88
A WELL-TAILORED SAFETY NET
Table 13.1 LMS personal account structure Source of funds
% of taxable payroll
% of funding for personal accounts
New revenue for Social Security
Payroll tax hike
1.5%
50%
$83 billion
High-earner tax
1%
33.3%
$55 billion
Social Security
0.5%
16.7%
Total
3%
100%
$138 billion
Source: Adapted from Social Security Administration, Office of the Chief Actuary, ‘‘Estimated Financial Effects of ‘A Nonpartisan Approach to Reforming Social Security—A Proposal Developed by Jeffrey Liebman, Maya MacGuineas and Andrew Samwick,’ ’’ November 17, 2005, http://www.ssa.gov/ OACT/solvency/Liebman_20051117.pdf.
That means just 0.5 percent of wages, or one-sixth of the total account, is carved out of Social Security’s existing revenue stream. Looking at the bigger picture, the LMS plan increases the taxes flowing into Social Security—albeit into personal accounts within Social Security—by roughly 19 percent, and it carves out less than 4 percent of Social Security’s existing revenue base to help fund the accounts. This is clearly a very constructive proposal that bares no resemblance to a number of Republican plans that create personal accounts without bringing significant new resources to Social Security. Yet, it’s not clear whether Democrats would give this approach the slightest consideration. ‘‘I’m not even going to talk about private accounts,’’ House Ways and Means Chairman Charles Rangel, Democrat of New York, said after Democrats regained control of Congress in 2006.6 Based on the rhetoric, one might think that there was something sacrosanct about at least 12.4 percent of each worker’s wages flowing directly to Social Security, leaving no opportunity for personal savings. But this appears to be a somewhat arbitrary line. It turns out that, under the LMS plan, revenue from the full 12.4 percent payroll tax and a bit more would still flow directly to Social Security, at least after a brief phase-in period. That’s because, while Social Security gets the vast majority of its revenue from the payroll tax, it also gets revenue from taxes on the benefits of individuals with income of more than $25,000 and couples with income of more than $32,000. By 2013, these taxes are projected to bring in revenue in excess of 0.5 percent of wages subject to the payroll tax,7 which would be more than enough to finance the only portion of the LMS account plan that carves revenues out of Social Security. What is more, in the LMS plan, much of the cash directed to accounts from Social Security’s existing revenue base would come within the first decade as the tax increases are phased in gradually. But if the full tax increases were applied from the start, rather than phased in over nearly a decade, and the size of the accounts was trimmed to about 2.7 percent of wages instead of 3 percent, then they could be financed entirely out of new revenue. Nevertheless, it seems virtually certain that Democrats would close ranks against this plan. That is because the Democrats’ real agenda isn’t simply to block a carve-
A BRIDGE TOO FAR
89
out of Social Security’s resources, but to direct significant new resources to preserve the program’s traditional benefit structure. A closer look at the LMS plan reveals why Democrats have held firm against any plan that trades a hefty portion of Social Security’s traditional benefits for personal account benefits. The safety net ensured by the proposal leaves a good deal to chance, because it is highly dependent upon investment returns, even for very modest earners. Consider first how a career-average earner ($42,000 in 2009) would fare under the LMS plan after all changes are phased in after 2055, assuming investment returns equal to safe government bonds. At that point, an average earner would face a 9 percent reduction in their Social Security income compared to promised benefits, and a worker who retires at 65, the new early-retirement age under the LMS plan, would face an additional 20 percent penalty, partly a reflection of the expected retirement age being moved to 68. Together, these reductions could cut Social Security annual income by 25 percent, compared to the current law Primary Insurance Amount, or base benefit. On top of that, while workers wouldn’t be eligible to receive their traditional Social Security benefit until 65, they would be able to cash in the account portion of their benefit for a lifetime annuity as early as 62, which would result in a further cut in their personal-account annuity. Under this scenario, despite the 1.5 percentage point tax increase and higher eligibility age, retirement income would be roughly in line with that available to one who retires at 62 under the current system and faces a 30 percent early-retirement penalty. Still, because of all the extra revenue the plan brings to Social Security, the benefit reductions in the LMS plan are moderate compared to many other proposals. But what if investment returns are unusually poor—worse even than the return on government bonds? To some extent, the risks would be limited; worker contributions would be invested in broad funds that track the performance of major stock indexes, which are far less volatile than individual stocks. Some portion of the accounts would also go into high-grade corporate bond funds and Treasuries. But because future investment returns are uncertain, it makes sense to ensure that the safety net can withstand negative outcomes. Yet, as show in table 13.2, even modest earners would have substantial exposure to investment risk in the LMS plan. An average earner would face a nearly 44 percent cut in their traditional benefit, leaving the personal account to make up the difference. And even low-wage earners would face a 37 percent cut in their guaranteed benefit. While above-Treasury investment returns would limit benefit reductions, the uncertainty tied to financial markets makes it of greater importance to focus on the potential downside when considering retirement income adequacy. As shown in table 13.3, if investment returns are historically poor—yielding just 1.5 percent above the inflation rate, compared to the projected 2.9 percent return for government bonds —that could open a significant hole in the safety net. Under such a scenario, Social Security could replace as little as 28.2 percent of wage income for an average earner, or about $11,850 based on today’s wages. On top of the cuts to retirement income, these workers would face higher payroll taxes, making it somewhat more difficult to save outside Social Security, and they’ll also have to wait until age 65 to claim benefits.
90
A WELL-TAILORED SAFETY NET
Table 13.2 Benefits under LMS plan in 2055 with Treasury returns Annual income (2009 wage level)
Traditional benefit as % of promised benefit
Personal account annuity as % of promised benefit
Total LMS benefit as % of promised benefit
$18,900
62.6%
23.0%
85.6%
$42,000
56.2%
30.9%
87.1%
$67,200
54.2%
37.4%
91.6%
$106,800
52.9%
50.4%
103.3%
Source: Adapted from Social Security Administration, Office of the Chief Actuary, ‘‘Estimated Financial Effects of ‘A Nonpartisan Approach to Reforming Social Security—A Proposal Developed by Jeffrey Liebman, Maya MacGuineas and Andrew Samwick,’’’ November 17, 2005.
At a 2006 forum at the American Enterprise Institute (AEI), David Certner, legislative policy director at AARP (formerly the American Association of Retired Persons), summed up what he found unacceptable about the LMS plan. ‘‘The low earners are getting less . . . and obviously taking on more risk as well, and the high earners are actually getting more.’’ 8 A closer look at table 13.2 reveals that, assuming investment returns equal to government bonds, the highest wage earners would actually see a net increase in their Social Security income compared to current benefit promises, while average and below-average earners would face substantial reductions. This shift to a less-progressive distribution of benefits is tied to the plan’s personal account structure. While the traditional Social Security system provides for some degree of income redistribution from higher earners to lower earners, personal accounts do not. As seen in table 13.2, over a full career, accounts equal to 3 percent of wage income would replace more than 50 percent of Social Security income for top earners but just 23 percent for low earners—a 27 percent gap. But the LMS plan makes up for less than half of this gap by preserving more of the traditional benefit for low earners. Thus, the net result is a regressive shift in the benefit structure. Table 13.3 Benefits under LMS plan in 2055 with 1.5% real returns Annual income (2009 wage level)
Personal account annuity as % of base benefit
Total LMS benefit as % Total LMS benefit as % of base benefit of base benefit (retire at 67) (retire at 65)
$18,900
17.5%
80.1%
69.4%
$42,000
23.2%
79.4%
68.8%
$67,200
28.1%
82.3%
71.4%
$106,800
37.8%
90.7%
78.6%
Source: Author’s calculations of impact of lower returns on benefit levels reflected in Social Security Administration’s analysis of LMS plan from November 17, 2005.
A BRIDGE TOO FAR
91
In theory, the LMS authors could have proposed the same 3 percent accounts without sacrificing any degree of progressivity. However, they determined that less progressivity in the benefit structure was warranted in light of ‘‘the fact that high earners under the plan are paying a disproportionate share of the new revenues.’’9 Recall that their plan would apply the full 12.4 percent payroll tax on income over the cap (up to roughly $195,000 based on 2009 wages) but award no benefits based on those additional contributions. Their plan summary suggests a concern that imposing benefit cuts on the highest earners in addition to disproportionately large tax hikes might ‘‘undermine support for the universal social insurance system.’’10 Interestingly, the AEI forum revealed that it was Liebman, the Democrat in the group, who insisted on the less-progressive distribution of benefits. However, the questions of equity raised by this plan are one of the primary reasons why the sensible goal of LMS to store up a large amount of savings to ease the future burden of caring for an aging population may not be practical to achieve on the scale they recommend, at least in the context of Social Security reform. To understand why, consider this: Under the LMS plan, a low earner ($18,900 in 2009) would face an implied 3.4 percent annual tax rate increase, while someone who earns the maximum income subject to Social Security taxes under current law ($106,800) would face just a 1.3 percent implied rate hike. This finding, detailed in table 13.4, is further explored in chapter 16, which provides a more in-depth look at the progressivity of various Social Security plans. Another key question facing such large personal account plans relates to risk—not merely investment risk but political risk. While the authors propose that account balances be used to buy a lifelong annuity upon retirement, if such a requirement can’t withstand political pressure, then individuals might draw down these balances and be left to subsist on their reduced traditional benefit. Finally, there is a question about the degree to which the economic benefit from storing up savings could be diluted by disincentives associated with a significant increase in the marginal tax rate on work income for higher earners. In this regard, it’s important to consider that there may be upward pressure on tax rates outside of Table 13.4 Implied tax hikes under LMS plan—a narrow view Annual income (2009 wage level)
LMS plan % benefit reduction
Implicit tax hike as % of income from LMS benefit reduction
LMS tax hike as % of income, including 1.5% payroll tax hike
$18,900
14.4%
1.9%
3.4%
$106,800
-3.3%
-0.2%
1.3%
Note: Benefit reductions reflect Treasury-level investment returns. Comparison reflects a theoretical worker who earns the maximum amount subject to the Social Security payroll tax; at some point in their careers, most workers with earnings near this level would also face the LMS plan’s 12.4 percent tax on income over the payroll ceiling. Source: Author’s calculations based on Social Security Administration, Office of the Chief Actuary, ‘‘Estimated Financial Effects of ‘A Nonpartisan Approach to Reforming Social Security—A Proposal Developed by Jeffrey Liebman, Maya MacGuineas and Andrew Samwick’’’ November 17, 2005.
92
A WELL-TAILORED SAFETY NET
Social Security to address other priorities while avoiding a worrisome increase in federal debt. For all of these reasons, this compromise appears to be a bridge too far for both political parties. Because the LMS plan exposes modest earners to significant risk in addition to benefit cuts, Democrats can argue that it doesn’t meet the challenge of ensuring a robust safety net for our aging population. Yet Republicans, and perhaps some Democrats, are likely to balk at the extent of the tax increases in a plan that raises more revenue than any alternative proposal. So the logical question is whether their differences can be reconciled. Wouldn’t smaller tax increases conflict with the Democrats’ goal of a more dependable safety net, rendering a compromise impossible? Not necessarily. The key to a compromise is to use the resources now flowing into Social Security more efficiently. By redesigning the safety net so that support is strongest where it needs to be strongest, and fewer resources are expended where support is less critical, we can limit both risk and sacrifice for the lowest earners, while limiting tax hikes and still providing a safety net that meets the needs of higher earners.
Chapter 14
The Hilda & David and Irene & Bernie Tests
I often think about Social Security in terms of my grandparents, Irene & Bernie and Hilda & David. I had the rare privilege of celebrating both couples’ 60th wedding anniversaries over Thanksgiving weekend in November 1994. And although they lived worlds apart, financially speaking, both would come to depend on Social Security. David operated a small print shop in Brooklyn, and he and Hilda retired to a modest one-bedroom condo in Fort Lauderdale. For them, Social Security meant everything—comfort, dignity and the peace of mind to enjoy their retirement. And they were grateful for every day they got. Bernie worked on Wall Street for much of his career, and he and Irene even lived on Park Avenue for a few years before they retired to an elegant, but hardly luxurious, beachfront condo in Palm Beach. He worked until he was well past 70, waiting to claim Social Security in order to earn a delayed-retirement bonus so that he and Irene would have a greater degree of income security if they lived to a ripe old age. And only partly because Bernie was a dreamer until the end and took some ill-advised financial risks, Irene now counts on Social Security like a security blanket on a stormy night. At 94, she is still going strong and hopefully will be relying on Social Security for years to come. The experience of my grandparents demonstrates why Social Security must preserve a robust safety net in very old age for people of all income levels. What was true for them is only likely to grow more so for future retirees as longevity increases and the era of employers providing workers with lifelong defined benefit pensions is left behind. But that doesn’t mean that keeping the promise of income security in very old age is incompatible with a reform of Social Security that helps ease the nation’s daunting budget challenges; rather, my grandparents’ examples point the way to a new framework for reforming Social Security in a way that is both affordable and effective.
94
A WELL-TAILORED SAFETY NET
Their experience suggests to me that any sensible reform of Social Security needs to pass both the Hilda & David test and the Irene & Bernie test. In other words, it needs to make sense not just for the working class, but even for the upper-middle class. The Hildas and Davids of the future are likely to need solid income support from Social Security throughout their retirement years. And if they live as long, or longer, than my mother’s parents did, they’ll grow to rely even more heavily on Social Security in their twilight years. The Irenes and Bernies of tomorrow may not need to depend much on Social Security in their 60s. But when they get into their 80s and especially their 90s, they too are going to need a robust safety net. Since we can’t realistically afford to pay all benefits, we should try to make changes where they will do the least harm, and there is a general consensus that we should ask for less sacrifice from those who depend on Social Security for the bulk of their retirement income. Because modest earners have much less capacity for sacrifice, a onesize-fits-all approach to Social Security reform simply can’t save very much without unwinding critical protections. Therefore, a cost-efficient approach to Social Security reform has to be very progressive, with the burden of sacrifice falling more heavily on higher earners. Most Social Security reform plans to date have proposed bigger benefit cuts for higher earners than lower earners, and that makes perfect sense, but only up to a point. For example, it’s perfectly sensible to prescribe bigger future benefit cuts at age 65 for workers earning $60,000 a year than for $30,000-earners. But what about when those $60,000-earners are 90? Should they still face significant benefit cuts? At some point, it’s no longer sensible—or progressive—to substantially reduce benefits for retirees who live to a very old age and face great risk of outliving their savings. Because other Social Security reform proposals prescribe a level of benefit reductions for retirees at age 65 that will also apply when they turn 95, policy makers are faced with two bad choices: Either they prescribe significant cuts that unwind the safety net in very old age, or else they have to limit the need for cost-saving measures and rely heavily on tax increases. Yet even the Liebman-MacGuineas-Samwick (LMS) and Diamond-Orszag plans—which both raise substantial new tax revenues for Social Security—agruably fall somewhat short when it comes to the ultimate test for Social Security reform—does it ensure a very solid safety net for seniors late in life? While the goal of an effective safety net requires minimizing benefit cuts in very old age, the twin goals of affordability and effectiveness could both be served by shifting the deepest benefit cuts to the initial years of retirement. Frontloading benefit cuts could help shift Social Security’s incentives more toward delayed retirement, making it less likely that individuals would retire too early thinking their savings and benefits would be sufficient, only to realize too late that they had misjudged their needs. Although it’s neither sensible nor progressive to have a safety net that tempts people to retire too early and be left to scrape by on an inadequate level of incomesupport in very old age, Social Security’s early retirement age of 62 sends a clear signal that retiring at 62 is a sensible choice. Yet there’s little reason to think it will
THE HILDA
&
DAVID AND IRENE
&
BERNIE TESTS
95
continue to be sensible as benefits are reduced, longevity increases and earlyretirement penalties reach 30 percent. Raising the earliest retirement age to 64 while lifting the expected retirement age an extra year to 68 would limit the maximum early-retirement penalty to 25 percent, providing both cost savings to Social Security and bigger benefit checks for many retirees. Logic suggests that such a shift in the retirement ages and the frontloading of benefit cuts are the most constructive ways to cut Social Security’s costs because, of all the options, they would be most likely to keep workers on the job longer and produce a higher level of retirement income. However, the idea of significant benefit cuts early in retirement raises questions about both equity and benefit adequacy. Because higher earners tend to live longer lives, and disparities in life expectancy appear to be widening, frontloading benefit cuts could disproportionately impact lower earners. In addition, higher earners generally have more capacity to save and more pension coverage than lower earners, often perform less-strenuous jobs and tend to have a broader range of work skills, all of which suggest they are better able to handle less support from Social Security in their sixties. All of these important concerns are reflected in the design of A Well-Tailored Safety Net, which would firm up much earlier in retirement for lower earners than higher earners. But benefit cuts would gradually unwind, leaving a robust safety net in very old age for people at all income levels. There’s no other way we can achieve a costefficient Social Security program that meets the needs of both the working class and the upper-middle class. The most constructive—and user-friendly—way to approach Social Security reform is to focus on the actions individuals must take to overcome reduced benefits —saving more and working longer. Both of those responses need to be built into any reform proposal, if we want to have an effective safety net. As with other proposals that seek to preserve a viable safety net, benefit cuts are only one part of the equation. But the same kind of considerations that yield a progressive approach to cost-savings also dictate a progressive approach to revenue increases. Just as the frontloaded approach to benefit reductions needs to reflect the greater difficulty lower earners may have in extending their careers, any tax increases should reflect the fact that lower earners are relatively cash-constrained and would have greater difficulty adapting to less support from Social Security in the early years of eligibility. Yet, while there is a very strong case for a progressive approach to both benefit cuts and revenue increases, an overly progressive approach could be inconsistent with the key goals of effectiveness and affordability, and it would call into question Social Security’s fairness. As noted in the prior chapter, this risk of making Social Security so progressive that it could come to be seen more as an income-transfer program than a social– insurance program led LMS proposal co-author Jeffrey Liebman to resist any benefit cuts for the highest earners who would be hit with a disproportionately large tax
96
A WELL-TAILORED SAFETY NET
increase under his plan.1 Yet, the decision to at least partly compensate for the highest earners’ extra tax burden on the benefit side of the equation produced an unfortunate outcome: a regressive shift in Social Security’s distribution of benefits. Liebman was surely right to want to provide a meaningful safety net even for the upper-middle class, and right to want to treat higher earners fairly. But these goals don’t require that we shield higher earners from benefit cuts. Rather, they suggest that disproportionately large tax increases on higher earners are ill-suited to redesigning Social Security in a way that that is affordable, effective and fair. The clearest path for achieving an affordable and effective Social Security reform is to prescribe tax increases and benefits cuts that are roughly proportional to income, a reasonable gauge of the ability to sacrifice, but to have the benefit cuts gradually unwind in order to ensure a robust safety net in very old age for all income groups. These are central principles for Social Security reform that I will translate into specific recommendations in coming chapters.
Part IV
Laying Out a Solution
Chapter 15
Old-Age Risk-Sharing
As I argued in the prior chapter, if we want an effective safety net, then Social Security reform can’t simply be about cutting away support. Because benefit levels aren’t especially generous to begin with—and even less so once penalties for early retirement are factored in—policy makers need to prescribe changes that workers can realistically respond to and retire with solid support. The two ways that workers can adjust their behavior to cope with benefit changes are by saving extra and working longer. To the extent that it is feasible for them to do so, asking workers to extend their careers to retire with full benefits is preferable to relying on additional saving. This seems particularly true when it comes to modest earners who are already strapped for cash; because policymakers can’t presume that those who have little to spare will increase their saving voluntarily, an explicit tax increase, i.e. more forced savings, is the only way of ensuring that a benefit cut doesn’t result in a cut in retirement income. Further, the likely upward trajectory of tax rates to cover the much bigger budget challenge tied to health care reinforces the case for reforming Social Security in a way that limits the need for tax increases. Therefore, the first and arguably most important step in reforming Social Security is to close as much of the program’s financing shortfall as is feasible in a way that is tied to extended careers, rather than explicit tax hikes. Yet doing so without undercutting the retirement safety net requires a recognition that workers may not be any more responsive to voluntary efforts at promoting delayed retirement than they are to voluntary savings exhortations. As a starting point, it’s worth considering that we would have to raise the expected retirement age roughly to 71 in order to close about half of Social Security’s entire funding shortfall.1 This, of course, is out of the question. In fact, the Social Security actuaries only provide estimates for the savings that would come from raising the retirement age as high as 70,2 probably because nobody who wants to be reelected would think about raising it higher. But even raising the retirement age to 70 is the
100
A WELL-TAILORED SAFETY NET
wrong approach if we want to have a safety net that does an effective job of protecting those who live into their 80s and 90s and face increasing risk of outliving their savings. There are two big problems with a retirement age of 70. First, if we raise the retirement age that high while keeping the early retirement age at 62, then those retiring early could face a 43 percent penalty for early retirement.3 Many people might continue working to avoid such a severe benefit reduction, but those who are most desperate for extra income might well grab hold of a government lifeline at the first chance they get, even at the cost of income security in old age. Alternatively, some have suggested raising the early retirement age to 65 in tandem with an increase in the expected retirement age to 70.4 This would limit the maximum early retirement penalty to 30 percent—the same penalty workers will face for retiring at 62 once the retirement age rises to 67. But there’s good reason to conclude that this 30 percent cut in a benefit that isn’t particularly generous to begin with is too big to be consistent with an effective safety net in very old age. Consider that after a 30 percent early-retirement penalty, a career $30,000-earner now reaching the Normal Retirement Age (NRA) would get about $8,370 a year after Medicare premiums, or roughly $160 a week. That seems like a long way from income security, particularly for those who live long enough to deplete their savings. In effect, increasing the expected and earliest retirement ages to 70 and 65, respectively, would only move—rather than repair—the large crack in Social Security’s foundation. So why not lift the earliest retirement age even further—to 66 or 67—if that would do more to ensure a robust safety net in very old age? The reason is that Social Security needs to strike an appropriate balance between many workers’ need for income-support at the end of their careers and the need for income security late in life. Right now, the balance is tilted toward supporting end-of-career income needs to such an extent that it may actually encourage workers to end their careers early. Moreover, as longevity increases, the appropriate balance will shift more in the direction of income security in very old age. Still, it will be important to strike the right balance so as not to put a disproportionate and perhaps unrealistic burden on lower-income workers. There is a great risk that lifting the eligibility age too high and, in effect, pulling out the floor of income support, could set up many workers for a hard fall. It should be noted that the idea of raising the NRA to 70, which was central to the Social Security debate in the late 1990s, rarely surfaces anymore. In recent years, only a single lawmaker has proposed raising the NRA past 68.5 Yet this apparent shift is less than meets the eye; rather the difference is largely one of packaging. Now, instead of explicit increases in the retirement age, the more common approach is to propose reducing benefits based on a formula that takes into account increasing longevity. Politicians have apparently embraced this latter approach because—unlike a straightforward increase in the so-called Normal Retirement Age—it doesn’t carry an implication that workers will have to stay on the job longer. But while indexing benefits for longevity may be less visible to the general public, the end result is pretty much
OLD-AGE RISK-SHARING
101
the same: Workers would face a comparable benefit reduction unless they do stay on the job longer.6 Although raising either the earliest or expected retirement ages under Social Security is politically unpopular, pretty much everybody is supportive of encouraging workers to voluntarily delay retirement. Workers who claim Social Security benefits at 63 instead of 62, for example, would receive a benefit check that is more than 7 percent bigger each year. A retiree who worked until 65 would get a benefit nearly 24 percent bigger than if they worked to 62. And to the extent that more workers remain in the workforce longer, Social Security will take in more revenue, closing a portion of the funding shortfall without benefit cuts or tax increases. An Urban Institute study referenced in chapter 11 found that if all workers opt to extend their careers by five years, it would erase 20 percent of Social Security’s deficit in coming decades.7 Over time, as early-retirement penalties grow, benefits are reduced and longevity increases, retirement at 62 will more clearly become a bad financial decision, but for now, about half of all workers do it. To some extent, people might be settling for a benefit that looks like enough to live on, only to realize—once health expenses rise and personal savings dwindle—that they retired too early. But for many workers, waiting to receive Social Security benefits presents a difficult trade-off: Not only would workers give up a year of potential retirement, but if they die in their 70s, they could end up getting less income from Social Security than if they had retired at 62. All of this begs a question: If delayed retirement is good for workers’ finances and good for Social Security’s finances—not to mention the economy—then shouldn’t we make it a more attractive option for workers to extend their careers? The logic for improving the incentives—and reducing the disincentives—for delayed retirement is compelling, especially when it can be done in a way that solves much of Social Security’s funding shortfall while bolstering the safety net in very old age. One logical way of reducing the disincentive for delayed retirement involves personal accounts and a right of inheritance. The risk that many workers who delay retirement will leave money on the table if they die in their 70s will be diminished if they can leave a portion of their Social Security contributions to their heirs. I’ll discuss this more fully in chapter 18. But personal accounts aren’t a money-saver, by themselves, so the more pertinent question is this: How can we tilt the scales in favor of delayed retirement in a way that shores up Social Security’s finances? Doing so requires providing less of an incentive to retire early, and that means even steeper benefit cuts for those who do. But in designing such an approach, it’s important to keep one thing in mind: We simply don’t know the extent to which shifting incentives aimed at producing more work can actually motivate more work. Unless we approach this policy challenge with a healthy skepticism, we could build in unrealistic expectations that won’t be met and further weaken the safety net in very old age. That overriding concern about preserving a robust safety net in very old age led me to develop an alternative to the two modest cost-saving ideas on the table that would restructure benefits to favor increased work.
102
A WELL-TAILORED SAFETY NET
Social Security now awards benefits based on a formula that derives workers’ average career wage level using their highest 35 years of earnings,8 but one reform option would be to factor in 40 years of earnings.9 Because five lower-earnings years would be averaged into the total, this would yield a lower career wage level and, as a result, somewhat smaller benefit checks, particularly for those who didn’t earn anything in some of those extra years. The appeal of this approach is that workers would have a greater incentive to work a 36th year and longer because that extra work would have a measurable effect in increasing their Social Security income in retirement. Such a change would promote delayed retirement, while also addressing one facet of Social Security that raises a question of fairness. As the Urban Institute’s Eugene Steuerle told Congress in 2005, ‘‘Someone who works 45 years at $35,000 gets substantially fewer benefits than someone who works 35 years at $45,000.’’10 Yet, in general, counting only 35 years of work toward benefits contributes to Social Security’s progressive distribution of benefits, even if it may not be the most constructive way of achieving that result. That’s why, as seen in table 15.1, counting 40 years of work could have a disproportionate impact on the lowest earners, who tend to work fewer years and who could face a 7 percent benefit cut, on average. Still, it is not a given that lowest earners would fare the worst under this approach. A bipartisan Social Security reform plan from two former Congressmen, Arizona Republican Jim Kolbe and Texas Democrat Charlie Stenholm, adopted the 40-year average-wage formula but also proposed a relatively generous poverty-related minimum benefit that would negate the reductions at the bottom end of the income spectrum.11 Kolbe and Stenholm’s recommendation that 40-year workers receive a benefit equal to at least 120 percent of a wage-adjusted poverty level—one that keeps pace with wage growth instead of merely inflation—served as the basis for the minimum benefit in A Well-Tailored Safety Net. Holding harmless the lowest wage earners would, of course, reduce the cost savings to Social Security from counting 40 years of work, which is among the reasons that the Kolbe-Stenholm version of this plan might erase less than 10 percent of Table 15.1 The regressive impact of a 40-year benefit formula in 2050 Household income level
Number of zero-wage years under a 40-year benefit formula
Average benefit reduction
Highest quintile
0.9
3.4%
nd
1.9
4.4%
2.7
5.3%
4.4
5.9%
13.2
7.1%
2
highest quintile
Middle quintile nd
2
lowest quintile
Lowest quintile
Source: Mark Sarney, Office of Retirement and Disability Policy, ‘‘Distributional Effects of Increasing the Benefit Computation Period,’’ Table 2, Table 5, August 2008, http://www.socialsecurity.gov/policy/ docs/policybriefs/pb2008-02.pdf.
OLD-AGE RISK-SHARING
103
Social Security’s shortfall.12 Another reason is that their plan counts wages from the 41st year of work and beyond in the average-wage tally but still divides by 40 years. For workers with long careers, this change could have the effect of raising both their average wage and benefit payments compared to the 35-year formula. Although the goals of the Kolbe-Stenholm approach are worthy—incentivizing longer careers while making Social Security more affordable and equitable—I opted against incorporating such an approach for two principal reasons. First, the inequitable treatment of those who work relatively longer careers would be ameliorated, over time, due to the progressive saving offset outlined in chapter 16. For example, career-average earners would be required to deposit an extra 1 percent of wages in a personal account that would, after a full career, cover 10 percent of that worker’s Social Security benefit. Average earners who work for more than 40 years would see a net benefit increase under this provision because the account accumulations would exceed the 10 percent offset. On the other hand, those who work 35 years would see a slight decrease. However, because the saving offset would phase out at life expectancy as the account balances are drawn down, 35-year workers wouldn’t face any benefit reduction in very old age compared to Social Security’s current formula. This points to the second reason: Unless this change in the benefit formula proved to be a powerful work incentive, which is far from certain, it would create both winners and losers relative to the present system, and the losers would see benefit reductions regardless of whether they were 62 or 92. Although the Kolbe-Stenholm version of the plan avoids a hit to the lowest earners that would make this approach more objectionable, modest-wage earners whose benefit falls somewhat above the poverty threshold could still face a significant cut in the context of a reform provision that does relatively little to help Social Security’s finances. Thus, it’s unclear the extent to which this approach would be of help in achieving an affordable and effective safety net.13 Now let’s look at the other approach that has been proposed to shift incentives in favor of more work in a way that would shave about 10 percent off of Social Security’s financing gap. This idea, advocated by President George W. Bush’s 2001 Social Security commission, as well as Kolbe and Stenholm, among others, would increase both penalties for early retirement and bonuses for working past the expected retirement age.14 As seen in table 15.2, instead of a 30 percent early-retirement penalty, those who retire at 62 would be hit with a 37 percent cut. For a $30,000-earner retiring at 62, this would mean a benefit of about $8,880 a year (before Medicare premiums) instead of $9,870 a year under current law. And those penalties would remain in effect whether a retiree is 62 or 92. Retiring at 62 would certainly be less attractive under this proposal, and that might induce some people to keep working a few years. But it’s safe to say that many of the workers who go ahead and retire anyway will be those who live from paycheck to paycheck and could desperately use some extra cash. Meanwhile, those who delay retirement past 67 would get a 10 percent bonus for each extra year of work, instead of 8 percent under current law. While moderate earners might be more likely to delay
104
A WELL-TAILORED SAFETY NET
Table 15.2 Discouraging early retirement at the expense of income security
Age of retirement
Current law earlyretirement penalty
Proposed earlyretirement penalty
Proposed annual benefit for $30,000-earner (2009 wage level)
67
0%
0%
$14,100
66
6.7%
8.3%
$12,930
65
13.3%
16.7%
$11,750
64
20%
25%
$10,580
63
25%
31%
$9,730
62
30%
37%
$8,880
Note: Early retirement penalties would be somewhat smaller than shown prior to 2022, when the Normal Retirement Age will reach 67. Source: Adapted from Office of the Chief Actuary memorandum of January 31, 2002: ‘‘Estimates of Financial Effects for Three Models Developed by the President’s Commission to Strengthen Social Security,’’ Model 3 Basic Provisions, Modify Actuarial Reduction and Increment Factors, p. 11,http:// www.ssa.gov/OACT/solvency/PresComm_20020131.pdf.
retirement, this incentive is most likely to help higher earners who would have kept working anyway. There’s certainly nothing wrong with helping workers to overcome benefit cuts by increasing the bonus for delaying retirement, but it doesn’t make sense to do it with a proposal that would disproportionately hurt those who will be most hard-pressed by having to extend their careers—including the lowest wage earners. In contrast to the 40-year benefit formula discussed above, the bigger early retirement penalties would not be offset by the poverty-related minimum benefits in both the Kolbe-Stenholm and 2001 commission plans, as those minimum benefit levels apply to the Primary Insurance Amount, which is set ahead of adjustments for early retirement. While it makes an enormous amount of sense to provide stronger incentives to keep working and less incentive to retire early, we need to do it in a way that—at the very least—doesn’t widen the crack in Social Security’s foundation. In other words, we shouldn’t further penalize retirees by unwinding the safety net they will depend on in very old age just because they decide to claim their Social Security benefit as soon as the government makes it available. Yet, because we can’t count on a vigorous behavioral response to increased incentives for delayed retirement, it’s a fair bet that this 2001 commission approach would do just that. So how can we discourage early retirement and improve incentives for delayed retirement, while finding enough savings to close a big part of Social Security’s funding shortfall, and still provide critical income-security protections in very old age? Finding the answer to that question gets at the essential challenge of Social Security reform: How can we best put to use each and every Social Security dollar in providing an efficient and effective safety net? None of the cost-saving measures on the table offers a satisfactory answer to that question, so I developed a new approach that I call Old-Age Risk-Sharing because
OLD-AGE RISK-SHARING
105
it is designed to address the great risk that retirees will live long enough to deplete their savings—a risk that is growing as longevity increases. While today’s Social Security program provides the same benefit for retirees at 62 that they’ll get at 92—which makes it easier to retire early, but often at the cost of income security late in life— Old-Age Risk-Sharing would direct a greater share of Social Security’s limited resources where they are most critically needed. Benefit cuts would be steepest when workers are in their early 60s to discourage them from retiring too early. But unlike in the 2001 commission approach discussed above, those cuts would phase out in order to keep the safety net intact and avoid benefit cuts in very old age, when people can least afford them. Meanwhile, delayed retirement incentives are greatly improved, offering workers who extend their careers the opportunity to overcome benefit cuts early in retirement and obtain greater income security in old age. The key distinction between Old-Age Risk-Sharing and the two reforms discussed above that link cost savings with program changes to spur delayed retirement is that only Old-Age Risk-Sharing is designed with the equivalent of a fail-safe function. If the threat of lower benefits implied by basing benefits on 40 years of work or applying steeper early-retirement penalties fails to produce the desired goal of additional work, then these approaches would further weaken a safety net that is already becoming threadbare for those who live to a very old age. While a failure of either policy, on its own, might not have a dramatic impact on the safety net, the reality is that neither of these approaches takes us far toward a solvent Social Security system, so major benefit reductions will be needed outside of these policies. That makes either approach a questionable building block for a comprehensive reform proposal. By contrast, Old-Age Risk-Sharing, which would close roughly 33 percent of Social Security’s full financing gap, is designed to ensure both a robust safety net in very old age and adequate income support early in retirement—even if efforts to encourage delayed retirement do not have a significant impact. One way Old-Age Risk-Sharing builds in fail-safe protections is by making benefit cuts progressively smaller for lower earners to avoid cutting away critical safety net protections for those who have little personal savings to fall back on. Yet because the benefit cuts will phase out for everyone, even higher earners will have a safety net they can count on in very old age. And for workers who extend their careers in order to earn more generous delayed retirement benefits, these benefit cuts would effectively phase out much more quickly and leave retirees with an added degree of income security in old age. So how would this work in practice? The best way to understand the implications of a reform proposal is to look at how it would change benefits for workers who experience its full impact. So for this proposal, which would phase in over 20 years starting in 2013, that means looking at its impact on workers who will turn 62 in 2032 and beyond. When Old-Age Risk-Sharing is fully phased in at that point, $21,000earners would face a 10 percent benefit cut at retirement. But, as seen in table 15.3, their benefit would automatically increase 0.5 percentage points a year for the
106
A WELL-TAILORED SAFETY NET
Table 15.3 Old-Age Risk-Sharing for low earners in 2035 Age
Old-Age Risk-Sharing benefit cut
Benefit as % of promised benefit (before other provisions)
65
10%
90%
70
7.5%
92.5%
75
5%
95%
80
2.5%
97.5%
85 and up
0%
100%
Note: Benefit levels assume worker claims benefits at 65.
next 20 years—on top of cost-of-living increases—until they receive their full scheduled benefit. Ideally, though, these workers would decide to delay receiving benefits until one year past the new Normal Retirement Age of 68. As seen in table 15.4, one delayed retirement credit equal to 10 percent of their base benefit for each extra year of work would enable these workers to retire with their full annual benefit. Thus, for these low earners, 69 would be the Target Retirement Age. And after 69, benefits would continue to rise 0.5 percentage points a year, eventually reaching 110 percent of the promised benefit. While these lower earners might have to economize or work part-time if they claim benefits before their Target Retirement Age, the safety net would grow firmer as their benefit cuts unwind. Further, as seen in table 15.5, thanks to the introduction of an enhanced minimum benefit of 120 percent of the wage-adjusted poverty level for a 40-year worker, these lower earners could still receive a more generous benefit than the current system offers. Two points are worth making here regarding the proposed minimum benefit in A Well-Tailored Safety Net that is further detailed in the appendix. First, because it would provide a smaller benefit to those who work fewer than 40 years—a 30-year worker would get a benefit equal to 100 percent of the adjusted poverty level—low Table 15.4 One extra work year erases Old-Age Risk-Sharing for low earners
Age
Old-Age Risk-Sharing benefit cut (2032 and later)
Delayed retirement credit as % of base benefit
Benefit as % of base benefit (before other provisions)
69
10%
10%
100%
74
7.5%
10%
102.5%
79
5%
10%
105%
84
2.5%
10%
107.5%
89 and up
0%
10%
110%
Note: Benefit levels reflect increase in the Normal Retirement Age to 68.
OLD-AGE RISK-SHARING
107
Table 15.5 New minimum benefit could offset full impact on low earners of Old-Age Risk-Sharing and increase in Normal Retirement Age
Age
Current law Social Security benefit
Minimum benefit after Old-Age Risk-Sharing and NRA increase for 40-year worker
Potential net benefit increase
65
$9,060
$9,360
3.3%
70
$9,060
$9,620
6.1%
75
$9,060
$9,880
9%
80
$9,060
$10,140
11.9%
85 and up
$9,060
$10,400
14.7%
Note: Benefit levels are based on the 2009 wage level of $18,900 for a low earner and reflect both the impact of Old-Age Risk-Sharing after 2032 and increase in Normal Retirement Age to 68. Source: Figures reflect Department of Health and Human Services, ‘‘2009 Federal Poverty Guidelines’’ and future income-replacement rate for low earners listed in Table VI.F10 of the 2009 Trustees Report.
earners could initially see a net benefit reduction. Second, the more generous benefits for low earners wouldn’t come at the expense of higher payroll tax rates or steeper benefit cuts. Rather the funding source would be a gradual broadening of Social Security to include the roughly 25 percent of the work force of state and local governments not currently covered under the program. Bringing all new hires of those states and localities into Social Security is consistent with the goal of a fair and proportional sacrifice, because there is nothing fair about shielding a certain class of workers from the sacrifice everyone else faces. Now let’s look at how Old-Age Risk-Sharing would affect those with higher earnings. After 40 years, career-average earners ($42,000 in 2009) would face a 20 percent benefit cut when they retire, but that would unwind over 20 years as their benefit rises 1 percentage point a year. As shown intable 15.6, with two extra years of work until age 70—their Target Retirement Age—average earners could earn a 20 percent delayed retirement credit and retire with an unreduced annual benefit.
Table 15.6
Age
Two extra work years erase Old-Age Risk-Sharing for average earners Old-Age Risk-Sharing benefit cut (2032 and later)
Delayed retirement credit as % of base benefit
Benefit as % of base benefit (before other provisions)
70
20%
20%
100%
75
15%
20%
105%
80
10%
20%
110%
85
5%
20%
115%
90 and up
0%
20%
120%
Note: Benefit levels reflect increase in the Normal Retirement Age to 68.
108
A WELL-TAILORED SAFETY NET
Table 15.7 Three extra work years erase Old-Age Risk-Sharing for high earners
Age
Old-Age Risk-Sharing Delayed retirement credit benefit cut (20532 and later) as % of base benefit
Benefit as % of base benefit (before other provisions)
71
30%
30%
100%
76
22.5%
30%
107.5%
81
15%
30%
115%
86
7.5%
30%
122.5%
91 and up
0%
30%
130%
Note: Benefit levels reflect increase in the Normal Retirement Age to 68.
As seen in table 15.7, those earning at least 150 percent of the career-average wage ($63,000 and up in 2009) would face a 30 percent benefit cut and would have a Target Retirement Age of 71, when they would be able to retire with their full benefit thanks to a 30 percent delayed retirement credit. But no matter when they retire, the benefit cut would unwind by 1.5 percentage points a year, until it phases out 20 years after retirement. This combination of smaller initial benefits and bigger rewards for delayed retirement tilts the scales in favor of more work to a greater degree than any other proposal on the table, and it does so without unwinding the safety net in very old age. Under today’s Social Security system, a worker who forgoes a benefit at the Normal Retirement Age will get a benefit the following year that is 8 percent bigger. The 2001 commission approach discussed above would provide a 10 percent increase. But under this more powerful incentive structure, an average earner would receive a traditional benefit that is 12.5 percent bigger than the one passed up in favor of work a year earlier.15 Old-Age Risk-Sharing, in effect, asks higher earners to work a bit lonDear Annger than lower earners to retire with their full benefit. This makes sense when you consider that lower earners tend have less advanced education and often begin their careers at an earlier age. What is more, life expectancy is increasing more for higher-income groups than lower earners. It’s also hard to argue, for example, that a low-paid janitor should face the same effective increase in the retirement age as a lawyer who works at a desk and goes out for tablecloth lunches. Higher earners are very likely to have greater personal and pension savings that enable them to depend less heavily on Social Security in the early years of eligibility. They also have greater resources to plan for their financial futures. And in general, higher earners will have a skill set that presents them with a broader range of work opportunities in their 60s. All of these factors suggest that higher earners are better able to prepare for a Social Security safety net that firms up later than it does now. But, in reality, even relatively high earners can have physically demanding jobs, so career income is a less-than-perfect gauge of the ability to delay retirement. This makes the design of a fair and proportional sacrifice tied to work-inducing benefit
OLD-AGE RISK-SHARING
109
changes inherently more complicated and even less exact than one tied to straightforward tax increases. Rather than prescribing deeper upfront benefit cuts to close the rest of Social Security’s shortfall via Old-Age Risk-Sharing, I made a judgment that the goals of affordability, effectiveness and fairness would be better served by a gradual increase in the Normal Retirement Age to 68 that spreads the sacrifice more equally, along with an increase in the earliest retirement age to 64 to close part of the crack in Social Security’s foundation. Together, these approaches to scaling back the promise of early retirement would erase close to half of Social Security’s shortfall without undercutting the safety net for those who live to an advanced age. This approach to reform reflects the reality that there is some degree of choice as to when non-disabled workers retire; it’s impossible to conclude otherwise when the average retirement age has fallen from 68 to 63 since 1950, even as the health of older workers has improved. That isn’t to minimize the real personal sacrifice that is required for workers to extend their careers, but there’s absolutely no question that it makes much more sense to ask for sacrifice from workers still in their sixities rather than retirees in their late eighties and nineties. And only by expending Social Security’s resources wisely, can we afford to provide a robust safety net in very old age. Consider the alternatives: deep benefit cuts in very old age, when seniors can illafford them; additional tax increases on the working class; and disproportionately large tax increases on higher earners, who are likely to face additional tax hikes to deal with the much-bigger budget problems outside of Social Security. None of these options meets the challenge of providing a safety net that is affordable, effective and fair, and all of these options may be unacceptable from a political standpoint. The real uncertainty nearly everyone will face is whether they have saved enough to live in comfort and dignity, even if they live to a very old age. Despite the tremendous financial challenges facing Social Security, Old-Age Risk-Sharing changes the program in a way that the answer is more likely to be yes. Unlike today’s Social Security system, in which early retirement is, in effect, accorded a higher degree of importance than income security in very old age, Old-Age Risk-Sharing gradually rebalances the scales to reflect the growing risk that seniors will outlive their savings. Workers will be less likely to retire too early and more likely to earn delayed retirement credits, and benefit cuts will unwind, ensuring a robust safety net for the very old.
Chapter 16
A Fair and Constructive Sacrifice
Some make the argument that the Social Security payroll tax is a regressive one because low wage earners face the same 12.4 percent rate as higher earners and the effective rate is lower for those with income above the $106,800 payroll tax ceiling. This argument focuses solely on the tax side, while ignoring Social Security’s progressive benefit structure. For example, Social Security will replace about 55 percent of wage income for someone with low career earnings ($18,900 in 2009) who retires at the expected retirement age, while average earners ($42,000) will receive a benefit equal to 41 percent of income and the income-replacement rate for workers at the top of the scale ($106,800) is 27 percent.1 Much of this progressivity in Social Security’s allotment of retirement benefits is reversed by the fact that higher earners tend to live longer and receive benefits for more years. Although a few studies have concluded that this disparity in longevity fully offsets the progressive impact of the benefit formula for those who work through to retirement, a recent analysis by the Congressional Budget Office concluded otherwise.2 But taking all life outcomes into the equation, including support for the disabled and payments to children of deceased workers, there is wide agreement that Social Security generally provides relatively higher rates of return on the contributions of lower wages earners. Weighing all of the evidence and the broad public support for Social Security, it’s reasonable to start with the presumption that the current benefit structure pretty much treats workers at all income levels fairly on a relative basis, although the data do point to an exception for higher-income one-earner couples that are advantaged by spousal benefits.3 This presumption is important in considering who should sacrifice and how much, because if Social Security treats people fairly at present, that will be less true in coming decades when there will be relatively fewer wage earners to take care of each retiree. As seen in Figure 16.1, Social Security, will unavoidably
112
A WELL-TAILORED SAFETY NET
Figure 16.1 Social Security’s diminishing rate of return
become less of a good deal for workers retiring a few decades from now, meaning they won’t get the same return on their payroll-tax contributions as the first batch of baby boomers now beginning to retire. Because lower earners have less capacity to save and overcome benefit cuts, then we should make sure—at the very least—that we don’t do anything to make Social Security’s distribution of benefits less progressive. Unfortunately, just about every cost-saving proposal on the table would disproportionately hurt low or moderate earners. How can we gauge the progressivity of a proposal? There is a very simple standard. We can simply calculate what percentage of annual wages a worker would have to save over the course of a career to overcome benefit cuts, assuming uniform investment returns equal to the long-term rate on government bonds (minus an account
A FAIR AND CONSTRUCTIVE SACRIFICE
113
fee).4 Even if a proposal doesn’t actually raise payroll taxes, the percentage of wages that must be saved in order to overcome benefit cuts amounts to an implied tax hike. Using this standard, let’s first take a look at the proposal pitched by President George W. Bush known as progressive price indexing. Benefits are now pegged to wage growth, with Social Security replacing a steady share of income for each new cohort of retirees. But under this plan, benefits for the highest earners would only grow as fast as inflation, meaning income-replacement rates would decline as economy-wide wages increase over time. This approach is certainly progressive for the lowest earners, who wouldn’t face any benefit cuts. And, on the surface, it might appear to favor moderate earners who would be partly shielded from the cuts facing higher earners. By 2075, career $42,000-earners would face a 27.9 percent benefit cut, while $67,200-earners would face a 41.8 percent cut and those earning $106,800 and up would be hit with a 49.3 percent reduction. While progressive price indexing might increase Social Security’s progressivity from a broad perspective, upon closer inspection, it doesn’t really live up to its ‘‘progressive’’ billing. As seen in table 16.1, to overcome these benefit cuts, $42,000-earners would have to save nearly twice as much of their income as $200,000-earners. And $67,200-earners would have to save more than 10 times as much as million-dollar earners. Because progressive price indexing would only partially close Social Security’s shortfall, other changes would be required. The distribution of sacrifice could change substantially if, for example, a tax was levied on income above $106,800. Yet it’s, nevertheless reasonable to examine the fairness of progressive price indexing by itself, not least because President Bush sold it as a solution that proved Social Security could be saved without tax hikes. Since the Bush plan only erased about half of Social Security’s shortfall—even without borrowing related to personal accounts—Utah Sen. Robert Bennett’s plan went a step further by also reducing benefits based on increases in life expectancy, Table 16.1 Not-so-Progressive Price Indexing Income level (2009 wages)
Progressive Price Indexing benefit cut in 2075
% of annual income needed to overcome benefit cut
$18,900
0%
0%
$42,000
27.9%
2.7%
$67,200
41.8%
3.4%
$106,800
49.3%
2.9%
$200,000
49.3%
1.5%
$1 million
49.3%
0.3%
Source: Adapted from Office of the Chief Actuary memorandum, ‘‘Estimated Financial Effects of a Comprehensive Social Security Reform Proposal Including Progressive Price Indexing,’’ Table B2, OASDI Benefit Before Offset, February 10, 2005, http://www.ssa.gov/OACT/solvency/RPozen_20050210 .pdf.
114
A WELL-TAILORED SAFETY NET
Table 16.2 Bennett plan cuts hit those in the middle Income level (2009 wages)
Bennett plan benefit cut in 2075
% of annual income needed to overcome benefit cut
$18,900
14.6%
1.9%
$42,000
38.5%
3.7%
$67,200
50.3%
4.0%
$106,800
56.8%
3.4%
$200,000
56.8%
1.8%
$1 million
56.8%
0.4%
Note: Benefit levels reflect no suspension of progressive price indexing before 2072. Source: Adapted from Social Security Administration’s February 12, 2009, analysis of Sen. Bennett’s ‘‘Social Security Solvency Act of 2009,’’ Table B, to reflect The 2009 OASDI Trustees Report, http:// www.ssa.gov/OACT/solvency/RBennett_20090212.pdf.
and the combination of the two measures would be more regressive than progressive price indexing alone. As detailed in table 16.2, under Bennett’s plan, $42,000-earners would have to set aside a higher percentage of their income than $106,800-earners, and $18,900-earners would have to save more than $200,000-earners to overcome benefit cuts. Thus, the Bennett plan, by putting more of the sacrifice on the lowest earners than the highest earners, raises an obvious question of fairness. Clearly, both the Bush and Bennett proposals would be far more progressive than simply applying uniform benefit cuts of a scale large enough to close the projected financing gap. But a more practical way of considering the progressivity of various Social Security reforms is to consider how the burden of a less-generous retirement safety net will be shared across income groups. In both the Bush and Bennett plans, as table 16.1 and 16.2 make clear, those workers whose wages don’t exceed Social Security’s payroll-tax ceiling would bear a disproportionate share of the sacrifice needed to make Social Security solvent. It follows, then, that any proposal that rules out a tax increase above the payrolltax cap of $106,800 would leave moderate earners to shoulder a disproportionate share of the burden that demographic forces will place on Social Security. Yet it’s also important to evaluate a Social Security fix in the context of the broader tax code, keeping in mind that Democrats plan to raise the top marginal income tax rate to 39.6% from 35%, while preserving the Bush-era tax cuts for households earning less than $250,000 a year. What is more, Democratic leaders in Congress have proposed paying part of the cost of universal health care by imposing a further rate increase on high earners. Thus, while the highest earners have more capacity to sacrifice, they face potentially large tax increases to address the enormous future budget shortfalls outside of Social Security. This broader perspective makes it difficult to draw hard and fast conclusions about what is fair in the context of Social Security reform, but fairness is an inevitable part
A FAIR AND CONSTRUCTIVE SACRIFICE
115
of the debate. This is especially true if the scope of reform is broadened, as I believe it should be, to settle the $2.4 trillion worth of IOUs in the trust fund that at least arguably reflect Washington’s long reliance on excess payroll taxes, instead of higher income tax rates, to fund basic government programs. Yet fairness must be informed by what is constructive for the economic well-being of the nation. To limit the drag on economic growth from both tax increases and rising federal debt levels, Social Security reform must aim to provide a robust safety for our aging population in a cost-efficient manner, and this can only be achieved with sacrifice that is broadly shared across all income groups. What’s more, only by asking for a broadly shared sacrifice can we make sure that Social Security still treats everyone fairly and merits broad political support. So what is the appropriate standard of fairness in the context of Social Security reform? The Liebman-MacGuineas-Samwick (LMS) plan provides a useful window into this question. First, let’s consider the explicit revenue increases. The proposal would raise Social Security’s payroll tax 1.5 percentage points to 13.9 percent. In addition, it would, over time, apply the current 12.4 percent payroll-tax on work income between $106,800 and roughly $195,000, based on today’s wages, affecting the highest-paid 6 percent of workers. As shown in table 16.3, the obvious inequity in this distribution of tax hikes is the disproportionately large rate increase that would face workers with wages moderately above the current payroll-tax ceiling. A $150,000-earner would pay an extra $6,960 a year in taxes, compared to $1,600 for a $106,800-earner. It hardly seems fair for someone who earns 40 percent more to face a tax increase that, in dollar terms, is more than four times as great. Explicit tax increases reflect only a portion of the sacrifice involved in Social Security reform, so it’s important to look at a plan’s full distribution of sacrifice to judge its progressivity. But this broader view, detailed in table 16.4, not only does little to resolve questions about the treatment of very high earners, but it also highlights Table 16.3 A less-than-proportional tax hike Income level (2009 wages)
LMS plan effective payroll-tax rate increase
Extra tax bill
$18,900
1.5%
$280
$42,000
1.5%
$630
$67,200
1.5%
$1,010
$106,800
1.5%
$1,600
$150,000
4.6%
$6,960
$200,000
6.2%
$12,540
$1 million
1.1%
$12,540
Source: Adapted from Office of the Chief Actuary November 17, 2005, memorandum, ‘‘Estimated Financial Effects of ‘A Nonpartisan Approach to Reforming Social Security,’’’ Detailed Plan Provisions, http://www.ssa.gov/OACT/solvency/Liebman_20051117.pdf .
116
A WELL-TAILORED SAFETY NET
Table 16.4 Distribution of sacrifice under LMS plan Income Level (2009 wages)
Implicit tax hike as % of income from LMS benefit reduction
Explicit tax hike as % of income under LMS plan
Total implied tax hike under LMS plan
$18,900
1.9%
1.5%
$650
$42,000
1.3%
1.5%
$1,160
$67,200
0.7%
1.5%
$1,470
$106,800
-0.2%
1.5%
$1,400
$200,000
-0.1%
6.2%
$12,340
$1 million
0%
1.2%
$12,340
Note: Benefit reductions reflect Treasury-level investment returns. Source: Author’s calculations reflect Office of the Chief Actuary November 17, 2005, memorandum: ‘‘Estimated Financial Effects of ‘A Nonpartisan Approach to Reforming Social Security.’ ’’ Implicit benefit reductions are derived from Table B1 and Table B1a.
concerns about the LMS plan’s regressive impact on lower earners. On the low end of the income spectrum, table 16.4 shows that the LMS plan sacrifices a substantial degree of progressivity, meaning that their approach to reform does not direct a bigger portion of Social Security’s resources where they are most critically needed. On the higher end, the LMS tax hike appears too progressive, and perhaps too large in the context of other more daunting budget challenges. At first blush, one might suppose that efforts to address these inequities on opposite sides of the income spectrum would be mutually exclusive. In reality, relative to the LMS proposal, the interests of both lower earners and higher earners would be served by a reform of Social Security that produces a fair, effective and affordable safety net. What might a fairer distribution of sacrifice look like? As a starting point, consider that it would take a 2 percent tax on all wage income—including income over the payroll-tax ceiling—to close the roughly 50 percent of Social Security’s shortfall that would remain after the provisions tied to delayed retirement spelled out in the prior chapter.5 This figure assumes that the extra tax on earnings below $106,800 would be deposited in individual accounts whose accumulating balances would permit a gradual reduction in the traditional Social Security annuity without any reduction in retirement income. It also assumes that those whose careers are cut short due to disability wouldn’t be penalized for building up smaller accounts. This approach of a flat tax increase on all wages and salaries would offer a middle ground between proposals that reject any tax increase over the payroll-tax ceiling and those that hit the highest earners with a disproportionately large tax hike. Nevertheless, there’s good reason to think that a flat 2 percentage point tax increase would neither be the fairest nor the most constructive way of closing the rest of Social Security’s financing gap.
A FAIR AND CONSTRUCTIVE SACRIFICE
117
One of the key principles running through A Well-Tailored Safety Net is that Social Security reform shouldn’t be considered in a vacuum, but rather it must take into account the broader budget realities. That’s why, for example, I emphasize the need to address Social Security’s full shortfall and not ignore the nearly one-third of it represented by the unfunded trust fund. It’s also why we shouldn’t shift estate tax revenues to Social Security and pretend that we have saved anything. Yet this perspective of how Social Security interacts with the rest of the budget doesn’t simply suggest the need for a cost-efficient approach to reform; it also points to a constructive way of closing part of the financing gap while reinforcing the safety net. With the nation’s health care system on the operating table this year, Washington’s policy surgeons have focused so intently on their outsized patient that they seem not to have noticed that Social Security is attached at the hip. That’s unfortunate, because a clearer recognition of the interconnectedness of the two policy challenges could make it easier to solve both. On top of Social Security’s demographic challenges that will leave fewer workers contributing payroll taxes for each retiree, spiraling health costs mean that a bigger portion of worker pay will be devoted to non-taxable health benefits. Social Security’s actuaries project that as rising health insurance costs outpace economic growth, untaxed compensation will increase from less than 19 percent of total pay to nearly 31 percent over the next 75 years.6 Changing this trajectory—by changing the tax-free status of employer-provided health insurance—could have a measurable impact on Social Security’s finances by bringing in more payroll tax dollars, along with more income tax and Medicare payroll tax dollars to help pay for a broad expansion of health care coverage. Yet unlike other approaches to raising tax revenue, limiting the tax exclusion for employersponsored insurance offers a big side benefit: it could help to restrain out-ofcontrol health care costs that pose the biggest long-term threat to the nation’s fiscal and economic future. Health policy analysts point to the tax-free status accorded to health insurance acquired through the workplace as among the reasons that the U.S. spends much more per-capita on health care than other developed nations, though without deriving better outcomes.7 Making employer-provided health insurance tax-free distorts economic behavior by creating an incentive for more expensive policies with low deductibles and co-pays that limit a worker’s taxable out-of-pocket health costs.8 This, in turn, has the effect of putting distance between consumers and the consequences of their health care spending—namely, the bills. Intuitively, this would seem to create the conditions for people to consume more health services than they need, while greasing the skids for the health care industry to raise prices. And closer analysis bears this out. A paper by economist Jason Furman published in the spring of 2008—before he became the Obama campaign’s economic policy director and a White House economic adviser—cited studies suggesting that ending the tax-free status of employer-provided insurance could lead to a 9–38 percent drop in health spending by the privately insured, with ‘‘little if any worsening in health outcomes.’’9
118
A WELL-TAILORED SAFETY NET
Thus, unlike the surtax on the wealthy that Democratic leaders in the House proposed to pay for health care reform, curbing or phasing out the tax benefits for employer-provided care could help ensure that health care reform doesn’t just offer short-term pain relief—subsidies to make insurance attainable for the tens of millions without it—but contributes to a cure that keeps coverage affordable over the long term. Despite this logic, and with centrist Democrats apparently cool to the House’s surtax on the wealthy, there was much doubt as of this writing as to how Washington would come up with the $1 trillion-plus realistically needed to pay for health care reform. One of the principal reasons many Democrats have resisted the idea of taxing employer-provided health insurance is that doing so could raise the tax bills of middle-class households, contrary to President Obama’s pledge to shield households earning less than $250,000 from tax hikes. Yet broadening the focus to include not just health care reform but Social Security as well could help to inform this debate. A review of the three proposals on the table that might provide a starting point for a Democratic-led Social Security fix—including the two co-authored by current White House advisers Peter Orszag and Jeffrey Liebman and another by former Social Security commissioner Robert Ball—reveals that each prescribes payroll tax increases on all wage earners. This isn’t surprising because any realistic path to closing Social Security’s shortfall without a broad-based tax increase would be likely to weaken the safety net for both the disabled and the very old. So the real question isn’t whether the middle class will face a bigger tax bill, but what kind of tax increases are most likely to produce constructive economic outcomes and the highest standard of living. In this context, limiting the tax exclusion on employer-provided health care appears much more attractive. If economists are correct that taxing employer-provided coverage can help restrain health care costs by removing counterproductive incentives and making health spending more transparent, then the result wouldn’t just be increased tax revenues; limiting the portion of a worker’s compensation devoted to health spending would also produce higher wage growth In the context of Social Security reform, there’s another reason that closing a portion of the financing gap by taxing employer-provided health coverage should appeal to those who wish to preserve a strong retirement safety net. Any extra portion of compensation that would be taxed as wages would raise a worker’s career earnings level and result in a bigger monthly benefit check relative to other means of closing that gap. By contrast, while raising the payroll-tax rate would reduce the need for benefit cuts, the extra tax dollars wouldn’t yield additional benefits under Social Security’s formula. Just how far could taxing employer-provided care go toward closing Social Security’s deficits? At least on paper, eliminating the tax exclusion could erase close to half of Social Security’s 75-year gap.10 But the extent to which Social Security’s financial improvement would flow through to the overall budget would depend on how much of the extra Social Security payroll-tax dollars—more than $5 trillion in present value over 75 years—are counted on to pay for universal health care.11
A FAIR AND CONSTRUCTIVE SACRIFICE
119
At least initially, the proposals on the table to tax all employer-provided care as wage income would rely on most of these extra tax dollars to expand coverage, yet Social Security would owe extra benefits worth well over $1 trillion in present value on the additional payroll-tax contributions.12 That’s not to throw cold water on the idea, but partly out of political realism, A Well-Tailored Safety Net adopts a much more gradual approach, targeting the more modest savings that would come from slowing the growth of payroll-tax-exempt health spending to the overall growth rate of the economy. Initially, this tax change could be designed to impact only those with the richest benefits packages, while keeping more affordable coverage options tax-free and without penalizing companies whose policies cost more because their work forces are older or sicker, or due to regional cost disparities.13 But over time, limiting the growth of non-taxable compensation to the growth rate of the economy could act as a restraining influence on even relatively moderate-priced plans. Few options for raising taxes have such a clear rationale. When rising health costs are among the biggest threats facing both government and household finances, it simply makes little sense to provide tax incentives that facilitate those price increases, put health insurance out of reach for more Americans and undermine Social Security’s future. Although the additional payroll tax dollars flowing to Social Security would come from the middle class and working class, a broader perspective reveals why limiting the tax-free status of employer-provided health care would be a progressive policy. High earners, who get the biggest benefit from the tax exclusion because they face higher income-tax rates and are more likely to have so-called Cadillac health plans through an employer, would contribute more toward covering the uninsured. Combined with other reforms to slow runaway health care costs, taxing employerprovided care could yield an increase in wages that might otherwise get eaten up by faster growth in non-wage compensation. The additional Social Security contributions would be compensated with bigger retirement benefits, relative to other options that would close a comparable portion of the program’s financing gap. Finally, it’s important to remember that the likely alternative is a bigger payroll tax increase, or else bigger benefit cuts than would otherwise be needed. As I noted earlier, after the savings introduced in the prior chapter from Old-Age Risk-Sharing and an increase in the Normal Retirement Age to 68, it would take a 2 percentage point tax on all wages to close the remaining 50 percent of Social Security’s shortfall. The savings generated from slowing the growth of payroll-tax-exempt health spending to the overall growth rate of the economy would close nearly 20 percent of the overall financing gap, limiting the remaining gap to the equivalent of a 1.3-percentage-point payroll-tax hike on all wage income.14 So let us now discuss how this remaining gap should be closed. In weighing what is fair, it’s important to consider that, as with the LMS plan, A Well-Tailored Safety Net would lift the expected retirement age to 68 and gradually hike the eligibility age for Social Security, both to reflect increasing longevity and limit earlyretirement penalties. These are changes that would disproportionately impact
120
A WELL-TAILORED SAFETY NET
modest earners, and even more so when combined with a significant payroll-tax hike. Higher earners may have the pensions and 401(k) assets to retire early anyway. But lower earners, who typically start work at an earlier age and perform more physically demanding work, may have no choice but to make up for less early support from Social Security with additional work. While more saving also could help lower earners adjust to a higher early retirement age, a substantial payroll-tax hike will only further stretch their limited resources and make it even tougher to save outside of Social Security. Thus, the need to adapt Social Security to accommodate gains in life expectancy by lifting the eligibility age offers an argument against applying the same payroll-tax hike on lower earners that is faced by above-average earners. The logic of A Well-Tailored Safety Net that focuses more of Social Security’s limited resources where they are most critically needed in order to limit the need for tax hikes suggests the appropriateness of a new standard of fairness in which sacrifice is roughly proportional to income. In the context of Social Security reform, this means smaller rate increases for those with modest wages. And it means that those with income over the payroll-tax ceiling wouldn’t be hit with a significantly bigger effective rate hike than those who earn $106,800. What exactly would this mean in practical terms? As seen in table 16.5 a worker earning up to half the average wage ($21,000 in 2009) would see a 0.5-percentagepoint payroll-tax hike; an average earner ($42,000) would see a 1-percentage-point rate hike; those earning at least 150 percent of the average wage would see a 1.5percentage-point rate hike; and all work income over the payroll-tax ceiling would face a 2-percentage-point tax hike after a short phase-in period. As in the LMS plan, the full tax increase applied to wages up to the payroll-tax ceiling would be deposited directly into individual accounts under A Well-Tailored Safety Net. Although average and below-average earners would face proportionately smaller rate hikes, 1.5 percent of wages would be collected from all workers, with any extra amount deposited in a 401(k)-type individual account.
Table 16.5 A smaller, more proportional tax hike Annual Income (2009 wages)
Well-Tailored Safety Net effective payroll-tax rate increase
Extra tax bill
$18,900
0.5%
$95
$42,000
1.0%
$420
$67,200
1.5%
$1,010
$106,800
1.5%
$1,600
$150,000
1.6%
$2,470
$200,000
1.7%
$3,470
$1 million
1.9%
$19,470
A FAIR AND CONSTRUCTIVE SACRIFICE
121
So far in this chapter I have only discussed explicit tax hikes. This is a narrow but important window in gauging a plan’s progressivity, because it is the only aspect of Social Security reform that relies on contributions that come from above the payroll-tax ceiling. Next, I’ll provide a broader view of a solvency provision that I describe as a progressive saving offset. Under A Well-Tailored Safety Net, as in the LMS plan, the full tax increase applied to wages up to the payroll-tax ceiling would be deposited directly into individual accounts. For most workers, assuming investment returns equal that of safe government bonds, the money saved up in these accounts will come close to fully offsetting benefit cuts that will fully phase in over 40 years. But, as shown in table 16.6, for above-average earners, the 1.5 percent of wages set aside in a personal account would fall short of completely offsetting benefit cuts. In effect, these higher earners will face a tax increase of greater than 1.5 percentage points, meaning that they would have to save a greater share of their income to fully offset benefit cuts. So why not simply impose a bigger tax increase and require them to save more? To some extent, individuals at these higher income levels have the resources and financial savvy to plan for a secure retirement with less government support. There is also another important factor to consider: The benefit cuts would only remain in effect until retirees reach their average life expectancy. In effect, workers would be asked to set aside more savings for their retirement that would help bridge the gap if they live as long as they can expect to, but Social Security will continue to absorb the risk that retirees will live beyond the average life expectancy. Thus, insufficient savings would not unwind the safety net for retirees in their late 80s, 90s and beyond. Nor, under this approach, could poor investment returns pose a risk to the safety net in very old age, if risk-based investments were permitted. As further explained in the plan’s detailed description in the appendix, those whose careers are cut short by disability would receive enhanced protections. All this is in keeping with the goal of a cost-efficient Social Security program and a well-tailored safety net that treats people fairly. While it’s reasonable to ask higher earners to save somewhat more on their own to compensate for benefit reductions, we shouldn’t skimp where the safety net needs to be strongest. Under this progressive saving offset, those who have more capacity to save are asked to save more, and Social Security benefits are reduced to reflect those savings. Table 16.6 Distribution of sacrifice under a progressive saving offset Income Levels (2009 wages)
Add-on account as % of base benefit
Progressive saving offset in 2056 and later
Implicit tax hike as % of income from progressive saving offset
$18,900
3.8%
4.5%
0.6%
$42,000
10.3%
10%
1.0%
$67,200
18.7%
20.5%
1.6%
$106,800
25.2%
34.5%
1.9%
Source: Author’s calculations.
122
A WELL-TAILORED SAFETY NET
Assuming these tax and benefit changes are phased in starting in 2016, this component of A Well-Tailored Safety Net would erase 25 percent of Social Security’s $7.7trillion funding gap15, and the 2 percent tax on income over the payroll-tax ceiling, would close 13 percent of the gap. 16 Curbing the growth of spending on taxexempt employer-provided care to the rate of GDP growth would close another 16 percent of the gap, meaning that together these tax and saving measures would solve more than half of the problem. Having laid out the principal means of closing Social Security’s financing gap in the past two chapters, the next two chapters will focus not on making ends meet but on changing the program in ways that help and encourage workers to achieve better economic outcomes and a stronger level of support in retirement. Then, in the final chapter, I’ll provide a detailed view of how A Well-Tailored Safety Net measures up against alternative approaches to reform. However, providing such a detailed view, including the impact on taxes and benefits, is complicated by the nature of the provision curbing tax-exempt health care spending and the unpredictable path of health care costs. For one thing, a large percentage of workers don’t get health care through the workplace, so the provision wouldn’t directly affect them. And, ideally, this tax change will help slow the increase in health care costs, meaning that higher payroll tax receipts – and a higher level of benefits – wouldn’t be primarily the result of higher taxes, but of higher wages. Therefore, to provide a more straightforward comparison in the final chapter, I will use what might be considered Plan B—an alternative version of A WellTailored Safety Net that closes almost the entire Social Security shortfall without depending on a change in the tax treatment of health care. Instead, Plan B calls for a bigger tax increase on wages above the payroll-tax ceiling—3 percentage points versus 2 percentage points. And, as seen in table 16.7, it also prescribes bigger benefit cuts—though not bigger explicit tax increases—for above-average earners as part of a progressive saving offset that would still expire at life expectancy. While this alternative is perhaps not as fully as constructive as capping the growth of tax-free employer-provided health care coverage, providing a fallback option makes sense for two reasons. First, limiting this tax exclusion is highly controversial in some quarters, most notably among unions that have negotiated generous benefit Table 16.7 Distribution of sacrifice under a modified progressive saving offset Income Level (2009 wage level)
Progressive saving offset in 2056 and later
Progressive saving offset in 2056 and later (Plan B)
Implicit tax hike as % of income from progressive saving offset
$18,900
4.5%
4.5%
0.6%
$42,000
10%
10%
1.0%
$67,200
20.5%
25%
1.9%
$106,800
34.5%
45%
2.5%
Note: Implicit tax hike reflects phase-out of benefit cuts. Source: Author’s calculations.
A FAIR AND CONSTRUCTIVE SACRIFICE
123
packages, so the political obstacles to this approach may be difficult to overcome. Second, allowing for a clearer comparison with the leading reform proposals on the table will make it possible to look in isolation at one of the principal questions underlying this work: Can A Well-Tailored Safety Net that limits tax increases meet the nation’s fiscal challenges and responsibilities of caring for an aging population at least as effectively as the higher-tax proposals for reforming Social Security?
Chapter 17
A Helping Hand, Not an Empty Promise
Penalties for early retirement—the crack in Social Security’s foundation—already threaten to undermine the promise of income security in old age. As we’ve seen, even if Social Security could afford to pay all benefits, those who retire at 62 will face a 30 percent early-retirement penalty in each and every year, for as long as they live. That means Social Security will replace just 29 percent of wages for an average earner, instead of the 41 percent of income replaced for one who retires at the Normal Retirement Age. And if there are significant benefit cuts on top of this—as there must be—then Social Security could provide cold comfort for retirees who live long enough to deplete their savings. Simply put, we have to either give up the promise of income security in old age or scale back the promise of a government-financed early retirement. The choice, while not easy, is a clear-cut one. Trying to preserve Social Security’s inefficient and unsturdy structure would be counterproductive to the financial well-being of future retirees. ‘‘There is a striking contrast between reforms that simply reduce (Social Security) benefits and reforms that increase work effort while partially reducing benefits,’’ Urban Institute researchers have noted.1 ‘‘[W]orkers on average achieve significantly higher net incomes when additional work is involved—the earnings increases can more than offset any benefit cut—and significantly lower net incomes when only benefit cuts are involved.’’ One of the great challenges—and opportunities—of Social Security reform is to provide workers a helping hand, not an empty promise. Instead of encouraging workers to retire as early as 62, even at the cost of income security in old age, the goal should be to help them adapt to less support from Social Security in the early years of eligibility and extend their careers long enough to retire with real income security. But up until now, the few Social Security proposals that have called for an increase in the earliest retirement age have mostly ignored the challenge—and in some cases hardship—that workers may face if they can’t count on any support from Social
126
A WELL-TAILORED SAFETY NET
Security in their early 60s. Addressing the challenge is an integral part of A WellTailored Safety Net. There is no sensible alternative to raising the age at which workers can claim their Social Security benefits, and A Well-Tailored Safety Net would gradually raise the earliest retirement age to 64. This change would go some distance in closing the crack in Social Security’s foundation by limiting the maximum early retirement penalty future retirees will face to 25 percent, instead of 30 percent under current law. Because other benefit cuts will phase out later in retirement, this means a retiree in very old age would get a benefit that is at least 7 percent bigger under A WellTailored Safety Net than someone who claimed their full benefit at 62—even if no Social Security cuts were necessary. For an average earner, that would be the difference between receiving $12,920 and $12,050 in annual benefits. Lower earners would do even better compared to today’s Social Security program because of the addition of a generous minimum benefit to help keep them above the poverty line, as adjusted for future wage growth. The extra support provided to low-wage earners under this new minimum benefit was one of the considerations that led me to propose lifting Social Security’s earliest retirement age to 64, instead of 65 as suggested in the Liebman-MacGuineasSamwick plan. Another consideration was the front-loaded nature of the benefit cuts under Old-Age Risk-Sharing, which would reduce the incentives for retiring too early and help tilt Social Security’s incentives in favor of delayed retirement. Because these enhanced protections and more constructive incentives are built into A Well-Tailored Safety Net that prescribes no cuts to Social Security’s basic benefit formula in very old age, I judged that 64 was an appropriate age for balancing the competing needs of income support early in retirement and income security late in life. Still, A Well-Tailored Safety Net recognizes that workers may need a cushion of income support even before age 64 and that the financial interest of near-retirees— as well as the nation—would be better served if they can remain in the work force for additional years. Here is where the helping hand comes in. Under today’s Social Security system, many workers face a difficult trade-off: Claiming income support in the earliest years of eligibility can mean sacrificing a significant degree of income security in very old age, yet many opt to prioritize their more immediate interest.2 But A Well-Tailored Safety Net would strike a balance between these two competing interests rather than forcing workers to choose one or the other. Instead of offering early retirement at 62, Social Security would begin to offer a substantial and growing level of income support, without unwinding the safety net these workers will depend on in very old age. In other words, the earliest eligibility age will stay the same, even as the earliest retirement age increases, but workers who begin receiving benefits at 62 wouldn’t face steeper early retirement penalties than those who claim their benefit at 64. Any change in the timing of benefits will come gradually, so those who are nearing retirement will only have to make a modest adjustment. As seen in table 17.1, workers turning 62 in 2022 would still be able to receive 75 percent of their benefit at the
A HELPING HAND, NOT AN EMPTY PROMISE
127
Table 17.1 Benefit eligibility for workers turning 62 in . . . Age
2017
2022
2028 and later
62
96%
75%
50%
63
97%
79%
58%
64
97%
83%
67%
65
98%
88%
75%
66
98%
92%
83%
67
99%
96%
92%
68
100%
100%
100%
Maximum early-retirement penalty
25%
25%
25%
Earliest retirement age
62.2
63
64
Note: Early-retirement penalty reflects increase in the Normal Retirement Age to 68. The earliest retirement age is the age at which workers could claim their full benefit if they don’t receive any benefits beforehand.
earliest eligibility age. And even after full change is implemented by 2028, workers could still get 50 percent of their benefit at 62, with the other half ramping up in equal increments through 68. Rather than pulling the floor of support away from workers late in their careers, Social Security’s role in the initial years of eligibility would shift to income support from income replacement. As benefit eligibility ramps up, workers would have the wherewithal to reduce the number of hours they work. Alternatively, workers would be able to claim their full benefit at the new Earliest Retirement Age detailed in table 17.1, if they don’t receive any benefits beforehand. And A Well-Tailored Safety Net would also offer workers a third option of receiving 50 percent of their benefit for as long as they like. This will increase workers’ flexibility to delay getting their full benefit and enhance their income security in old age by further limiting penalties for early retirement. A second way in which Social Security would extend a helping hand to enable workers to cope with scaled-back benefit eligibility in their early 60s is tied to the progressive saving offset introduced in chapter 16. Recall that those earning at least 150 percent of the average wage ($63,000 in 2009) would face an additional Social Security payroll tax of 1.5 percent, but those earning less than 150 percent of the average wage would face proportionally smaller tax increases. An average earner ($42,000) would face an extra 1 percent tax and those earning up to 50 percent of the average wage would face a 0.5 percent tax. However, partly for the sake of simplicity and efficiency, an additional 1.5 percent of wages will be collected from everyone, with any extra amount deposited in a 401(k)-type individual account, as seen in table 17.2. While modest earners would be given a strong push to save more, they wouldn’t be forced to do so and would be able to access the funds if needed before retirement. But, ideally, this extra saving opportunity afforded to average and below-average
128
A WELL-TAILORED SAFETY NET
Table 17.2 Saving to cope with scaled-back early eligibility Income level (2009 wages)
Extra % of wages collected
% of wages deposited in personal account
% of wages available for supplemental savings
Up to $21,000
1.5%
0.5%
1.0%
$42,000
1.5%
1.0%
0.5%
$63,000 and up
1.5%
1.5%
0%
earners under A Well-Tailored Safety Net would provide these workers with a source of income support after they turn 62 and face scaled-back benefit eligibility. After a full career of deposits, this account could provide the equivalent of 1.5 years’ worth of early-retirement benefits to someone earning 50 percent of the average wage ($21,000 in 2009). A third way to help ensure that those of modest means have the financial resources to cope with an increase in the earliest retirement age is to improve their prospects for staying in the workforce. This would also serve the nation’s economic interest, because the economy will be better able to produce the resources to meet the challenges of an aging population if we make better use of the productive resources of older workers. One logical way of trying to achieve these twin goals is to give employers an incentive to provide flexible or low-intensity opportunities for older workers. This could be achieved by eliminating the 6.2 percent payroll tax paid by employers on the first $10,000 in wages paid to workers 62 and older, reducing tax bills by $620 per older employee. As seen in table 17.3, this approach would target the tax cuts to relatively moderate-skilled jobs where this incentive is more likely to make a difference, rather than jobs paying above-average wages. Further, this approach would give employers an incentive to provide work opportunities with flexible hours for older workers who are winding down their careers. For example, an employer paying one fulltime older worker $30,000 would get a $620 tax cut. But an employer paying two half-time workers a total of $30,000 would get a $1,240 tax cut. Table 17.3 Employer Social Security payroll tax rates for workers 62 and up in 2022 Income level (2009 wages)
Employer payroll taxes on first $10,000 in wages
Employer payroll taxes on wages above $10,000
$10,000
$0
$0
0%
$20,000
$0
$620
3.1%
$30,000
$0
$1,240
4.1%
$100,000
$0
$5,580
5.6%
Effective payroll tax rate
A HELPING HAND, NOT AN EMPTY PROMISE
129
Payroll tax cuts structured in this way will be particularly helpful for workers who have had physically demanding jobs. These tax cuts will help extend their careers by providing their employers an incentive to transition them to lower-intensity work or at least to give them the flexible hours that will allow them to wind down their careers with dignity. And if neither of those is a workable option, the employer tax cuts would give them a leg up in finding alternative work. As we’ve seen, even proposals that call for major tax increases would still require sizeable benefit cuts on top of severe early-retirement penalties. Instead of being content to see workers retire too early and wind up scraping by on an insufficient benefit check, the compassionate thing to do is to help them to extend their careers, while still providing a backstop of support. ‘‘The choice of retirement age is the most important portfolio choice most workers will make—far exceeding in importance such issues as whether to invest their 401 (k)s in stocks or bonds,’’ conclude scholars at the Urban Institute. And lower earners, for whom Social Security provides the bulk of retirement income, have the most to gain by remaining in the workforce, they note.3
Chapter 18
A Well-Tailored Account Structure
When I completed the original design of A Well-Tailored Safety Net in 2006, the account structure I proposed looked almost exactly the same as the one I’m proposing now, with one important difference: It would have carved out a portion of Social Security’s existing revenue stream. To briefly explain, my original approach bore little resemblance to the account proposals that have driven Democrats to distraction, and only a modest resemblance to the Liebman-MacGuineas-Samwick (LMS) approach, which—although far more constructive than its predecessors—still changes Social Security in a way that the party of FDR appears unlikely to accept. My approach would have required zero borrowing—both because the carve-out was relatively small and because a tax on income over the payroll-tax ceiling would have been applied several years before the accounts got off the ground. Further, because account size would be proportional to income, the carve-out portion would have represented less than 5 percent of a low earner’s benefit and 10 percent for an average earner. Risk would have been further reduced by conservative account options that carried crystal clear labels: ‘‘no risk,’’ ‘‘low risk’’ (up to 33 percent in stock index funds) and ‘‘higher risk’’ (up to 50 percent in stock index funds). Finally, there would have been no chance that poor returns would diminish retirement income in very old age because the accounts would only cover a portion of benefits through life expectancy, at which time Social Security’s full defined benefit would have kicked in.1 I provide this background about a feature of A Well-Tailored Safety Net that I have now cast aside for several reasons. First, although the debate over personal accounts is often portrayed in starkly ideological terms, I offer this window into my thinking, in part, to demonstrate that I have approached this issue from a non-ideological perspective. More to the point, my shift in thinking on the merits of carving out a portion of Social Security’s existing revenue to finance personal accounts raises three questions
132
A WELL-TAILORED SAFETY NET
that are central to this chapter: 1) Why did I believe that carve-out accounts were appropriate in A Well-Tailored Safety Net? 2) Why do I now find the case for carveout accounts less compelling? 3) And, most importantly, how can we take advantage of the most constructive and progressive features of personal accounts while leaving aside their more controversial and questionable aspects? First, let’s talk about what has changed since 2006. One obvious difference is that the current White House occupant is much less likely to be open to a proposal that carves revenues out of Social Security. That’s an important consideration since the cost of not reforming Social Security is much higher than the cost of doing so without a carve-out. However, President Barack Obama’s election wasn’t the primary reason I saw fit to alter my approach. Nor was another obvious development—the onset of a brutal bear market— central to my reconsideration of the merits of carve-out accounts. Rather than being a reason to shun equities in the future, the consequences of dangerous risk-taking throughout the financial sector has spurred reform efforts that may provide a more stable foundation for investing. However, the fact that major market indexes have now made no headway since the late 1990s helps to explain why one of the rationales offered by proponents of such accounts—the opportunity for favorable investment returns—is not, by itself, sufficient to warrant a carve-out of Social Security resources. The main development that reshaped my thinking was the rapid deterioration of the fiscal outlook that struck at another central argument for personal accounts— the fiscal responsibility or national savings rationale. As detailed at some length in the first two chapters, Washington has shown itself over the past quarter-century of Social Security surpluses to be incapable of saving for the future. Rather, it’s fair to conclude that elected officials have felt somewhat less restrained in their spending because the surplus revenue coming from Social Security has helped make annual budget deficits appear relatively more manageable. Thus, instead of seizing the opportunity to set aside real resources that would make Social Security more affordable in coming decades while expanding the pool of savings now available to invest in the productive capacity of the economy, our political leaders have done just the opposite. It is this missed opportunity that has been one of the most compelling arguments for creating personal accounts to, in effect, wall off a portion of the extra revenues flowing to Social Security. Without a reliable way of setting aside real resources— and a risk that Congress would fritter away yet another opportunity for saving— there was reason to question how much could be gained by generating even bigger near-term Social Security surpluses through applying a tax on income over the payroll-tax ceiling and beginning to scale back benefit promises. But in the intervening years since 2006, the link between carve-out personal accounts and the goal of increasing national savings has grown much more tenuous. Because the severe recession and rise in joblessness have taken a bite out of Social Security’s revenue stream, the substantial Social Security surpluses that had been expected through 2016 now amount to little more than a rounding error.2 What is
A WELL-TAILORED ACCOUNT STRUCTURE
133
more, because large-scale budget deficits are projected through the end of the decade, with no let up in sight, if any modest surpluses do flow to Social Security, they would hardly distort the true fiscal picture. Now, with the era of large and growing Social Security deficits almost upon us, there’s no longer a risk that the rest of the budget will grow fat on any extra savings and revenue produced by reforming Social Security. Rather, as seen in the example in table 18.1, carve-out accounts would now come exclusively at the cost of bigger Social Security deficits, at least in the intermediate term. Unless the rest of the budget—which faces much greater stresses—shrinks to accommodate bigger Social Security deficits, then carve-out accounts would have no impact on national savings, either positive or negative. Thus, they are now largely a question of when we will pay Social Security’s bills—either as they come due, or decades later as personal accounts balances begin to displace some of the need for government benefit checks. Now, at the very least, it’s reasonable to presume that there is no particular advantage to be had by paying our bills later. That means the case for carve-out accounts has to be strong enough on its own without any regard to the goal of increasing national savings. I should note, however, that storing up resources in so-called add-on accounts financed by an increase in the payroll-tax rate also could come at the expense of less deficit reduction in the near term. That’s because under such add-on accounts, even though the taxes are levied up front, they aren’t effectively paid until retirement in the form of lower traditional benefits that would be replaced with proceeds from the accounts. Therefore, the case for both add-ons and carve-outs rests on the third central argument for personal accounts within Social Security—what I’ll call the progressive ownership rationale. This, in my view, is the strongest argument for such accounts. The case can be boiled down to two different but related issues: incentivizing more work and providing those with little savings outside of Social Security with the opportunity to leave a modest inheritance. In part, this rationale for personal accounts reflects the logic that workers may be more likely to retire later and achieve better financial outcomes if there is more of a link between contributions and retirement income. To some extent, such a link is obscured by the complexity of Social Security’s formula for determining benefits and by the fact it only counts 35 years of work. Further, workers may be more likely Table 18.1 The deficit impact of carve-out personal accounts in 2025—an example
Social Security financing gap Savings (incl. tax on high earners) Carve-out personal account deposits
($200 billion) $150 billion ($110 billion)
TOTAL — Net Social Security deficit
($160 billion)
134
A WELL-TAILORED SAFETY NET
to claim benefits at the first chance they get—even at the cost of income security in very old age—if they have no opportunity to leave behind an inheritance no matter how long they delay retirement. And at present, Social Security generally restricts that opportunity to a few targeted groups: primarily those who leave behind either a child under 18 or a surviving spouse with lower career earnings.3 The question of inheritance within Social Security came to the forefront in 2005, when President George W. Bush said that the system is unfair to African-American males who, on average, have shorter average life spans.4 Democrats and associated advocacy groups argued reasonably that adding a right of inheritance to Social Security could be a bad deal if it came at the expense of the generous and critical support provided to the minor children and surviving spouses of those who die early.5 But in arguing against plans to carve out a chunk of Social Security’s revenues for personal accounts, critics of the Bush plan also appeared to argue against a broader right of inheritance for the low-income workers they were devoted to serving. Their argument was that African-Americans make out OK as a group because they are more likely to receive disability benefits before the retirement age, and their children are more likely to receive survivor benefits.6 But while it may be important to think about Social Security from a group perspective, it isn’t sufficient. We also need to think about it in terms of individuals and families, and for too many families with modest incomes who suffer a tragic loss of life, Social Security is a bad deal. Despite all the good that Social Security does—and the protections it provides for the elderly, disabled and children are critical—there is an element of injustice in this. If we reform Social Security without providing workers any right of ownership, this injustice is likely to be exacerbated in two ways. First, if we raise the payrolltax rate beyond 12.4 percent without turning Social Security into more of a true savings program, then modest earners would have an even harder time accumulating the financial assets that will provide them both pride of ownership and peace of mind. Secondly, workers trying to overcome benefit cuts by extending their careers could stay on the job into their late 60s and die without being able to leave any of their Social Security savings to their loved ones. That should be unacceptable both because it is unjust and because it runs counter to Social Security’s goal of providing a secure retirement. Without any opportunity for ownership, workers with modest incomes may be more likely to conclude that delaying retirement isn’t worth the risk of foregone benefits and opt to settle for a lower benefit that may turn out to be inadequate as they grow older. This progressive ownership rationale, in my view, is clearly strong enough to merit that any additional payroll-tax contributions on top of the 12.4 percent rate be deposited in personal accounts. But the unprecedented surge in federal debt resulting from government efforts to stem the financial and economic crises led me to examine whether my original stance in favor of a modest carve-out of Social Security revenue was still optimal. In other words, did this progressive ownership rationale trump the goal of prompt deficit reduction? And in looking at that question, it hit upon me that there is a way to retain much of the progressive benefit of a carve-out without
A WELL-TAILORED ACCOUNT STRUCTURE
135
actually having to do the carve-out. While it is not a perfect solution, I judged that, given budget and political realities, it was good enough to dispense with a carve-out. The solution I settled on is really quite simple. Instead of taking resources away from Social Security and depositing them in personal accounts, the accounts would be credited with a claim on Social Security’s future resources as represented by the government bonds in the trust fund. Because these bonds would remain within Social Security, where they would accrue interest, this approach would avoid the need for any up-front borrowing to finance account deposits. This type of account structure, which bears a resemblance to the one adopted as part of Sweden’s public pension reform, could be considered a notional account because it tracks the implied value of an account that does not directly hold any assets. However, in the context of other important reforms of Social Security, this account would be much more than a notion. Most obviously, those who die before claiming Social Security benefits would have a claim on future trust fund assets that is just as valuable—and inheritable—as real Treasury bonds. Thus, apart from any superior returns that might be attainable from investing in stocks, a carve-out account would have no built-in advantage over this alternative structure in adding a right of inheritance to Social Security. The other progressive ownership rationale for carve-out accounts is to encourage delayed retirement, and this alternative approach of crediting the accounts with a claim on trust fund bonds is also well-suited to achieve that goal for two reasons. First, this potential inheritance would remove any risk that workers might delay retirement only to wind up forfeiting all of their benefits. Secondly, the accounts would be linked to a more robust incentive structure for promoting delayed retirement that includes lump-sum transfers out of these more-than-notional accounts, in addition to bigger benefit checks for life. Thus, the most useful way to describe these accounts may be as Delayed Retirement Accounts. In effect, workers retiring early would give up this opportunity for increased ownership because their Delayed Retirement Account balances would simply be cashed in for the traditional Social Security annuity. Before detailing this new approach to accounts, I should note that a number of serious Social Security reform advocates have proposed a different alternative to a carve-out—a simple add-on account financed by an additional 2 percent, or so, of wages that would be required to offset prescribed benefit cuts.7 While such accounts could go a long way in providing ownership within Social Security and might produce an important increase in national savings, I opted against this kind of approach primarily out of concern that such a tax increase ($400 for a $20,000-earner) would cause hardship for those who have little room in their budgets. By contrast, this proposal, tied to the progressive saving offset introduced in chapter 16, would limit the tax hike on lower earners to just 0.5 percent of wages ($100 for a $20,000-earner). Because A Well-Tailored Safety Net prescribes sacrifice that is roughly proportional to income, a simple add-on structure is insufficient to provide a meaningful ownership opportunity.
136
A WELL-TAILORED SAFETY NET
It also is important to understand that while add-on accounts of significant scale do not arouse the same intensity of opposition as carve-outs, largely because they devote additional resources to retirement income, they do raise some similar questions regarding how tamper-proof the safety net would be.8 The possibility that low-income workers would have significant resources built up in accounts that, at least in theory, provide some degree of ownership raises the question of whether any law that denies them access to those funds will be politically sustainable. If, for example, a requirement that individuals cash in their full accounts for lifetime annuities upon retirement were eased, individuals would be at risk of depleting their account funds, leaving them to scrape by in very old age. As I will explain, of all the account proposals introduced to date, A Well-Tailored Safety Net is the only one that effectively addresses these concerns that giving individuals control of their Social Security account assets would undermine income security in old age. First, because account size is directly related to income, accounts would make up a very small percentage of the overall benefit for the modest earners most dependent upon Social Security. In addition, low- and average-wage earners would be allowed to tap a portion of their account savings before retirement—the voluntary savings portion. And upon retirement, individuals would retain ownership of their account balances, giving them the opportunity to leave a modest inheritance. Yet this ownership wouldn’t reduce their Social Security income in old age because benefit cuts phase out at life expectancy. Finally, workers who delay retirement are rewarded with greater ownership that they would enjoy in addition to bigger Social Security checks for as long as they live. In other words, ownership and income security can go hand in hand. Now let’s first look at the full account structure, including the additional 1.5 percent of wages collected related to the progressive saving offset and the Delayed Retirement Account portion equal to another 1.5 percent of wages. As table 18.2 shows, all workers would see account deposits of 3 percent of wages each year, though a portion of that 3 percent would represent savings outside of Social Security for average and below-average earners. One consequence of this account structure that doesn’t include a real carve-out and provides for moderate earners to save extra outside of Social Security is that the amount of resources devoted to accounts within Social Security would be small relative to other account plans. This raises a question of whether it would be efficient to Table 18.2 A well-tailored account structure Income level (2009 wages)
Add-on account deposit as % of income
Treasuries earmarked to Delayed Retirement Account as % of income
Supplemental savings as % of income
Total account size as % of income
$21,000
0.5%
1.5%
1%
3%
$42,000
1%
1.5%
0.5%
3%
$63,000
1.5%
1.5%
0%
3%
A WELL-TAILORED ACCOUNT STRUCTURE
137
provide an equity investment option for these add-on accounts. In other words, could any management and transaction costs be shared widely enough so that account fees would be minimal? I concluded that the answer is likely not, which leads to my recommendation that the add-on accounts be limited to government bonds for the foreseeable future. Those who think any degree of investment risk is inconsistent with Social Security’s mission will see that as the right decision; however, it comes with a trade-off. One of the advantages of devoting more resources to investment within Social Security is that it could provide a low-fee investment platform for modest earners saving additional funds outside of Social Security. Without such an efficient account structure, it logically follows that this voluntary saving also would be limited to Treasuries.9 Thus, this solution is not a perfect one, but given our fiscal situation, it may be the best one available. Now let’s look at potential accumulations over a 40-year career for a worker earning 50 percent of the average wage, or $21,000 in 2009. As seen in table 18.3, including the extra saving opportunity outside of Social Security, such a worker could accumulate about $49,000—the equivalent of about 3 years of benefits—that they could pass on to family members if they die before claiming Social Security. To be clear, this right of inheritance would complement—not take away from— Social Security’s survivor benefits. For example, because the add-on account portion is designed to compensate for reduced benefits, it would be transferred to a surviving spouse’s account so as not to reduce future retirement income. However, the inheritance related to Delayed Retirement Accounts would represent a commitment of new resources, and that does raise a question of cost. One answer is that the cost of failing to provide such a modest degree of ownership within Social Security would be even greater in terms of reduced fairness and work incentives, in addition to the impact on families and communities where life expectancy is lower. For perspective, this inheritance would reflect just 12 percent of a worker’s contributions into Social Security. Further, because the vast majority of workers will live long enough to claim benefits—89 percent of workers born in 1985 are expected to survive to age 65 in 205010—and lower earners are more likely to benefit from this inheritance, the cost would not be extravagant. I estimate that providing an inheritance for those whose families would not already receive generous survivor benefits would increase Social Security’s projected annual cost by less than 1 percent even after 40 years of deposits. Now let’s return to the important question of how these accounts will work for those who don’t die before claiming benefits. The key design consideration is to
Table 18.3 Account accumulations for low earner with 40 years of work, 2009 dollars Add-on account Delayed Retirement Account
$8,165
$24,495
Supplemental savings account
Total
$16,330
$48,990
Note: Account totals assume Treasury-level investment returns; supplemental savings is voluntary. Source: Author’s calculations
138
A WELL-TAILORED SAFETY NET
make sure that the ownership opportunity does not come at the cost of income security, but rather contributes to it. First, let’s look at the add-on portion of the account. After retirement, workers would maintain a degree of ownership over these account balances, which would be drawn down in equal inflation-adjusted amounts through life expectancy to help workers overcome the benefit cuts they will face under the progressive saving offset. For example, after a full career, 10 percent of an average earner’s Social Security income would come from the add-on account, fully compensating for the reduction tied to saving, and 90 percent would come from the traditional benefit. While the accounts would be depleted as a retiree reaches the average life expectancy, the 10 percent benefit cut would expire at that point and Social Security’s full old-age safety net would kick in. The account of an average earner who works more than 40 years would more than offset the 10 percent reduction, providing an incentive to keep working. Further, this control of the accounts post-retirement would, to some extent, limit the risk that a worker who delays retirement would get less from Social Security if they should die in their 70s. Implicit in the approach to personal accounts laid out in this chapter is a judgment that Social Security’s current incentive structure is not well designed to encourage longer careers. The evidence seems to support such a conclusion: In 2007, only about 4 percent of workers opted to claim retirement benefits after the Normal Retirement Age, now 66, but soon to rise to 67.11 Of course, there are many reasons that workers may retire early that have nothing to do with the design of Social Security, and the extent to which different incentives could influence retirement decisions is unknown. However, because longer careers can have a positive impact on retirement security, government finances and the economy, it certainly makes sense to try. One approach that has been offered would be to raise Social Security’s delayed retirement credit from 8 to 10 percent, meaning a worker who passed up benefits for a year would get a check that is 10 percent bigger the next year.12 This would definitely provide more generosity to those who would have delayed retirement anyway, and it might well have some impact on others’ retirement decisions. But, in my view, this approach doesn’t go far enough to address the central trade-off inherent in giving up a year of benefits: doing so would only pay off for those who live long enough to reap the rewards. As currently structured, those who extend their careers but die before the average life expectancy (now 83) could end up getting less of a good deal from Social Security than if they had retired at the Normal Retirement Age or earlier, though a surviving spouse with lower career earnings could benefit from the delayed retirement bonus. Yet, the fact that about half of workers claim benefits at 62 and only 4 percent after 66 suggests that near-term gratification is driving retirement decisions more than concern about income security late in life. Thus, while increased incentives may help, there is also logic in making delayed retirement incentives more reflective of both risks that workers have to balance in deciding when to retire: The risk they will live
A WELL-TAILORED ACCOUNT STRUCTURE
139
to a very old age and the risk that they won’t. That is why, under A Well-Tailored Safety Net, the rewards for delaying retirement would be increased and half would be paid immediately—out of a worker’s Delayed Retirement Account. Recall that in connection with the Old-Age Risk-Sharing provision, the delayed retirement credit would gradually rise to 10 percent of a worker’s base benefit, or what the Social Security Administration calls the Primary Insurance Amount. After mid-century, once benefit changes are fully phased in, an average earner would face an upfront 20 percent benefit reduction, which could be overcome with two years of delayed retirement credits. Half of this new 10 percent credit would still serve to bolster one’s Social Security benefit and provide an added degree of insurance against the risk retirees will live to a very old age and deplete their savings. The other half would be awarded as a lumpsum transfer that reflects the present value of the benefits that would be paid through life expectancy.13 Because of the combination of upfront benefit reductions and more generous delayed retirement incentives, the lump-sum transfer paid to average earners for working one year past the Normal Retirement Age—almost $10,000 based on today’s wage levels—would equal almost 90 percent of what they would have been paid in traditional benefits if they had simply retired. On top of that, the Social Security checks they will receive every month for the rest of their lives will be considerably bigger. If this, at first, sounds like too much of a good deal to be affordable, that is because further explanation is required. The first thing to understand is that there is a tradeoff in balancing the rewards of delayed retirement between upfront payments and greater income security for those who live to an advanced age. Because half of the 10 percent credit is awarded when it is earned, a retiree’s benefit check would only be 5 percent bigger in very old age, which is less the 8 percent benefit increase under the current structure. While a strong safety net in very old age is clearly one of the most important goals for Social Security reform, there are two reasons why this trade-off makes sense as part of A Well-Tailored Safety Net. First, it’s important to remember than few workers take advantage of the present system’s delayed retirement awards. Therefore, if redesigned incentives lead more people to extend their careers and earn more credits than they would have under today’s system, then the safety net in very old age would be firmer—even if the long-term credit were reduced to 5 percent. The other key thing to remember is that this delayed retirement incentive is designed to complement the prescribed changes to Social Security’s benefit structure in which benefit cuts are reduced early in retirement to avoid the need for cuts in very old age. A simple example will help illustrate this point. Consider an alternative approach to Social Security reform that cuts benefits even by a very modest 10 percent, but makes those cuts apply to all retirees, whether age 65 or 95. As seen in table 18.4, even if a worker facing such modest cuts were to work four years past the Normal Retirement Age and enjoy the current 8 percent benefit increase, A
140
A WELL-TAILORED SAFETY NET
Table 18.4 Benefit levels at age 92 under two delayed retirement scenarios Years of work past Normal Retirement Age
Benefit as % of base benefit with a 10% reduction and 8% delayed retirement credits
Benefit as a % of base benefit under A Well-Tailored Safety Net with 5% delayed retirement credits
1
97.2%
105%
2
104.4%
110%
3
111.6%
115%
4
118.8%
120%
Well-Tailored Safety Net would provide a comparable level of income support in very old age. A closer look reveals another reason why this more attractive delayed retirement incentive would be not only affordable but also highly progressive. This relates to the progressive saving offset, which asks higher earners to set aside more savings and, in effect, pay a bigger share of their benefits through life expectancy, after which they would receive Social Security’s full support. As seen in table 18.5, the 10percentage-point delayed retirement credit would be reduced prior to life expectancy based on the extent to which add-on savings makes up a portion of a retiree’s benefit. This makes sense when you consider that delaying retirement for a year would mean that the add-on accounts would earn extra interest, and these larger balances could be spread over a fewer number of years through life expectancy, yielding bigger annual payments. The net result is that one year of delayed retirement would yield a 9.55percentage-point benefit increase for a $18,900-earner, but a 6.5-percentage-point increase for the highest earners—that is, until they reach life expectancy, when the progressive saving offset phases out and both income groups would receive similar treatment. Returning to the link between Delayed Retirement Accounts and lump-sum transfers, I should note that while workers would have the option of receiving a Table 18.5 A progressive delayed retirement incentive
a
Income level (2009 wages)
Progressive saving offset as % of base benefit
Delayed retirement credit prior to average life expectancy as % of base benefita
$18,900
4.5%
9.55%
5%
$42,000
10%
9.0%
5%
$67,200
20.5%
7.95%
5%
$106,800
34.5%
6.55%
5%
Benefit increase after life expectancy as % of base benefit
The delayed retirement credit includes a lump-sum transfer equal to the present value of a 5-percentage-point increase in the base benefit through life expectancy.
A WELL-TAILORED ACCOUNT STRUCTURE
141
straightforward payment, the default option would simply be to have these funds transferred to their add-on account, where they would be drawn down through life expectancy or passed on to a survivor. Clearly, there is no good reason to prioritize account withdrawals, but there is every reason to design incentives so they are attuned to widely divergent goals and needs of older workers. If we fail to make the rewards of delayed retirement both more attractive and more immediate, then we will, in effect, to discourage workers from doing everything they can to secure a comfortable retirement and we will deny those with modest incomes a potential path for improving the lives of their children and grandchildren. This raises one final issue in designing well-tailored incentives. To this point, I have only written about lump-sum transfers in connection with work past the Normal Retirement Age. Such a change might raise little concern, in part because it would be accompanied by bigger benefit checks, but also because it would apply only to workers who had stayed on the job much longer than their peers and had presumably acquired a large degree of income security. Nevertheless, I think it is worth considering whether such an option should be extended to those who work past the age of 65. This would permit workers to split their 6.7-percentage-point benefit increase for each year of work between 65 and 68, with half of this increase going to raise benefit levels and half going as an upfront payment into their personal account. The default option should unquestionably be to devote the full increase to a bigger traditional benefit, which would likely be more rewarding in the long run for the majority of older workers. But we should remember that retirement decisions often aren’t made on the basis of long-term financial interest, and there’s a risk that if we delay effective work incentives until past 68, too many workers would see them as being out of reach and still claim benefits as early as possible.
Chapter 19
Measuring Up
There are many considerations in evaluating a Social Security reform plan, but a good place to start in examining whether a proposal meets the most critical objectives is to recall just why the so-called third rail of American politics is back on the agenda. Although Social Security’s finances have taken a moderate hit due to recessionary job losses and the 2009 trustees report moved up the data of trust fund exhaustion until 2037, neither of those developments quite explain the urgency with which some leading Democrats now want to address this challenge. After all, the trust fund still provides the legal authority to make good on all benefit promises for the next quarter-century. The urgency is that budget deficits are of such a scale and federal debt is growing at such a dangerous trajectory that the government is clearly in no position to take on trillions of dollars in unfunded Social Security promises. As House Majority Leader Steny Hoyer put it, Social Security reform ‘‘gives us . . . an immediate chance to take pressure off of the budget.’’1 Although the fiscal impact of a Social Security reform proposal is generally gauged by the extent to which it erases the 75-year financing gap, this benchmark doesn’t give enough weight to the next few decades in which our budget problems are expected to come to a boil. While Social Security’s long-term solvency remains an important goal, policy makers need to be equally focused on reducing the $5.4 trillion debt the nation would accumulate by paying all of Social Security’s promises from 2016, when the program is on track to start running a permanent cash deficit, through 2036, just before the theoretical trust fund is exhausted. This shorter view is critical for evaluating the extent to which a proposal produces savings and helps the intermediate-term budget picture or leaves a big burden of debt to be shouldered by succeeding generations of workers and retirees. First, let’s look at the near-term fiscal impact of Sen. Robert Bennett’s approach, which rejects tax increases and gradually reduces benefits to offset future increases in Social Security outlays related to increasing life expectancy and rising wages. Over
144
A WELL-TAILORED SAFETY NET
time, these benefit cuts would become more severe than those in almost any other plan on the table, but the gradual nature of the cuts means that the budget savings also would be relatively slow in coming. Under the Bennett plan, the nation would still have to accumulate $4.1 trillion in debt to pay all of Social Security’s promises from 2016 through 2036. That’s 75 percent of the $5.4 trillion debt that would pile up with no changes at all.2 Now, let’s look at the Diamond-Orszag plan, which combines an upfront tax increase on the highest earners with broad-based benefit cuts and payroll-tax increases that are gradually phased in over the 75-year period. This proposal, among the most fiscally responsible to date, would reduce the Social Security-related debt the nation will incur through 2036 by about 55 percent to roughly $2.9 trillion.3 It’s important to understand that the additional debt reduction in the DiamondOrszag plan relative to the Bennett plan is due to its tax increase on higher earners. In other words, a decision not to apply such a tax hike on income above the payroll-tax ceiling amounts to a decision to pass on an extra burden of Social Security debt to future generations. Yet a similar analysis also applies to both the Diamond-Orszag and Bennett proposals’ approach of imposing a level of sacrifice that starts modestly but continues to grow deeper throughout the next 75 years. By opting to minimize sacrifice for workers who will retire in the next couple of decades, both proposals choose instead to pass on trillions in Social Security debt and make future workers bear a larger share of the sacrifice. Such a go-slow approach, while perhaps more politically palatable, ignores the shape of Social Security’s financing gap, which is projected to reach 1.3 percent of GDP on an annual basis by 2030 and then proceed to stabilize, more or less, for the next 45 years.4 Under the Bennett plan, instead of a 1.3 percent GDP gap, the annual cost of carrying an extra $4.1 trillion in debt and filling Social Security’s remaining cash shortfall would approach 1.9 percent of GDP in 2037 and continue rising thereafter.5 In other words, the extra interest burden would more than negate any savings from cutting benefits. In evaluating these proposals, I assumed that no changes would be implemented before 2013, allowing time for the economy and job market to recover. Given the expected fragility of recovery, I also think it may be unrealistic to expect any Social Security payroll-tax rate hikes to take effect before 2016, though that view didn’t impact my analysis of the Diamond-Orszag plan. So how can we go about addressing Social Security’s near-term cash shortfalls and avoid passing a big burden of debt to future generations? One constructive option for reducing near-term debts is to hold the growth of tax-exempt spending on employerprovided health care to the rate of GDP growth. In conjunction with the other strands of A Well-Tailored Safety Net, this could reduce the additional $5.4 trillion in Social Security-related debt through 2036 by nearly 80 percent to less than $1.2 trillion. However, as noted at the end of chapter 16, A Well-Tailored Safety Net also offers an alternative approach that doesn’t rely on altering the tax-free status of employer health coverage. Because measuring the impact of such a tax change on
MEASURING UP
145
both individual tax and benefit levels is complicated by the uncertain path of health care costs, the rest of this chapter will focus on Plan B to enable a clear comparison between A Well-Tailored Safety Net and other existing Social Security reform proposals. This alternative design of A Well-Tailored Safety Net involves a slightly higher new tax on wages above the payroll tax ceiling (3 percent versus 2 percent) and a bigger saving offset (i.e. benefit cuts) for above-average earners as detailed in table 16.7. Plan B would still reduce the projected debt increase related to Social Security’s unfunded promises through 2036 by 70 percent to $1.6 trillion—leaving more debt than Plan A with the curb on tax-exempt employer health spending, but significantly less than other approaches on the table. Among the reasons that A Well-Tailored Safety Net stands apart from other proposals in its near-term budget savings is the Old-Age Risk-Sharing approach that frontloads benefit reductions in order to encourage delayed retirement and avoid benefit cuts in very old age. Under this approach, which prescribes proportionally smaller benefit cuts for lower earners, the maximum benefit cut any retiree will face would come in the first year benefits are paid and the full benefit cut would gradually phase out over 20 years. My analysis shows that in 2030, the savings from Old-Age RiskSharing would be more than 40 percent greater than applying a flat benefit cut over the same 20 years. Turning to the Liebman-MacGuineas-Samwick plan’s near-term impact on federal debt levels, I should emphasize that the comparison is not a straightforward one. Even if the LMS plan were adjusted to eliminate borrowing for personal accounts, it would still amass $4 trillion in debt through 2036.6 However, this result is not due to any lack of near-term sacrifice, but rather to the authors’ decision to direct most of the plan’s savings and extra tax revenue in coming decades to personal accounts, not to federal deficit-reduction. My analysis of the impact of these Social Security reform proposals that have been scored by Social Security’s Office of the Chief Actuary should be considered accurate, if not precise, because it’s a fairly straightforward undertaking to update the actuaries’ work based on the latest projections for Social Security. But because the actuarial staff has yet to subject A Well-Tailored Safety Net to its computing programs, my own conclusions with respect to its impact on Social Security’s financial status come with a higher margin of error. Yet one stark contrast with the LMS plan helps explain why accepting the general accuracy of my analysis should not require any great leap of faith on the part of readers, despite that plan’s bigger tax increases. Consider that under the LMS plan, those who earn the maximum subject to Social Security taxes over an entire career and retire at the Normal Retirement Age of 68 would see an annual benefit increase of about 7 percent compared to promised benefit levels.7 By contrast, under A Well-Tailored Safety Net, the same maximum earner would initially face a nearly 50 percent benefit cut, though the reductions would phase out to provide a robust safety net in very old age. Let’s now look at how the plans stack up in closing the 75-year financing gap, which remains an important benchmark for judging whether a proposal would restore stability to Social Security’s finances over the long term. Social Security’s
146
A WELL-TAILORED SAFETY NET
75-year gap between benefits and tax revenues is equal to $7.7 trillion in net present value, meaning we would need $7.7 trillion on hand right now in an interest-bearing account to make up the difference. The Bennett plan, for all its tough medicine, would erase about 65 percent of Social Security’s shortfall, leaving future taxpayers to shoulder the remaining unfunded liability of $2.7 trillion in net present value, in addition to the brunt of the prescribed benefit cuts. Both the Diamond-Orszag and LMS plans would reduce this shortfall by about 80 percent—perhaps a bit more for the LMS plan under certain assumptions—to about $1.5 trillion in present value. By comparison, the primary version of A Well-Tailored Safety Net that curbs the growth of tax-exempt spending on employer-provided health care would fully erase the $7.7 billion shortfall, while Plan B would still reduce it by about 95 percent to $400 billion. While present value can be a useful way of quantifying the size of a long-term budget problem in a way that is reasonably accessible, it is somewhat less useful in understanding the future scope of a problem if it is not addressed in a proactive fashion. In that case, a more useful measure is how much a proposal would increase federal debt as a percentage of GDP at the end of 75 years. Figure 19.1 looks at how these reform proposals stack up on that score, with the Bennett plan, on the high end, raising debt levels by more than 40 percent of GDP and the Plan B version of A Well-Tailored Safety Net, on the low end, raising debt levels by about 8 percent of GDP. For perspective, keep in mind that this debt would come on top of government debt that is already projected to surpass 80 percent of GDP within a decade, and public debt much in excess of 100 percent of GDP could jeopardize the AAA credit rating that helps keep down federal borrowing costs.8 Although the Diamond-Orszag and LMS proposals fall somewhat short of erasing the full cost Social Security’s unfunded promises, they both would generate fairly sizable annual surpluses at the end of the 75-year period. Rather than being intentional, the surpluses reflect recent changes in the Social Security actuaries’ assumptions since the proposals were introduced that have moderately narrowed the scale of the program’s annual deficits in the latter half of the century. The above analysis of the proposals’ long-term fiscal impact includes such surpluses, and any changes to limit sacrifice would come at the cost of less debt reduction. However, in the analysis to follow I assume that surpluses emerge starting around 2060 in the Diamond-Orszag plan, and that incremental benefit cuts and tax hikes can be frozen by 2065. On the other hand, because the LMS plan surpluses don’t emerge until after its reductions are phased in and because the authors’ intent was to build up savings within Social Security, I leave its provisions unaltered. Keeping in mind that these figures are, to some degree, a moving target, let’s now turn to a third essential criteria for evaluating the impact of a Social Security reform plan: the additional tax burden it imposes. The higher the tax burden, the less wherewithal workers would have to save outside of Social Security, and the less flexibility the government would have to devote additional resources to the rest of the budget, including health care, where the fiscal challenges are much more daunting. But the level of tax increases needs to be viewed in the context of a plan’s impact on both
MEASURING UP
147
Figure 19.1 Increase in federal debt under various reform plans
federal debt levels and the strength of the safety net. That’s why the absence of any tax hikes in the Bennett plan, which comes at the expense of both debt reduction and retirement security, is more of a weakness than a strength. A look at the proposals that do include tax increases in an effort to maintain an effective safety net shows that the Diamond-Orszag approach of slowly ramping up tax increases as the bills come due, rather than setting aside resources in advance, could eventually result in a bigger tax burden. Taking into account both the tax hike on ordinary workers and the tax increase on higher earners with work income above the payroll-tax ceiling, the overall tax increase imposed by the Diamond-Orszag plan at the end of the 75-year period is roughly equivalent to a 3.1 percent payroll-tax hike. By comparison, the LMS proposal calls for the equivalent of a 2.6 percent payroll-tax hike, while A Well-Tailored Safety Net requires the equivalent of a 1.7 percent payroll-tax hike.9
148
A WELL-TAILORED SAFETY NET
The varying dependence on tax increases among the plans is most visible when it comes to lower earners with the least capacity to find extra room in their budgets. Under the Diamond-Orszag plan, although tax increases would begin modestly, low earners would face an extra 2 percent tax on wages at the end of the 75-year period. By comparison, the LMS plan imposes an additional 1.5 percent tax on wages, while a low earner would face a much more modest 0.5 percent payroll tax increase under A Well-Tailored Safety Net, providing a greater opportunity to save outside of Social Security. Based on these three key measures we’ve looked at, the fiscal benefits of A WellTailored Safety Net are apparent: Less debt in both the near term and the long term; and a much more limited need for tax increases. But a Social Security reform proposal can’t be judged only on the bottom line. It’s not enough to offer an affordable safety net; the safety net also must be effective. Now let’s turn to the critical issue of benefit adequacy. After all, savings outside of Social Security could be depleted if a retiree lives long enough, but Social Security provides an income stream that retirees can count on no matter how long they live. A key test is how a proposal treats average earners ($42,000 in 2009) toward the end of the 75-year period, after all benefit cuts are phased in. While it’s not surprising that A Well-Tailored Safety Net which raises taxes would provide a stronger safety net than the tax-free Bennett plan, table 19.1 highlights one of the costs of depriving Social Security of additional resources. Even though it would do much less to curb Social Security’s unfunded promises, it would only be able to afford a modest level of income support for retirees, whether age 65 or 95. Both the Diamond-Orszag and LMS plans, which both raise substantially more revenue than A Well-Tailored Safety Net, provide a more useful—and tougher—comparison. Table 19.2 suggests that average earners could initially fare better under the Table 19.1 A comparison of benefits for average earner in 2075 % of base benefit to average earner retiring at 65
Benefit payment based on 2009 wage level of $42,000
Age
Bennett proposal
Well-Tailored Safety Net
Bennett proposal
Well-Tailored Safety Net
65
53.4%
64.3%
$9,190
$11,070
70
53.4%
68.3%
$9,190
$11,760
75
53.4%
72.3%
$9,190
$12,450
80
53.4%
76.3%
$9,190
$13,140
85
53.4%
80.3%
$9,190
$13,830
90
53.4%
80%
$9,190
$13,780
95
53.4%
80%
$9,190
$13,780
Note: Benefit levels reflect no suspension of progressive price indexing before 2072. Source: Adapted from Social Security Administration’s February 12, 2009, analysis of Sen. Bennett’s ‘‘Social Security Solvency Act of 2009’’ to reflect 2009 Trustees Report.
MEASURING UP
149
Diamond-Orszag and LMS plans. But those proposals, which offer a benefit that is the same at age 65 as it is at 95, would do somewhat less for retirees who live long enough to deplete their savings and come to depend even more heavily on Social Security. A Well-Tailored Safety Net would provide a bigger benefit to a retiree with average earnings than the Diamond-Orszag proposal starting at age 78. And by age 85, an average retiree would get a benefit that is 8.5 percent bigger under A WellTailored Safety Net—equivalent to an extra $1,090 more each year (in 2009 dollars). Relative to the LMS proposal, A Well-Tailored Safety Net would provide a bigger benefit starting at age 80, and an extra 6 percent, or about $800 a year, starting at age 85. One might expect that a Social Security reform plan that imposes bigger tax increases and accumulates more debt would provide a more effective safety net, but table 19.2 suggests that is not necessarily the case. While A Well-Tailored Safety Net provides less support to average earners in the early years of benefit eligibility, it offers a somewhat more robust safety net in very old age, when Social Security’s support is most critical. It’s also worth considering that A Well-Tailored Safety Net could lead to better personal financial outcomes, particularly compared to the Diamond-Orszag plan, because workers would be able to save more outside of Social Security; they would have less incentive to retire too early; and in part because of the complementary reforms introduced in chapter 17 to help workers retire with maximum support, they’d have greater incentive, flexibility and opportunity to extend their careers. One further point: Table 19.2 overstates the relative income support provided by Diamond-Orszag because it doesn’t reflect the fact that workers could claim their full benefit at age 62 and face much steeper early retirement penalties. Looking at the big
Table 19.2 A further comparison of benefits for average earner in 2075 % of base benefit to average earner retiring at 65
Benefit payment based on 2009 wage level of $42,000
Age
DiamondOrszag
LMS
Well-Tailored Safety Net
DiamondOrszag
LMS
Well-Tailored Safety Net
65
73.7%
75.5%
64.3%
$12,690
$13,000
$11,070
70
73.7%
75.5%
68.3%
$12,690
$13,000
$11,760
75
73.7%
75.5%
72.3%
$12,690
$13,000
$12,450
80
73.7%
75.5%
76.3%
$12,690
$13,000
$13,140
85
73.7%
75.5%
80.3%
$12,690
$13,000
$13,830
90
73.7%
75.5%
80%
$12,690
$13,000
$13,780
95
73.7%
75.5%
80%
$12,690
$13,000
$13,780
Source: Benefit levels reflect Congressional Budget Office, ‘‘Long-Term Analysis of the Diamond-Orszag Social Security Plan,’’ December 2004 and Social Security Administration, Office of the Chief Actuary, ‘‘Estimated Financial Effects of ‘A Nonpartisan Approach to Reforming Social Security - A Proposal Developed by Jeffrey Liebman, Maya MacGuineas and Andrew Samwick,’’’ November 17, 2005, http://www.ssa.gov/OACT/solvency/Liebman_20051117.pdf.
150
A WELL-TAILORED SAFETY NET
picture, there are three critical national priorities that Social Security reform can help to advance: preserving an effective old-age safety net as life expectancy increases; avoiding a debt spiral without onerous tax increases; and encouraging delayed retirement to help counter the demographic pressures that will slow the economy and add to budget problems. The above analysis shows that A Well-Tailored Safety Net is uniquely well designed to address the full scope of these challenges. Another consideration to weigh in evaluating benefit adequacy is a Social Security reform plan’s treatment of disabled beneficiaries. Policy in this area reflects a tension between Sen. Bennett’s view of a ‘‘broad consensus that the disabled should be held harmless’’10 and Diamond and Orszag’s concern about creating ‘‘even stronger incentives for workers to apply for disability rather than retirement benefits.’’11 The latter issue arises because, for example, a worker eligible for disability benefits at 62 wouldn’t face any early retirement reductions, while a 62-year-old claiming retirement benefits would face a 30 percent cut in annual benefits. Because the Normal Retirement Age doesn’t apply to the disabled, further increases beyond 67 would contribute to a wider disparity between disability and retirement benefits. But there is an important distinction between Social Security’s disability benefits— which only apply before the Normal Retirement Age—and the broader category benefits for the disabled that also include benefits after the NRA. As noted in chapter six, while Bennett’s approach would fully protect pre-retirement-age disability benefits, his proposal does not hold the disabled completely harmless—nor, for that matter, does A Well-Tailored Safety Net nor any of the other proposals reviewed in this book. Yet, there are different degrees of harm, and the Bennett plan’s treatment of postretirement age benefits for the disabled provides one of the strongest arguments for bringing in new revenue to Social Security. Both the Bennett and LMS plans would protect pre-retirement-age benefits for the disabled and then limit benefit cuts after the retirement age based on the proportion of years a worker was disabled between 22 and 62, meaning a worker disabled at 62 would feel the full force of the benefit reductions. Yet there is a big difference between the Bennett and LMS plans, because in the latter approach, workers would build up funds in personal accounts that would offset much of the prescribed benefit cuts. As a result, while workers disabled late in their careers could face a 38 percent benefit cut upon reaching the Normal Retirement Age under the Bennett plan, they would only face a maximum 9 percent reduction under the LMS plan, assuming at least Treasury-level investment returns. As also seen in table 19.3, the Ball and Diamond-Orszag plans take opposite approaches to benefits for the disabled. In the former, the disabled would initially receive full benefits, but reductions would accumulate over time as the Ball plan curbs cost-of-living increases. Under Diamond-Orszag, the disabled would initially face the full benefit cut prescribed for non-disabled workers, but the reductions would phase out over time and then benefits would continue to grow as the disabled are awarded a more generous cost-of-living-increase, referred to as a super-COLA.12 Diamond-Orszag would provide extra help to those disabled at an early age—the same group that would be hurt the most under the Ball plan.
MEASURING UP
151
Table 19.3 Benefits to an average earner disabled at age 62 in 2075—a comparison Age
Bennett
LMS
Well-Tailored Safety Net
DiamondOrszag
Ball
62
100%
100%
91.1%
85%
100%
65
100%
100%
92.1%
87.3%
99.1%
70
61.4%
91.0%
93.8%
91.3%
97.6%
75
61.4%
91.0%
95.6%
95.5%
96.2%
80
61.4%
91.0%
97.3%
99.9%
94.7%
85
61.4%
91.0%
99%
104.5%
93.3%
90
61.4%
91.0%
100%
109.2%
91.9%
Average benefit cut, age 62-90
31.9%
7.1%
4.1%
3.4%
4.1%
Note: Disabled workers who work fewer years than is typical from 22 to 62 would accumulate somewhat less in their personal accounts. This could result in slightly lower benefits under A Well-Tailored Safety Net than suggested above, and moderately lower benefits under the LMS plan. Source: Benefit levels reflect treatment of disability benefits under the various proposals as defined by the Social Security Administration’s Office of the Chief Actuary. Benefit reductions reflect Congressional Budget Office analysis of the Diamond-Orszag plan and Social Security Administration analyses for the other evaluated proposals
A Well-Tailored Safety Net adopts a combination of approaches that resemble both the LMS and Diamond-Orszag approach to disability benefits. The progressive saving offset tied to the add-on personal account is reduced proportionally based on years of disability. Under Old-Age Risk-Sharing, those disabled before 65 would face initial benefit reductions equal to less than 50 percent of those facing the nondisabled, which seems only fair because they have no capacity to respond to changed incentives. Benefit cuts would be further lowered based on age of disability and then gradually phase out over time. While there is no optimal way of cutting disability benefits, table 19.3 shows that A Well-Tailored Safety Net provides a level of support to the disabled that is roughly comparable with the higher-tax Ball and DiamondOrszag plans from the left side of the political spectrum. While Social Security reform proposals must be evaluated in comparison to other realistic options—rather than against the option of doing nothing—an internal comparison also can provide a clearer view of the design of A Well-Tailored Safety Net. For perspective, it’s useful to remember one of the main criticisms that Democratic–leaning analysts leveled at the approach embraced by President George W. Bush: that it would eventually produce a relatively flat benefit, regardless of income. This approach raised concern not just about benefit adequacy for those in the middle, but about the political implications of such a shift. ‘‘This raises the question of whether broad political support for Social Security can be sustained if workers pay very different amounts of payroll taxes but most workers receive the same level of
152
A WELL-TAILORED SAFETY NET
benefits,’’ wrote Jason Furman, now an economic adviser to President Barack Obama, in an analysis for the Center on Budget.13 While this political concern does suggest that Social Security needs to be reformed in a way that maintains its relevance for the upper-middle class as well as the working class, it doesn’t preclude a highly progressive approach to reform. Table 19.4 provides a detailed view of how A Well-Tailored Safety Net reconciles the logic of cost-saving measures that fall more heavily on higher earners with the goal of a safety net that meets the needs of retirees across the income spectrum. Once changes are fully phased in after mid-century, initial retirement benefits would be relatively flat compared to the current system, with low earners enjoying much firmer support early in retirement, meaning a higher income replacement rate. But as benefit cuts gradually unwind, the safety net would firm up for all income levels and benefits would come to closely reflect currently promised levels. This is the key to preserving a safety net that is affordable and effective. Next, I’ll provide a fuller analysis of the distribution of sacrifice in A Well-Tailored Safety Net based on the gauge of fairness set out in chapter 16: the percentage of annual income workers would have to save to overcome future benefit cuts. First, in evaluating whether the distribution is fair, it’s important to keep in mind that there are two ways for workers to overcome benefit cuts: saving more and working longer. Both responses are built into A Well-Tailored Safety Net, with each providing roughly half of the savings. In my view, it is reasonable to ask the highest earners to bear their share of the required saving through a tax hike on income over the payroll-tax ceiling. However, there’s less logic in applying additional tax increases in the context of benefit reductions that are designed with a goal of extending working careers. Therefore, as detailed in table 19.5, the biggest burden of sacrifice in A WellTailored Safety Net is borne by $100,000-earners, rather than million-dollar-earners. Finally, I’m going to wrap up this chapter by comparing A Well-Tailored Safety Net to another proposal that has yet to be introduced, but that strikes me as a logical Table 19.4 A safety net that firms up earlier for lower earners, but protects everyone A Well-Tailored Safety Net benefit in 2056 as % of base benefit
a
A Well-Tailored Safety Net benefit in 2056 based on 2009 wage levels
Age
Lowa
Avg.
High
Max.
65
81.3%
64.3%
49.7%
36.2%
$8,490 $11,070 $11,360 $10,560
70
83.5%
68.3%
55.7%
42.2%
$8,730 $11,760 $12,730 $12.300
75
85.8%
72.3%
61.7%
48.2%
$8,970 $12,450 $14,100 $14,050
80
88.1.%
76.3%
67.7%
54.2%
$9,210 $13,140 $15,470 $15,800
85
90.4%
80.3%
73.7%
60.2%
$9,450 $13,830 $16,840 $17,550
86
91.2%
80%
80%
80%
$9,530 $13,776 $18,278 $23,325
Low
Avg.
High
Max.
Reflects minimum benefit awarded based on 35 years of work. Benefit level would rise with more work years or fall with fewer.
MEASURING UP
153
Table 19.5 Distribution of sacrifice under A Well-Tailored Safety Net Annual income (2009 wage level)
Implicit tax hike as % of income from benefit reductions
Explicit tax hike as % of income
Total implied annual tax hike
$18,900
1.5%
0.5%
$370
$42,000
1.4%
1.0%
$1,010
$67,200
1.9%
1.5%
$2,270
$106,800
2.1%
1.5%
$3,810
$200,000
1.1%
2.2%
$6,610
$1 million
0.3%
2.8%
$30,610
approach for a Democrat-controlled Congress to weigh. But first, I’ll provide a little background as to how I arrived at this alternative. Given the immediate stresses on the budget and the Democrats’ rejection of carveout accounts in the past, I see little chance that an LMS-type proposal will receive serious consideration. I also find it difficult to believe that Congress would simply raise the payroll-tax rate and funnel those extra contributions straight out to beneficiaries, as proposed by Diamond-Orszag and Ball, rather than setting aside these savings in personal accounts. This view is influenced by my analysis that add-on savings accounts are more in keeping with a progressive and efficient reform of Social Security. It’s also informed by the fact that neither President Bill Clinton nor Bush was open to raising the payroll tax rate.14 This perspective leads me to conclude that a modified LMS plan, without a carveout and with a more progressive distribution of benefit cuts, is a logical approach for Democrats to consider, if they are serious about Social Security reform. Some Democrats might find this approach attractive because it includes a substantial increase in taxes on income over the payroll tax ceiling—enough to erase about 37 percent of Social Security’s financing gap. However, reflecting Obama’s disposition to limit the hit to workers with earnings moderately above $106,800, I have doubts that Congress would want to simply apply the full 12.4 percent payroll tax on wages up to roughly $195,000 in today’s terms, as called for by the LMS plan. An equivalent amount of income could be raised by imposing a roughly 5.3-percentage-point tax hike on all wages over the payroll-tax ceiling.15 I also presume that the earliest retirement age would be raised to 64, as in A Well-Tailored Safety Net, along with an increase in the Normal Retirement Age to 68. This modified LMS plan would still apply an additional 1.5 percent tax on wages. Those funds would be directed to individual accounts that would help workers overcome cuts in Social Security’s benefit formula. For the analysis to follow, I presume that accounts in the modified LMS approach would be cashed in for a lifelong annuity, despite the concern raised by Diamond and Orszag that such a requirement might not be politically sustainable. Because any degree of ownership of personal
154
A WELL-TAILORED SAFETY NET
account balances must come at a price of some reduction in benefits, the lack of any ownership in the modified LMS plan makes for a less-than-exact comparison. By contrast, in A Well-Tailored Safety Net, account balances would be drawn down through life expectancy and Social Security would bridge the gap in benefits postlife expectancy. Such a solution seems politically practical and reasonably well suited to an approach in which those who have lower life expectancies also have relatively small accounts, since they would be able to absorb less of the longevity risk that increases with income. But it would be of larger potential consequence in the modified LMS approach because the accounts of low earners financed by 1.5 percent of wages would be three times larger. A look at the distribution of sacrifice in the two plans detailed in table 19.6 shows that the brunt of the sacrifice is borne by the highest earners in the modified LMS approach and those with earnings at or near the payroll-tax ceiling in A WellTailored Safety Net. This difference reflects the Well-Tailored Safety Net approach that would close a much bigger portion of Social Security’s shortfall through measures to scale back the promise of early retirement in a progressive manner, in order to encourage delayed retirement and limit the need for tax hikes. While this modified version of the LMS plan has a more progressive distribution than the original, the change is at the bottom and top of the income scales, with average earners doing about the same. Still, the lowest earners bear a smaller burden of sacrifice in A Well-Tailored Safety Net relative to the modified LMS approach. The takeaway is that a disproportionate tax on income over the payroll-tax ceiling doesn’t serve to redistribute from the very highest earners to the lowest earners, but from the very highest earners to pretty high earners. Table 19.6 The choice Modified LMS plan
A Well-Tailored Safety Net
Income level (2009 wages)
Explicit tax hike
Annual tax bill
Total implied tax hike
$18,900a
1.5%
$285
2.4%
0.5%
$95
2.0%
$42,000
1.5%
$630
2.6%
1.0%
$420
2.4%
$67,200
1.5%
$1,010
2.6%
1.5%
$1,010
3.4%
$106,800
1.5%
$1,600
2.6%
1.5%
$1,600
3.6%
$200,000
3.3%
$6,600
3.9%
2.2%
$4,400
3.3%
$1 million
4.8%
$48,000
5.0%
2.8%
$28,000
3.1%
a
Explicit tax hike
Annual tax bill
Total implied tax hike
Implied tax hikes for $18,900 earners don’t reflect the impact of a new minimum benefit. Note: Reflects benefit levels in 2056, when all provisions in both proposals would be fully phased in. Source: Author’s assumptions of modified LMS plan reflect Office of the Chief Actuary’s assessment of LMS benefit levels, adjusted to provide for a more progressive distribution and to reflect no carve-out personal accounts. Implied tax hike reflects phase-out of benefit cuts in A Well-Tailored Safety Net.
MEASURING UP
155
This, then, is the most likely choice lawmakers will face if they take on their responsibility to reform Social Security: A high-tax plan that redistributes from the better-off to the well-off or A Well-Tailored Safety Net that is more affordable, more progressive at the lower end of the income spectrum and that would provide a stronger safety net in very old age. A key secondary choice, not reflected in the table above, is whether to close a portion of Social Security’s shortfall by limiting the growth of tax-free spending on employer-provided health coverage. A wide range of economists believe that this option could help restrain spiraling health care costs and produce higher wages, yielding additional Social Security tax contributions and higher benefit levels. This, in turn, would limit the need for a portion of the explicit rate hikes and benefit cuts that would be required in the absence of a change in the tax treatment of employerprovided health insurance.
Conclusion
Key Findings To Guide Social Security Reform • Social Security reform is really three challenges in one: a budget problem; an aging challenge; and a work force challenge. We should strive to address all three at the same time by restructuring benefits in a way that encourages workers to delay retirement but protects the safety net in very old age. The lesson of A Well-Tailored Safety Net is that we can address our budget problem while preserving a level of income security that is at least as strong as today’s Social Security system for those who live to an advanced age and face a great risk of depleting their savings. Policy makers have a simple choice: either maintain Social Security’s outdated structure that isn’t built to withstand coming demographic and fiscal pressures, or adapt it to accommodate the gains in life expectancy among our senior population and, in doing so, get more mileage out of Social Security’s limited resources. • There’s only one route to Social Security reform that can limit the need for tax hikes while still meeting the needs of both the working class and the broad middle class: A safety that firms up earlier in retirement for low earners than high earners but protects everybody’s benefits in very old age. This is the only approach to Social Security reform that can reconcile the need to preserve a strong safety net in very old age for all income groups with cost-saving measures that ask much more of high earners. The logical point of compromise between Republicans opposed to tax cuts and Democrats opposed to benefit cuts is an affordable and effective safety net that provides fair treatment across income groups and generations. • An approach to reducing benefits in which the cuts are sharpest in the initial year of retirement but grow progressively smaller until they phase out entirely has another advantage in addition to preserving the old age safety net: the cost savings come much faster. The Old-Age Risk-Sharing benefit cuts, which gradually unwind over 20 years, would save 40 percent more in 2030 than a flat benefit cut spread out over the same number of years, my analysis shows. • Gradual reductions in Social Security’s benefit formula that are stretched out over 75 years may not be well-tailored to meet the much nearer-term budget challenge. While a go-slow approach may be more politically palatable, it ignores the shape of Social Security’s financing gap, which is projected to reach 1.3 percent of GDP on an annual basis by 2030 and then proceed to stabilize for the next 45 years. Sen. Robert Bennett’s plan, which combines two such go-slow approaches, longevity indexing and progressive price indexing, would still leave a burden of $4.1 trillion in debt through 2036, or 75 percent of the debt that would be incurred if Social Security simply paid all unfunded benefit promises by redeeming trust fund bonds until they are all spent. Rather than a budget gap of 1.3 percent of GDP, the annual cost of carrying this extra debt plus
158
A WELL-TAILORED SAFETY NET
Social Security’s remaining cash shortfall under the Bennett plan would approach 1.9 percent of GDP in 2036 and keep on rising thereafter. Economists tend to like such indexing provisions because they take future policy changes out of the hands of the political system, but it’s more important to focus on the large problem we’re pretty sure we have, rather than potential problems we’re less sure about. • Reform approaches that adjust Social Security’s benefit formula tend to have two drawbacks: First, the same cut applies to retirees whether they are 62 or 92; Second, the progressive nature of the benefit formula makes it hard to adjust and produce substantial savings unless moderate earners take a big hit, as in progressive price indexing. A Well-Tailored Safety Net overcomes this challenge by introducing two highly progressive new reforms that are applied on top of the benefit formula, with both benefit cuts and tax increases (i.e. mandated savings) applied roughly in proportion to a worker’s earnings. In effect, the base benefit level, or Primary Insurance Amount unadjusted for age of retirement, would be unchanged; however the full phase-in of the base benefit would be delayed until later in retirement. A sacrifice that is roughly proportional to a worker’s career earnings level seems to be a reasonable standard of fairness. • A tax on income over the payroll-tax ceiling is consistent with a fair, effective and affordable reform of Social Security. As noted above, the Bennett plan would only moderately reduce the Social Security-related debt incurred through 2036, and it would raise federal debt levels by more than 40 percent of GDP at the end of 75 years. The implication is that a decision not to apply a tax increase on income over the payroll tax ceiling amounts to a decision to pass a significant burden of debt onto future workers. Further, while Republicans correctly note that coming demographic pressures on Social Security will result in a less-generous retirement safety net, a key question policy makers face is whether to make average earners bear a disproportionate share of this burden. Consider how the Bennett plan would impact young adults entering the work force about 15 years from now. One whose earnings averaged $1 million over the course of a career would have to set aside 0.4 percent of yearly wages to offset benefit cuts, while a $42,000-earner would have to save an extra 3.7 percent of wages. • A broad-based tax increase also is needed as part of an effective safety net. Consider this: In addition to applying a big tax hike on earnings above the payroll-tax ceiling, the reform proposals authored by former Social Security commissioner Robert Ball and liberal economists Peter Diamond and Peter Orszag both called for payroll-tax increases and reduced benefits for both the working class and the disabled. The goal of limiting Social Security reform’s impact on workers whose careers are cut short by disability makes new revenue essential. In the Bennett plan, an average earner disabled at 62 would face a 38 percent benefit cut upon reaching the Normal Retirement Age. In the Liebman-MacGuineas-Samwick proposal, which sets aside additional tax revenue in personal accounts, the same worker would face a 9 percent reduction. • The real question isn’t whether the middle class will face a bigger tax bill, but what kind of tax increases are most likely to produce constructive economic outcomes and the highest standard of living. In this context—and in the broader context of a progressive health-care reform—there is a strong case for limiting the tax exclusion for employer-provided health care. If economists are correct that taxing employer-provided coverage can help restrain health care costs by removing counterproductive incentives and making health spending more transparent, then the result wouldn’t just be
CONCLUSION
159
increased tax revenues; limiting the portion of a worker’s compensation devoted to health spending would also produce higher wage growth. This option should appeal to those who wish to preserve a strong retirement safety net because taxing a bigger portion of compensation as wages would raise a worker’s career earnings level and result in a bigger monthly benefit check relative to other means of closing Social Security’s financing gap. By contrast, while raising the payroll-tax rate would reduce the need for benefit cuts, the extra tax dollars wouldn’t yield additional benefits under Social Security’s formula. • One interesting idea that has been tossed around is to use revenue raised from a carbon tax or the auctioning of permits under a cap-and-trade system to help save Social Security.1 This is a reasonable idea that might counter one of the biggest obstacles to proactive policy making—the impression that taxpayers are giving up something but getting nothing in return. One thing that linking action on Social Security and climate change would not provide is a pot of free money. Nor would it alleviate the need for A WellTailored Safety Net that devotes fewer resources where they are less essential so that support is strong where it is most critical—in very old age. However, there does appear to be a certain compatibility between the progressive aims of A Well-Tailored Safety Net and the White House proposal to use cap-and-trade revenues to pay for its Making Work Pay tax credit of $400 per worker.2 Average earners ($42,000 in 2009) would be required to set aside an extra 1 percent of wages (phased in by 2015) in a personal account, most of which could be covered by the $400 tax cut. For low-wage earners ($18,900 in 2009), the Making Work Pay credit would easily cover their required extra savings of 0.5 percent of wages and still leave them with most of the intended tax cut. • Personal accounts financed by add-on tax contributions should be linked to Social Security reform. If efforts are limited to incentivizing more retirement savings outside of Social Security, the result could be a big hole in the safety net, since such saving would presumably be voluntary and workers would have the freedom to take their funds out of such accounts. • Add-on accounts are preferable to a straightforward tax hike because they are more likely to contribute to favorable outcomes for the economy and retirement security. Gradually raising the tax on wages—not to save for the future in individual accounts, but to write bigger benefit checks than Washington could otherwise afford—would only take us further in the direction of a nation that consumes but does not save. Further, introducing some degree of ownership within Social Security would reduce the risk faced by workers who extend their careers that they would end up getting less from Social Security if they die in their 70s. On the other hand, increasing payroll-tax rates without providing the incentive of ownership would make workers more likely to claim benefits at the first chance they get—even at the cost of income security in very old age. This is the exact opposite of what we need to do. Our best chance of saving Social Security without abdicating other critical responsibilities of government is encouraging workers to delay retirement and increasing the resources available for investment in the economy’s productive capacity so that the nation will be able to afford the promises it makes. • The case for carving out a portion of Social Security revenues to finance personal accounts is much weaker than it was a decade ago because one central rationale— protecting Social Security’s surpluses—is no longer operative now that those surpluses
160
A WELL-TAILORED SAFETY NET
have all but vanished. Now such accounts are largely a question of when we will pay Social Security’s bills—either as they come due, or years later as personal account balances begin to displace some of the need for government benefit checks. At the very least, it’s reasonable to presume that there is no particular advantage to be had by paying our bills later. • A Well-Tailored Safety Net offers an alternative approach that includes some of the most progressive features of carve-out accounts—the chance for wage earners with little savings outside of Social Security to accumulate financial assets and leave a modest inheritance—without any borrowing or undue risk. Instead of taking away resources from Social Security and depositing them in personal accounts, workers would be credited with a claim on Social Security’s future resources equal to 1.5 percent of annual wages. Although this type of account is known as a notional account, these accounts would be much more than a notion. Those who die before claiming Social Security benefits would have a claim on future program assets that is just as valuable—and inheritable— as real Treasury bonds. Providing such an inheritance for those whose families would not already receive generous survivor benefits would cost less than 1 percent of annual Social Security outlays, I estimate. These notional accounts also should be linked to ownership incentives, including lump-sum payments, to encourage delayed retirement. • Add-on accounts of significant scale could leave a hole in the safety net—even if they are invested entirely in bonds. The risk is that workers who have little savings outside of Social Security could find these accounts an enticing source of ready cash and that the political system will bend to accommodate them. Thus, the larger the scale of an add-on account for modest earners, the greater the risk to retirement security in very old age. This concern would be relatively minor in A Well-Tailored Safety Net, because low earners’ add-on accounts would be financed with a slim 0.5 percent of wages each year. Further, because an additional 1.5 percent of wages would be collected from all workers with any extra amount deposited in a supplemental savings account unrelated to Social Security’s benefit structure, average and below-average earners would be able to tap some of their contributions. Finally, there is no requirement to cash these account balances in for a lifelong annuity; the accounts would be drawn down in equal inflation-adjusted amounts through life expectancy, when the benefit reductions tied to this add-on savings account expire, leaving no risk to the safety net in very old age. • Social Security reform needs to strike a balance between workers’ need for income support toward the end of their careers and the need for income security late in life. Increasing longevity and the fact that promised benefit levels appear unaffordable suggests that, if anything, the balance needs to shift more toward income security in very old age. But the coming increase in the early-retirement penalty to 30 percent for workers who claim benefits at 62 means that the balance is shifting toward income support in a way that can be expected to undermine income security, and that’s before any benefit reductions to align Social Security’s promises and resources. A Well-Tailored Safety Net would maintain a more appropriate balance by reducing benefits early in retirement to avoid the need for benefit cuts in very old age and by lifting the earliest retirement age to 64. But this change can be made in a way that is mindful of some workers’ need for income support early in their 60s. Once it is phased in over 15 years, workers would still be able to claim 50 percent of the available benefit at 62, with the remainder becoming available in equal increments through age 68. In effect, benefits
CONCLUSION
161
would ramp up in a way that supports workers as they wind down their careers, but without sacrificing income security late in life. In connection with an increase in the Normal Retirement Age to 68, this approach would hold the maximum earlyretirement penalty at its current 25 percent. • Those state and local governments now exempted from Social Security should rejoin the system, with new hires helping to share the sacrifice that all other workers will face. A Well-Tailored Safety Net would direct this extra revenue to finance strengthened antipoverty protections for low career wage-earners and the disabled. • Social Security reform should address our actual budget problem, rather than the more manageable actuarial deficit politicians would prefer to have. With federal debt spiraling as a result of the recession and financial crisis, and the budget challenges outside of Social Security much more daunting, reform efforts need to focus on erasing most of the $5.35 trillion in debt that would be amassed if we pay all unfunded benefits through 2036—just before the trust fund is exhausted. The fact that the intragovernmental promissory notes held by the trust fund would keep Social Security officially solvent through 2036 must not distract Congress from this priority. To the extent that lawmakers want to pay these unfunded promises, they should strive to bring in new revenue to Social Security rather than presuming that the rest of the budget will have a lot of room to spare. It won’t: Even under relatively positive scenarios, Medicare and Medicaid costs are projected to double as a share of the economy by 2050. • Any delay in reforming Social Security is merely a decision to shift a bigger burden of tax hikes, benefit cuts and debt onto younger generations. It’s no wonder that about 70 percent of Americans under 45 didn’t believe they will receive the promised payments from Social Security, according to a 2005 New York Times/CBS News poll.3 Social Security reform is nothing less than a test of whether Washington can muster the discipline to make commitments it can afford, or whether the federal government will continue to make promises that breed distrust and send a signal that we don’t have the will to grapple with our financial challenges. Most importantly, Social Security reform is a test of whether we will treat younger generations fairly, while still keeping faith with our seniors.
Appendix 1
Details of A Well-Tailored Safety Net Social Security Reform Plan
Roughly half of Social Security’s shortfall would be closed by relying on additional saving and tax increases. Progressive saving offset: Average earners ($42,000 in 2009) are required to save an extra 1 percent of wages in a personal account; those earning 50 percent of the average would save an extra 0.5 percent; and those earning at least 150 percent of the average would save 1.5 percent of income. Savings would phase in from 2016 to 2018. For most workers, Treasury-only accounts would largely negate the impact of benefit cuts, which will phase in over 40 years, starting 2017. Above-average earners would have to save more on their own to overcome benefit cuts, but the provision expires at life expectancy, so insufficient savings won’t jeopardize the safety net in very old age. Fine print: After 40 years, those with career earnings up to 100 percent of Social Security’s average wage index will face a benefit cut equal to their career-earnings percentage divided by 10. Higher earners will have benefits reduced by 10 percent plus 0.175 times earnings as a percentage of the average wage index above 100 percent. Estimated Savings: 25 percent of Social Security’s $7.7 trillion present value gap. Tax over the payroll cap: A 2 percent tax on all wages above the ceiling on wages subject to Social Security taxes will be phased in from 2016 to 2019. Estimated Savings: 13 percent of Social Security’s $7.7 trillion present value gap.
164
APPENDIX
1
Table A1.1 The impact of a progressive saving offset in 2056 Income level (2009 wages)
Tax hike, % of income
Add-on account, % of base benefit
Progressive saving offset, % of base benefit
Net benefit, % of base benefit (before other measures)
$18,900
0.5%
3.8%
4.5%
99.3%
$42,000
1%
10.3%
10%
100.3%
$67,200
1.5%
18.7%
20.5%
98.2%
$106,800
1.5%
25.2%
34.5%
90.7%
Taxing employer-provided health coverage: The growth of payroll-tax-exempt spending on employer-provided health benefits will be limited to the rate of GDP growth. Estimated Savings: 18 percent of Social Security’s $7.7 trillion present value gap. Nearly half of Social Security’s gap would be closed by scaling back the promise of early retirement in a progressive way—to avoid the need for benefit cuts in very old age. Old-Age Risk-Sharing: Initial retirement benefits are reduced, but benefit cuts fully unwind over 20 years to ensure a robust safety net in very old age for retirees of all income levels. When fully phased in after 2032, a low earner would face an upfront 10 percent benefit cut, which could be offset with an enhanced 10 percent delayed retirement credit for working one year past the Normal Retirement Age. An average earner would face a 20 percent upfront cut that could be offset with credits from two extra years of work. High earners would face a 30 percent cut that could be offset with three extra years of work. Fine print: Initial benefits are reduced by an additional 0.5 percentage point a year starting in 2013 for those with up to 50 percent of average career earnings. The incremental annual reduction for higher earners is equal to 0.5 percentage point multiplied by career earnings percentage divided by 50 percent, maxing out at 1.5 percentage point for those with at least 150 percent of average career earnings. Estimated Savings: 33 percent of Social Security’s $7.7 trillion present value gap. Normal Retirement Age: The NRA would rise an extra year, going from 66 to 68 between 2017 and 2028, as the earliest retirement age climbs from 62 to 64. Together, this would limit early-retirement penalties to 25 percent, versus 30 percent under current law. Estimated Savings: 15 percent of Social Security’s $7.7 trillion present value gap.
APPENDIX
1
165
Table A1.2 Old-Age Risk-Sharing reduction in 2035 and later, retire at 65 Age
Up To 50% Of Career-Average Earnings
Career-Average Earnings
At least 150% Of Career-Average Earnings
65
10%
20%
30%
70
7.5%
15%
22.5%
75
5%
10%
15%
80
2.5%
5%
7.5%
85
0%
0%
0%
A complementary set of policies to lift up the floor of support for low earners and help workers adapt to less support from Social Security in the early years of eligibility. Poverty-related minimum benefit: Workers with 40 years of work will get a base benefit equal to 120 percent of the wage-adjusted poverty level before adjustment for age of retirement and solvency measures. Workers with 30 years of work will receive a benefit equal to 100 percent of the wage-adjusted poverty level, falling to 80 percent for 20-year workers. Disability benefits also will be subject to a povertyrelated minimum. Estimated Cost: 8 percent of Social Security’s $7.7 trillion present value gap. Phased-in benefit eligibility: Even after the earliest retirement age rises to 64 by 2028, workers would still have the option of receiving a benefit that ramps up as they wind down their careers—starting at 62. They could still receive 50 percent of their benefit at 62, with the other half ramping up in steady increments through age 68. This would provide workers with a substantial income supplement, while still shielding them from excessive early-retirement penalties that would deprive them of income security in very old age. Workers would also have two other options: They could wait until the new earliest retirement age and receive 100 percent of their benefit, or they could receive 50 percent of their benefit for as long as they like, increasing their flexibility to delay retirement while attaining a greater degree of income security in old age. Payroll-tax reductions: As an incentive to provide flexible, low-intensity jobs for older workers, employers will be exempted from payroll taxes on the first $10,000 in wages for workers 62 and older, phased in 2017-2022. The exemption would rise with future wage growth. Estimated Cost: The financial impact is reflected in the net savings for Old-Age Risk-Sharing under the expectation that revenue generated from additional work would offset this expense.
166
APPENDIX
1
Table A1.3 Benefit eligibility for workers turning 62 in . . . Age
2017
2022
2028 and later
62
96%
75%
50%
63
97%
79%
58%
64
97%
83%
67%
65
98%
88%
75%
66
98%
92%
83%
67
99%
96%
92%
68
100%
100%
100%
Maximum early-retirement penalty
25%
25%
25%
Earliest retirement age
62.2
63
64
Supplemental savings: While only those earning at least 150 percent of the average wage ($63,000 in 2009) would have to save an extra 1.5 percent of income, the full 1.5 percent of payroll will be collected from everyone by 2018, partly for ease of administration. Any extra amount would be deposited in a supplementary personal account that could provide the foundation for a universal payroll-deduction savings plan that has bipartisan support. While individuals could have pre-retirement access to these separate accounts, ideally these funds would be available for income support after age 62, when workers will face scaled-back benefit eligibility to shield them from steep penalties for early retirement. A new approach to personal accounts involves no borrowing and no risk, but still provides workers an opportunity to accumulate assets and leave a modest inheritance. Add-on accounts: As explained above, all workers would be required to save an additional share of wages in a personal account. With sacrifice roughly proportional to income, average earners will save an extra 1 percent of income; those earning half the average wage will save 0.5 percent of income; and those earning at least 150 percent of the average would save 1.5 percent of income. Account balances related to these add-on deposits would be drawn down in equal (inflation-adjusted) amounts through life expectancy, when the progressive saving offset expires. Balances would be subject to the same taxes as Social Security’s traditional annuity. Delayed Retirement Accounts: Each worker’s add-on account will be complemented by the equivalent of a risk-free Treasury portfolio equal to 1.5 percent of wages, providing an inheritable claim on trust fund bonds to those families who would not already receive generous survivor benefits. For workers retiring by 65, this portion of the account will be cashed in for their regular Social Security annuity.
APPENDIX
1
167
Table A1.4 A well-tailored account structure Treasuries Add-on account earmarked to Delayed Income level, % deposit as % of Retirement Account of average wage income as % of income
Supplemental savings as % of income
Total account size as % of income
Up to 50%
0.5%
1.5%
1%
3%
100%
1%
1.5%
0.5%
3%
150% and up
1.5%
1.5%
0%
3%
Estimated Cost: The cost of providing an inheritance to those who die before retirement would equal 1.5 percent of Social Security’s $7.7 trillion present value gap. Delayed retirement as a route to ownership: Workers who retire after the Normal Retirement Age will receive lump-sum payments equal to one half of an enhanced delayed retirement credit equal to 10 percentage points of a worker’s base benefit. Offering the incentive of ownership will encourage workers to delay retirement and thereby strengthen income security in old age. For work between 65 and 68, the default option would be for workers to receive the full increase in the form of bigger benefits; however, to the extent that modest earners may be more likely to extend their careers if the rewards of doing so are enjoyed immediately, it makes sense to provide workers a lump-sum option before the NRA. Fine print: The 10-percentage-point increase in the traditional benefit would be reduced prior to life expectancy based on the progressive saving offset to reflect the portion of a worker’s benefit tied to add-on savings. The full 5-percentage-point lump sum transfer for delayed retirement would be reduced for those who don’t meet a work requirement in each of the five years up to and including the year of delayed retirement. Lump-sum transfers would equal the present value of a 5-percentagepoint increase in benefits through life expectancy. Additional details State & local government workers: New state and local government workers will be covered under Social Security starting in 2013. This change will spread the sacrifice of Social Security reform more broadly and, more specifically, offset the cost of a stronger minimum benefit for lower earners. It also will give public employers the same incentive as private employers to provide opportunities for older workers. Estimated Savings: 8 percent of Social Security’s $7.7 trillion present value gap. Spousal benefit changes: Spousal benefits will fall to 45 percent of the primary earner’s base benefit (PIA) for spouses of those with career-average earnings, rising to 47.5
168
APPENDIX
1
Table A1.5 Initial Old-Age Risk-Sharing reduction for those disabled in 2032 and later Disability age
Up To 50% Of Career-Average Earnings
Career-Average Earnings
At least 150% Of Career-Average Earnings
65
5%
10%
15%
60
4.1%
8.3%
12.4%
55
3.5%
7.1%
10.6%
50
3.1%
6.2%
9.2%
percent for the spouse of someone with up to 50 percent of career-average earnings. These reductions mirror the progressive saving offset and also phase out at life expectancy. Ideally, spouses will overcome these cuts with account accumulations. Spousal benefits are collected according to Old-Age Risk-Sharing as it applies to the lowest earners. Benefits for spouses of above-average earners are capped at the level of benefits for the spouse of a career-average earner, with the change phased in over 40 years. Survivor benefit changes: Survivor benefits for retired spouses reflect both the OldAge Risk-Sharing and progressive saving offset reductions of the higher earner. Once the saving offset expires, benefits are reduced by the progressive saving offset of the spouse until life expectancy. Disability benefits and Old-Age Risk-Sharing: Those disabled by age 65 would face a cumulative benefit reduction equal to 60 percent of that faced by the nondisabled, but these reductions will be spread out over more years. Reductions would take effect upon claiming disability benefits and gradually phase out by 20 years past the NRA. As an example, an average earner disabled at 65 would face an initial benefit cut of 10 percent compared to 20 percent for the non-disabled. Disability benefits and progressive saving offset: Benefit reductions take effect upon claiming disability benefits but are lowered based on the age of disability with the purpose of stretching the add-on account balances through life expectancy, when the progressive saving offset expires. The saving offset also is proportionally lowered based on number of disability years before 62. In addition, the saving offset is reduced for above-average earners. The formula for the non-disabled (10 percent + 0.175 times career earnings as a percentage of the average wage above 100 percent) will phase down for those disabled by age 60 to 10 percent plus 0.15 times percentage of career earnings above 100 percent. Thus, before adjusting for the proportion of disability years, the offset for someone with 200 percent of career-average earnings would be reduced from 27.5 percent to 25 percent, when fully phased in.
Appendix 2
Solvency Impact of A Well-Tailored Safety Net, Billions of 2009 Dollars
Table A2.1 Limit tax-free Saving Tax over health offset cap care
Year
Old-Age RiskSharing
NRA To 68
State and Local
Min. benefit
Inherit
Net Balance
2013
0
0
0
0
0
1
(1)
0
30
2015
1
0
0
0
6
6
(2)
0
14
2020
17
0
2
28
26
14
(8)
(1)
(22)
2025
51
4
8
29
34
21
(13)
(1)
(67)
2030
86
24
21
32
43
25
(17)
2()
(68)
2035
109
53
39
34
49
31
(22)
(3)
(39)
2040
123
64
61
37
58
37
(25)
(4)
6
2045
129
66
84
41
63
38
(28)
(5)
40
2050
132
72
111
44
68
35
(30)
(6)
64
2055
138
77
142
48
73
30
(33)
(7)
78
2060
149
82
176
52
81
24
(35)
(9)
84
2065
160
90
207
55
89
20
(39)
(10)
89
2070
174
96
237
60
100
12
(42)
(12)
78
2075
188
205
262
65
109
7
(46)
(13)
52
2080
203
117
287
70
117
5
(50)
(15)
21
2083
214
120
303
73
129
1
(53)
(16)
1
Note: Old-Age Risk-Sharing impact factors in the shift in the earliest retirement age to 64, taxing employer-sponsored health coverage, and a reduction in the payroll taxes that employers owe on workers 62 and older. Source: Solvency impact is adapted from Office of the Chief Actuary analyses for a 3 percemt tax over the payroll ceiling, shifting the Normal Retirement Age to 68, taxing employer-sponsored health coverage, and bringing state and local government workers back into Social Security. Other measures are based on author’s own calculations.
Notes
PREFACE 1. Nancy Pelosi, interview by George Stephanopoulos, This Week, ABC, May 1, 2005.
INTRODUCTION 1. President-elect Barack Obama, transcript of interview by Washington Post, January 15, 2009, http://media.washingtonpost.com/wp-srv/politics/documents/obama_011509.pdf ? sid=ST2009011504146. 2. U.S. Government Accountability Office, ‘‘The Nation’s Long-Term Fiscal Outlook: March 2009 Update,’’ Data from Baseline Extended Simulation Based on Trustees’ Assumptions for Social Security and Medicare,http://gao.gov/special.pubs/longterm/march09baseline _extended_trustees.pdf . 3. The GAO Baseline Extended scenario assumes that spending outside of the three big entitlements and debt interest will fall to 8.5% of GDP. Going back to 1962, the Congressional Budget Office’s December 2007 ‘‘Long-Term Budget Outlook,’’ shows that such spending hasn’t dropped below its 1998 level of 8.6% of GDP. Historical spending is provided in Supplemental Data, Figure 1-1, http://cbo.gov/ftpdocs/88xx/doc8877/ SupplementalData.xls. 4. The GAO Baseline Extended scenario assumes a Debt-to-GDP ratio of 56.1% at the end of 2019. This compares to a ratio of 82.4% seen by the Congressional Budget Office’s ‘‘A Preliminary Analysis of the President’s Budget,’’ March 2009, p. 10, http://cbo.gov/ ftpdocs/100xx/doc10014/03-20-PresidentBudget.pdf . 5. U.S. Government Accountability Office, ‘‘The Nation’s Long-Term Fiscal Outlook: March 2009 Update,’’ Data from Alternative Simulation Based on Trustees’ Assumptions for
172
NOTES
Social Security and Medicare, http://gao.gov/special.pubs/longterm/march09alternative _trustees.pdf. 6. Peter Orszag, ‘‘Health Costs Are the Real Deficit Threat,’’ Wall Street Journal, op-ed page, May 15, 2009. 7. Center on Budget and Policy Priorities, ‘‘Statement of Robert Greenstein, Executive Director, on the New Report from the Congressional Budget Office,’’ March 20, 2009. 8. Congressional Budget Office, ‘‘A Preliminary Analysis of the President’s Budget,’’ March 2009, p. 10. 9. Social Security Administration, Office of the Chief Actuary, The 2009 OASDI Trustees Report, Supplemental Single-Year Tables, Operations of the Combined OASI and DI Trust Funds in Current Dollars, http://www.ssa.gov/OACT/TR/2009/lr6f8.html. 10. Social Security Administration, Office of the Chief Actuary, The 2009 OASDI Trustees Report, Supplemental Single-Year Tables, Table V.A2—Social Security Area Population and Dependency Ratios, 1950-2085, http://www.ssa.gov/OACT/TR/2009/lr5a2.html. 11. U.S. Bureau of the Census, State & County QuickFacts, Persons 65 years and old and over, percent, 2007, http://quickfacts.census.gov/qfd/states/12000.html . 12. Social Security Administration, Office of the Chief Actuary, The 2009 OASDI Trustees Report, Supplemental Single-Year Tables, Table IV.B2—Covered Workers and Beneficiaries, http://www.ssa.gov/OACT/TR/2009/IV_LRest.html#345423. 13. Social Security and Medicare Boards of Trustees, Status of the Social Security and Medicare Programs: A Summary of the 2009 Annual Reports, http://www.ssa.gov/OACT/TRSUM/ index.html (accessed June 11, 2009). 14. President-elect Barack Obama, transcript of interview by Washington Post, January 15, 2009, http://media.washingtonpost.com/wp-srv/politics/documents/obama_011509.pdf ? sid=ST2009011504146. 15. By some measures, the size of Social Security’s trust fund financing gap has remained fairly stable. But by the most relevant measure, which includes the cost of benefit promises related to Treasury’s $2.4 trillion debt to Social Security, the financing gap has escalated. 16. House Majority Leader Steny Hoyer, keynote address at Bipartisan Policy Center forum, ‘‘Unprecedented Federal Debt: Putting Our Fiscal House in Order,’’ May 6, 2009, 111th Cong., 1st sess., http://www.hoyer.house.gov/newsroom/index.asp?ID=1383. 17. Author’s calculations based on The 2009 OASDI Trustees Report projections about the future of the program and economic factors, including interest rates. 18. U.S. Bureau of the Census, 2008 National Population Projections, Summary Tables, Table 2: Projections of the Population by Selected Age Groups and Sex for the United States: 2010 to 2050, http://www.census.gov/population/www/projections/files/nation/summary/ np2008-t2.xls. 19. Social Security Administration, Office of Retirement and Disability Policy, ‘‘Income of the Population 55 or Older, 2006,’’ released in 2009, http://www.ssa.gov/policy/docs/ statcomps/income_pop55/2006/sect11.html#table11.4.
CHAPTER 1 1. Sen. Barbara Boxer, Candidate Statement, Official Voter Information Guide, California Secretary of State, 2004, http://vote2004.sos.ca.gov/voterguide/cand/sen_dem.htm. 2. Sen. Barbara Boxer, interview by Soledad O’Brien, American Morning, CNN, January 21, 2004.
NOTES
173
3. Sen. Barbara Boxer, ‘‘Peace of Mind vs. a Gamble,’’ speech delivered February 11, 2005, 109th Cong., 1st sess., http://boxer.senate.gov/issues/20050211_print.cfm. 4. Social Security and Medicare Boards of Trustees, Status of the Social Security and Medicare Programs: A Summary of the 2009 Annual Reports, http://www.ssa.gov/OACT/TRSUM/ index.html (accessed June 11, 2009). 5. Carolyn Lochhead, ‘‘Officials urge action on Social Security—GAO chief says dispute is over semantics,’’ San Francisco Chronicle, January 22, 2005. 6. Ibid. 7. Carolyn Lochhead, ‘‘Bush’s Social Security Proposal Takes a Hit—GAO finds problems with privatization,’’ San Francisco Chronicle, March 10, 2005. 8. Social Security Administration, Office of the Chief Actuary, Trust Fund Data, Old-Age, Survivors, and Disability Insurance Trust Funds, 1957-2008, updated January 29, 2009, http://www.ssa.gov/OACT/STATS/table4a3.html. 9. Office of Management and Budget, Budget of the United States Government, Fiscal Year 2010, Historical Tables, Summary of Receipts, Outlays, and Surpluses or Deficits: 17892014,Washington, DC: GPO, 2009, http://www.whitehouse.gov/omb/budget/fy2010/assets/ hist01z1.xls. 10. Social Security Administration, Office of the Chief Actuary, The 2009 OASDI Trustees Report, Supplemental Single-Year Tables, Operations of the Combined OASI and DI Trust Funds in Current Dollars, http://www.ssa.gov/OACT/TR/2009/lr6f8.html. 11. Carolyn Lochhead, ‘‘Officials urge action on Social Security—GAO chief says dispute is over semantics,’’ San Francisco Chronicle, January 22, 2005. 12. Rep. Nancy Pelosi, interview by George Stephanopoulos, This Week, ABC, May 1, 2005. 13. Congressional Budget Office, ‘‘A Preliminary Analysis of the President’s Budget,’’ March 2009, p. 10, http://cbo.gov/ftpdocs/100xx/doc10014/03-20-PresidentBudget.pdf. 14. Office of Management and Budget, Budget of the United States Government, Fiscal Year 2000, Analytical Perspectives, Washington, DC: GPO, 1999, p. 337. 15. Office of Management and Budget, Budget of the United States Government, Fiscal Year 2010, Analytical Perspectives, p. 345, Washington, DC: GPO, 2009, http://www.whitehouse .gov/omb/budget/fy2010/assets/spec.pdf. 16. Social Security and Medicare Boards of Trustees, A Summary of the 2007 Annual Social Security and Medicare Trust Fund Reports, ‘‘A Message from the Public Trustees,’’ p. 11, www.ssa.gov/history/pdf/tr07summary.pdf. 17. Rep. Jim Cooper, ‘‘The Incumbent Party,’’ Wall Street Journal, op-ed page, April 13, 2007. 18. John Rother, Director of Policy and Strategy, AARP, ‘‘Social Security: Do We Have to Act Now?’’ Senate Special Committee on Aging, February 3, 2005, 109th Cong., 1st sess. 19. Ibid. 20. Social Security Administration, Office of the Chief Actuary, The 2009 OASDI Trustees Report, Supplemental Single-Year Tables, Operations of the Combined OASI and DI Trust Funds in Current Dollars, http://www.ssa.gov/OACT/TR/2009/lr6f8.html. 21. Congressional Budget Office, ‘‘A Preliminary Analysis of the President’s Budget,’’ March 2009, p. 10. 22. Congressional Budget Office, ‘‘A Preliminary Analysis of the President’s Budget,’’ March 2009, Supplemental Data on Spending Projections, http://cbo.gov/budget/factsheets/ 2009b/oasdiTrustfund.pdf.
174
NOTES
23. Social Security Administration, Office of the Chief Actuary, The 2009 OASDI Trustees Report, Supplemental Single-Year Tables, Operations of the Combined OASI and DI Trust Funds in Current Dollars, http://www.ssa.gov/OACT/TR/2009/lr6f8.html. 24. Portions of this chapter were previously published under the title ‘‘Security Breach’’ by Investor’s Business Daily on April 24, 2007, and appear here courtesy of IBD.
CHAPTER 2 1. Social Security Administration, 1988 Public Trustees’ Symposia on the Implications of the Social Security Trust Fund Build-Up, http://www.ssa.gov/history/reports/trustees/ publictrustees.html. 2. Edward Gramlich, 1988 Public Trustees’ Symposia on the Implications of the Social Security Trust Fund Build-Up, ‘‘Comment on the Lewin-ICF Report,’’ June 27, 1988, http:// www.ssa.gov/history/reports/trustees/gramlich.html. 3. Herman Leonard, 1988 Public Trustees’ Symposia on the Implications of the Social Security Trust Fund Build-Up, Transcript of Policy Symposium, September 16, 1988, http:// www.ssa.gov/history/reports/trustees/transcript2.html. 4. Sen. Daniel Patrick Moynihan, U.S. Senate Floor, (recounting Today Show appearance with Sen. John Heinz), debate on Social Security Trust Funds, 101st Cong., 2nd sess., Congressional Record (April 23, 1990): S4806. 5. Sen. Kent Conrad, Senate Floor Debate on Removing Social Security from the Budget, Motion to Waive the Budget Act Re: S.3167, ‘‘A bill to cut Social Security contribution rates and return Social Security to pay-as-you-go financing,’’ 101st Cong., 2nd sess., Congressional Record (October 10, 1990): S14855. 6. U.S. Senate Roll Call Votes, 101st Cong., 2nd sess., Vote Summary on the Motion to Waive the Budget Act Re: S.3167, ‘‘A bill to cut Social Security contribution rates and return Social Security to pay-as-you-go financing,’’ October 10, 1990, http://www.senate.gov/ legislative/LIS/roll_call_lists/roll_call_vote_cfm.cfm?congress=101&session=2&vote=00262. 7. Office of Management and Budget, Budget of the United States Government, Fiscal Year 2010, Historical Tables, Summary of Receipts, Outlays, and Surpluses or Deficits: 17892014,Washington, DC: GPO, 2009, http://www.whitehouse.gov/omb/budget/fy2010/assets/ hist01z1.xls. 8. Douglas W. Elmendorf, Jeffrey B. Liebman and David W. Wilcox, ‘‘Fiscal Policy and Social Security Policy During the 1990s,’’ in American Economic Policy in the 1990s, Cambridge, Mass.: MIT Press, 2002, p. 93. 9. Weekly Compilation of Presidential Documents, Volume 41, Number 13, ‘‘Remarks in a Discussion on Strengthening Social Security in Cedar Rapids,’’ March 30, 2005, Washington, DC: GPO, 2005, p. 529, fdsys.gpo.gov/fdsys/pkg/WCPD-2005-04-04/pdf/WCPD-200504-04-Pg527.pdf. 10. House of Representatives Committee on Ways and Means, Subcommittee on Social Security, Social Security Improvements for Women, Seniors and Working Americans, 107th Cong., 2nd sess., March 6, 2002. 11. Board of Trustees of the Federal Old-Age and Survivors Insurance and Federal Disability Insurance Trust Funds, Report of the 1994-1996 Advisory Council on Social Security, January 1997, http://www.ssa.gov/history/reports/adcouncil/index.html. 12. Unlike more recent proposals that have involved investing a portion of existing Social Security payroll taxes, this third group would have applied a tax increase of 1.5 percent of taxable wages to finance the transition to personal accounts.
NOTES
175
13. Weekly Compilation of Presidential Documents, Volume 34, Number 23, ‘‘Remarks by the President to Saver Summit,’’ Washington D.C., June 4, 1998, Washington, DC: GPO, 1998, p. 1039, http://www.gpo.gov/fdsys/pkg/WCPD-1998-06-08/html/WCPD-1998-0608-Pg1036.htm. 14. Weekly Compilation of Presidential Documents, Volume 34, Number 15, ‘‘Address by the President to a National Forum on Social Security,’’ Kansas City, Mo., April 7, 1998, Washington, DC: GPO, 1998, p. 588, http://www.gpo.gov/fdsys/pkg/WCPD-1998-04-13/html/ WCPD-1998-04-13-Pg585-4.htm. 15. The Board of Trustees, Federal Old-Age and Survivors Insurance and Federal Disability Insurance Trust Funds, The 2009 OASDI Trustees Report, Table IV.B5: Components of 75-Year Actuarial Balance Under Intermediate Assumptions, Washington, DC: GPO, 2009, p. 61, http://www.socialsecurity.gov/OACT/TR/2009/IV_LRest.html#276411. 16. President-elect Barack Obama, transcript of interview by Washington Post, January 15, 2009, http://media.washingtonpost.com/wp-srv/politics/documents/obama_011509.pdf ? sid=ST2009011504146.
CHAPTER 3 1. Weekly Compilation of Presidential Documents, Volume 41, Number 14, ‘‘Remarks on Strengthening Social Security in Parkersburg,’’ April 5, 2005, Washington, DC: GPO, 2005, P. 561. 2. U.S. Senate, debate on Social Security, 109th Cong., 1st sess., Congressional Record (April 5, 2005): S3231, http://www.govtrack.us/congress/record.xpd?id=109-s20050405-45. 3. U.S. Senate, debate on Social Security, 109th Cong., 1st sess., Congressional Record (April 5, 2005): S3232. 4. Senator Mikulski of Maryland, U.S. Senate, ‘‘Checklist for Change,’’ 109th Cong., 2nd sess., Congressional Record (June 28, 2006): S6593, http://fdsys.gpo.gov/fdsys/pkg/ CREC-2006-06-28/pdf/CREC-2006-06-28-pt1-PgS6592.pdf. 5. Interest costs derived from Congressional Budget Office, ‘‘A Preliminary Analysis of the President’s Budget,’’ March 2009, p. 10; non-defense discretionary spending comes from Office of Management and Budget, Budget of the United States Government, Fiscal Year 2010, Updated Summary Tables, May 2009, p. 8, http://www.whitehouse.gov/omb/budget/fy2010/ assets/summary.pdf. 6. Congressional Budget Office, The Long-Term Budget Outlook, December 2007, Supplemental Data, Figure 1.1, http://cbo.gov/ftpdocs/88xx/doc8877/SupplementalData.xls . 7. C. Eugene Steuerle, ‘‘Alternatives to Strengthen Social Security,’’ Testimony before the House Committee on Ways and Means, 109th Cong., 1st sess., May 12, 2005. 8. Weekly Compilation of Presidential Documents, Volume 34, Number 23, ‘‘Remarks by the President to Saver Summit,’’ Washington D.C., June 4, 1998, Washington, DC: GPO, 1998, p. 1037, http://www.gpo.gov/fdsys/pkg/WCPD-1998-06-08/html/WCPD-1998-06-08Pg1036.htm. 9. Weekly Compilation of Presidential Documents, Volume 34, Number 50, ‘‘Remarks by the President in Opening White House Conference on Social Security,’’ December 8, 1998, Washington, DC: GPO, 1998, p. 2443, http://www.gpo.gov/fdsys/pkg/WCPD-1998-12-14/pdf/ WCPD-1998-12-14-Pg2441.pdf. 10. Congressional Budget Office, The Long-Term Budget Outlook, June 2009, Supplemental Data, Data for Figure in Box 1.2, http://cbo.gov/ftpdocs/102xx/doc10297/SupplementalData2009LTBO.xls.
176
NOTES
11. Peter Orszag, ‘‘Health Costs Are the Real Deficit Threat,’’ Wall Street Journal, op-ed, May 15, 2009. 12. Social Security Administration, Office of the Chief Actuary, The 2009 OASDI Trustees Report, Supplemental Single-Year Tables, Operations of the Combined OASI and DI Trust Funds, in Constant 2009 Dollars, http://www.ssa.gov/OACT/TR/2009/lr6f7.html. 13. Elmendorf, Liebman, Wilcox, ‘‘Fiscal Policy and Social Security Policy During the 1990s,’’ 2002, p. 87. 14. Author’s calculations based solely on information provided by the Office of the Chief Actuary in The 2009 OASDI Trustees Report, Supplemental Single-Year Tables, including measures for interest rates, GDP and Social Security’s cost gap until the trust fund expires, http:// www.ssa.gov/OACT/TR/2009/lrIndex.html. 15. The annual gap between Social Security’s cost and tax revenue as a percentage of GDP can be seen in Table VI.F4 on p. 184 of The 2009 OASDI Trustees Report. In this present-law scenario, which assumes that Treasury redeems all of the bonds in the trust fund by floating new public debt and the trust fund is exhausted in 2037, interest payments on this Social Security-related debt would exceed Social Security’s cost gap in every year starting 2042. By 2083, the debt that is a legacy of the trust-fund redemptions would reach 38.5 percent of GDP. 16. The projected average wage for 2009 and later years is provided by the Office of the Chief Actuary in The 2009 OASDI Trustees Report, Supplemental Single-Year Tables, Table V1.F6: Selected Economic Variables, Calendar Years 2008-2085.
CHAPTER 4 1. U.S. Senate, debate on Social Security, 109th Cong., 1st sess., Congressional Record (April 5, 2005): S3235, http://fdsys.gpo.gov/fdsys/pkg/CREC-2005–04–05/html/CREC2005–04–05-pt1-PgS3227–6.htm. 2. Social Security Administration, Stephen C. Goss, Chief Actuary, ‘‘Estimated Financial Effects of the ‘Social Security Personal Savings Guarantee and Prosperity Act of 2005,’’’ April 20, 2005, p. 7, http://www.ssa.gov/OACT/solvency/RyanSununu_20050420.pdf. 3. Author’s calculations based on ‘‘Estimated Financial Effects of the ‘Social Security Personal Savings Guarantee and Prosperity Act of 2005,’’’ Table 1, updated to reflect the latest financial status presented in The 2009 Trustees Report. 4. Office of the Chief Actuary, ‘‘Effects of the ‘Personal Savings Guarantee and Prosperity Act,’’’ Table 1.e: Specified and Total General Fund Transfers to the OASDI Trust Funds, April 20, 2005. 5. Office of the Chief Actuary, ‘‘Effects of the ‘Personal Savings Guarantee and Prosperity Act,’’’ April 20, 2005. 6. Robert Greenstein and Richard Kogan, Center on Budget and Policy Priorities, ‘‘The Ryan-Sununu Social Security Plan: ‘Solving’ the Long-Term Social Security Shortfall by Raiding the Rest of the Budget,’’’ April 26, 2005, http://www.cbpp.org/cms/?fa=view&id=222. 7. Office of the Chief Actuary, ‘‘Effects of the ‘Personal Savings Guarantee and Prosperity Act,’’’ Table 1.e: Specified and Total General Fund Transfers to the OASDI Trust Funds, April 20, 2005. 8. Stephen Moore and Peter Ferrara, Institute for Policy Innovation, ‘‘Social Security Reform and National Spending Restraint,’’ February 8, 2005. 9. Richard W. Stevenson, ‘‘GOP Divided as Bush Views Social Security,’’ New York Times, January 6, 2005.
NOTES
177
10. Jonathan Weisman and Jim VandeHei, ‘‘GOP Is Divided on Social Security Push,’’ Washington Post, January 7, 2005. 11. ‘‘Greenspan Likes Social Security Private Accounts, But Urges Caution,’’ SeniorJournal.com, February 17, 2005, http://seniorjournal.com/NEWS/SocialSecurity/ 5–02–16GreenspanSays.htm (accessed June 11, 2009). 12. Eric Pfeiffer, ‘‘FEF Says Personal Accounts Being Hijacked,’’ Beltway Buzz on National Review Online, March 9, 2005 (accessed June 11, 2009). 13. Goss, ‘‘Effects of the ‘Personal Savings Guarantee and Prosperity Act,’’’ Table 1.e: Specified and Total General Fund Transfers to the OASDI Trust Funds, April 20, 2005, p. 7. 14. Higher replacement rates for lower earners are a function of Social Security’s formula for deriving the Primary Insurance Amount. In 2009, new retirees turning 62 would receive about 90 percent of their career indexed earnings up to $9,000. For career earnings between $9,000 and $54,000, Social Security would replace 32 percent of earnings, but just 15 percent of earnings from $54,000 to the payroll ceiling of $106,800. For further explanation, see: http://www.ssa.gov/OACT/COLA/piaformula.html. 15. Congressional Budget Office, ‘‘How Pension Financing Affects Returns to Different Generations,’’ September 22, 2004, p. 5. 16. The White House, Office of the Press Secretary, ‘‘Fact Sheet: Setting the Record Straight,’’ February 3, 2005. 17. Social Security Administration, Stephen C. Goss, Chief Actuary, ‘‘Estimated Financial Effects of ‘A Nonpartisan Approach to Reforming Social Security,’’’ November 17, 2005, p. 4. 18. Goss, ‘‘Effects of the ‘Personal Savings Guarantee and Prosperity Act,’’’ Table 1.e: Specified and Total General Fund Transfers to the OASDI Trust Funds, April 20, 2005, p. 11.
CHAPTER 5 1. Social Security Administration, Office of the Chief Actuary, ‘‘Long Range Solvency Provisions: Provisions Affecting Level of Monthly Benefits,’’ B7, Summary Measures and Graphs, OASDI Cost Rates and Income Rates, http://www.ssa.gov/OACT/solvency/provisions/charts/ chart_run168.html. 2. The White House, ‘‘Personal Retirement Accounts’’ in Strengthening Social Security for the 21st Century, February 2005, p. 8, http://georgewbush-whitehouse.archives.gov/infocus/ social-security/200501/socialsecurity.pdf. The Office of Management and Budget estimated that although President Bush’s personal account would start small relative to the eventual scale of the accounts and would only phase in gradually, it would still require $754 billion in borrowing over the first decade. 3. Weekly Compilation of Presidential Documents, Volume 41, Number 17, ‘‘The President’s News Conference,’’ April 28, 2005, Washington, DC: GPO, 2005, p. 684, http://www.gpo .gov/fdsys/pkg/WCPD-2005-05-02/pdf/WCPD-2005-05-02-Pg683.pdf. 4. Edmund L. Andrews, ‘‘Beware the Easy Fix for Social Security,’’ New YorkTimes, May 15, 2005. 5. Social Security Administration, Stephen C. Goss, Chief Actuary, ‘‘Estimated Financial Effects of a Comprehensive Social Security Reform Proposal Including Progressive Price Indexing,’’ Table 1, February 10, 2005, updated by author to reflect The 2009 OASDI Trustees Report, http://www.ssa.gov/OACT/solvency/RPozen_20050210.pdf. 6. Department of Treasury, ‘‘Social Security Reform: Strategies for Progressive Benefit Adjustments,’’ Issue Brief No. 5, June 2008, p. 13.
178
NOTES
7. David Jackson, ‘‘Bush draws up ambitious domestic agenda,’’ USA Today, January 22, 2007. 8. Andrew J. Rettenmaier and Thomas R. Saving, National Center for Policy Analysis, ‘‘Social Security Reform without Illusion: The Five Percent Solution,’’ December 17, 2004. 9. Thomas Saving, e-mail exchange with author, April 21, 2008. 10. Paul Krugman, ‘‘A Gut Punch to the Middle,’’ New York Times, May 2, 2005; National Committee to Preserve Social Security and Medicare, ‘‘Social Security: ‘Price Indexing’ and Middle-Class Benefit Cuts.’’ 11. Social Security Administration, Office of the Chief Actuary, The 2009 OASDI Trustees Report, Supplemental Single-Year Tables, Table V.B1: Principal Economic Assumptions, Calendar Years 1960-2083, http://www.ssa.gov/OACT/TR/2009/lr5b1.html. 12. Social Security Administration, Office of the Chief Actuary, ‘‘Long Range Solvency Provisions: Provisions Affecting Level of Monthly Benefts,’’ B6: reduce PIA formula factors to that benefits grow by inflation rather than increases in real wages, http://www.ssa.gov /OACT/solvency/provisions/charts/chart_run176.html. 13. Office of the Chief Actuary, ‘‘Effects of Progressive Price Indexing,’’ February 2005, Table B1, OASDI Benefit Before Offset, Steady Maximum Earner. 14. Office of the Chief Actuary, ‘‘Effects of Progressive Price Indexing,’’ February 2005, p. 1. 15. Social Security Administration, Office of Retirement and Disability Policy, ‘‘Income of the Population 55 or Older, 2006,’’ released in 2009, Table 9.A1: Relative Importance of Social Security for Beneficiary Aged Units. 16. Weekly Compilation of Presidential Documents, Volume 41, Number 6, ‘‘Remarks in a Discussion on Strengthening Social Security in Tampa, Florida,’’ February 4, 2005, Washington, DC: GPO, 2005, p. 180-81, http://fdsys.gpo.gov/fdsys/pkg/WCPD-2005-02-14/pdf/ WCPD-2005-02-14.pdf. 17. Weekly Compilation of Presidential Documents, Volume 41, Number 17, ‘‘The President’s News Conference,’’ April 28, 2005, Washington, DC: GPO, 2005, p. 684. 18. N. Gregory Mankiw, ‘‘How to Screw Up Social Security,’’ Fortune, March 15, 1999, http://www.economics.harvard.edu/files/faculty/40_mar15.html. 19. Ibid. 20. Weekly Compilation of Presidential Documents, Volume 41, Number 17, ‘‘The President’s News Conference,’’ April 28, 2005, Washington, DC: GPO, 2005, p. 696.
CHAPTER 6 1. David Rosenbaum, ‘‘Bush Backed a Limited Bill on Benefits, Senator Says,’’ New York Times, June 22, 2005, http://query.nytimes.com/gst/fullpage.html?res=9D0CE4D9113 BF931A15755C0A9639C8B63 (accessed June 11, 2009). 2. Thomas, The Library of Congress, S.426, ‘‘A bill to amend title II of the Social Security Act to provide for progressive indexing and longevity indexing of Social Security old-age insurance benefits, 111th Cong., 1st sess. (accessed June 11, 2009). 3. Social Security Administration, Stephen C. Goss, Chief Actuary, ‘‘Estimated Financial Effects of the ‘Social Security Solvency Act of 2009’, S.426,’’ Table 1, February 12, 2009, updated by author to reflect The 2009 Trustees Report, http://www.ssa.gov/OACT/solvency/ RBennett_20090212.pdf.
NOTES
179
4. U.S. Senate, introduction of S.2427, ‘‘The Sustainable Solvency for Social Security Act,’’ 109th Cong., 2nd sess., Congressional Record (March 16, 2006): S2318. 5. Office of the Chief Actuary, ‘‘Effects of the ‘Social Security Solvency Act of 2009’, S.426,’’ February 12, 2009, Table B, Proposal Benefit Percent Reduction, Life Expectancy Indexing. 6. U.S. Senate, introduction of S.2427, ‘‘The Sustainable Solvency for Social Security Act,’’ 109th Cong., 2nd sess., Congressional Record (March 16, 2006): S2318. 7. U.S. Senate, introduction of S.2427, ‘‘The Sustainable Solvency for Social Security Act,’’ 109th Cong., 2nd sess., Congressional Record (March 16, 2006): S2317. 8. Social Security Administration, Office of the Chief Actuary, Fact Sheet on the Old-Age, Survivors, and Disability Insurance Program, Table A: Benefits in Current-Payment Status, December 31, 2008, http://www.ssa.gov/OACT/FACTS/fs2008_12.pdf. 9. Office of the Chief Actuary, ‘‘Effects of the ‘Social Security Solvency Act of 2009’, S.426,’’ February 12, 2009, p. 4-6. 10. Joint Committee on Taxation,Tax Expenditures for Health Care (JCX-66-08), Senate Committee on Finance, 110th Cong., 2nd sess., July 31, 2008, p. 2, http://finance.senate.gov /hearings/testimony/2008test/073108ektest.pdf. 11. Social Security Administration, Stephen C. Goss, Chief Actuary, ‘‘Estimated Financial Effects of the ‘Social Security Personal Savings Guarantee and Prosperity Act of 2008,’’’ May 21, 2008, Plan Specification, p. 2-4, http://www.ssa.gov/OACT/solvency/PRyan_20080521 .pdf. 12. Social Security Administration, Office of the Chief Actuary, ‘‘Internal Real Rates of Return under the OASDI Program for Hypothetical Workers,’’ April 2009, Table 3: Payable Benefits Scenario, http://www.ssa.gov/OACT/NOTES/ran5/index.html. 13. Office of the Chief Actuary, ‘‘Effects of the ‘Social Security Personal Savings Guarantee Act of 2008,’’’ May 21, 2008, Effects on Trust Fund Assets and Unfunded Obligations, p. 13. 14. H.R. 6110, Roadmap for America’s Future Act of 2008, 110th Cong., 2nd sess., Title I —Health Care Reform, Sec. 102: Changes to Existing Tax Preferences for Medical Coverage. 15. Office of the Chief Actuary, ‘‘Effects of the ‘Social Security Solvency Act of 2009’, S.426,’’ February 12, 2009, Payroll Tax Coverage of Employer Provided Group Health Insurance, p. 5. 16. Office of the Chief Actuary, ‘‘Effects of the ‘Social Security Personal Savings Guarantee Act of 2008,’’’ May 21, 2008, Effects on Trust Fund Assets and Unfunded Obligations, Table 1e: Revenue and Benefits from Coverage of Group Health Premiums. 17. Goss, ‘‘Effects of the ‘Social Security Solvency Act of 2009’, S.426,’’ February 12, 2009, Table 1b: Changes in OASDI Cash Flow, Bennett Proposal. 18. Office of the Chief Actuary, ‘‘Effects of the ‘Social Security Personal Savings Guarantee Act of 2008,’’’ May 21, 2008, Effects on Trust Fund Assets and Unfunded Obligations, Table 1c: OASDI Cash Flow to General Fund of the Treasury, Total 2008–82.
CHAPTER 7 1. Luiza Ch. Savage, ‘‘Hearing on Social Security Showcases Partisan Divide,’’ New York Sun, April 27, 2005, http://www.nysun.com/national/hearing-on-social-security-showcasespartisan/12879/ (accessed June 12, 2009).
180
NOTES
2. Center on Budget and Policy Priorities, Kris Cox and Richard Kogan, ‘‘Long-Term Social Security Shortfall Smaller than Cost of Extending Tax Cuts for Top 1 Percent,’’ March 31, 2008, http://www.cbpp.org/cms/index.cfm?fa=view&id=110. 3. The Board of Trustees, Federal Old-Age and Survivors Insurance and Federal Disability Insurance Trust Funds, The 2009 OASDI Trustees Report, Table IV.B6: Unfunded obligation for 1935 through 2083, Washington, DC: GPO, 2009, p. 63, http://www.ssa.gov/OACT/ TR/2009/IV_LRest.html#267528. 4. Social Security Administration, Office of the Chief Actuary, The 2009 OASDI Trustees Report, Supplemental Single-Year Tables, Operations of the Combined OASI and DI Trust Funds in Current Dollars, http://www.ssa.gov/OACT/TR/2009/lr6f8.html. 5. U.S. Government Accountability Office, ‘‘The Nation’s Long-Term Fiscal Outlook: March 2009 Update,’’ Data from Baseline Extended Simulation Based on Trustees’ Assumptions for Social Security and Medicare, http://gao.gov/special.pubs/longterm/march09baseline_extended_trustees.pdf. 6. Center on Budget and Policy Priorities, Cox and Kogan, ‘‘Long-Term Social Security Shortfall Smaller Than Cost of Extending Tax Cuts,’’ March 31, 2008. 7. Center on Budget and Policy Priorities, ‘‘Social Security and the Tax Cut: A Response to the Concord Coalition,’’ June 27, 2002, http://www.cbpp.org/cms/index.cfm?fa= view&id=2001. 8. Ceci Connolly and Jonathan Weisman, ‘‘The Choice for Voters: Health Care or Tax Cuts,’’ Washington Post, June 28, 2004, http://www.washingtonpost.com/wp-dyn/articles/ A10538–2004Jun27.html. 9. Author’s calculations based on Social Security Administration, Office of the Chief Actuary, ‘‘Long Range Solvency Provisions: Provisions Affecting OASDI Contribution and Benefit Base,’’ E1: Detailed single year tables, updated based on The 2009 OASDI Trustees Report, http://www.ssa.gov/OACT/solvency/provisions/tables/table_run132.html. 10. Urban Institute, Reform Options for Social Security: Raising the Taxable Maximum, http://www.urban.org/retirement_policy/sstaxableminimum.cfm. 11. Author’s estimate derived from Social Security Administration, Office of the Chief Actuary, ‘‘Long Range Solvency Provisions: Provisions Affecting OASDI Contribution and Benefit Base,’’ E4: Make 90 percent of the earnings subject to the payroll tax, Detailed single year tables. 12. Larry Rohter, ‘‘Obama Pulls Back on Social Security Plan,’’ The Caucus, New York Times, August 15, 2008, http://thecaucus.blogs.nytimes.com/2008/08/15/obama-pulls-backon-social-security-plan/?apage=2. 13. Ben Smith’s Blog, ‘‘A Tax Attack,’’ Politico, January 16, 2008. 14. Democratic Presidential Candidates Debate, October 30, 2007, MSNBC, Transcript: http://www.msnbc.msn.com/id/21562193/. 15. Jason Furman and Austan Goolsbee, ‘‘The Obama Tax Plan,’’ Wall Street Journal, August 14, 2008, http://online.wsj.com/public/article_print/SB121867201724238901.html. 16. Author’s estimate.
CHAPTER 8 1. Social Security Administration, Office of Retirement and Disability Policy, Annual Statistical Supplement, 2008, Benefits Awarded, Retired Workers, Table 6.B5: Number, average
NOTES
181
age, and percentage distribution, by sex and age, selected years 1940–2007, http://www .ssa.gov/policy/docs/statcomps/supplement/2008/6b.html#table6.b5. 2. Ibid. Calculation factors out disability conversions. 3. Social Security Administration, Office of the Chief Actuary, Effect of Early or Delayed Retirement on Retirement Benefits, http://www.ssa.gov/OACT/ProgData/ar_drc.html. 4. Social Security Administration, Online History, 1960s, http://www.ssa.gov/history/ 1960.html. (The 1961 legislation only changed the eligibility age for men; the earliesteligibility age for women was lowered to 62 in 1956.) 5. Social Security Administration, Office of the Chief Actuary, ‘‘Unisex Life Expectancies at Birth and Age 65,’’ June 2008, Table 1—Period Life Expectancies, http://www.ssa.gov/ OACT/NOTES/ran2/index.html. 6. U.S. Bureau of the Census, Statistical Brief, ‘‘Sixty-Five Plus in the United States,’’ May 1995, http://www.census.gov/population/socdemo/statbriefs/agebrief.html. 7. Social Security Administration, Office of the Chief Actuary, ‘‘Unisex Life Expectancies at Birth and Age 65,’’ June 2008, Table 1—Period Life Expectancies, http://www.ssa.gov /OACT/NOTES/ran2/index.html. 8. U.S. Bureau of the Census, Statistical Brief, ‘‘Sixty-Five Plus in the United States,’’ May 1995, http://www.census.gov/population/socdemo/statbriefs/agebrief.html. 9. Social Security Administration, Legislative History, ‘‘Summary of P.L.98-21, (H.R. 1900) Social Security Amendments of 1983-Signed on April, 20, 1983,’’ http://www.ssa.gov/ history/1983amend.html. 10. Social Security Administration, Office of Retirement and Disability Policy, ‘‘Income of the Population 55 or Older, 2006,’’ released in 2009, Table 11.1, 11.2, http://www.ssa.gov/ policy/docs/statcomps/income_pop55/2006/sect11.html#table11.1.
CHAPTER 9 1. Nancy Altman, ‘‘Bob Ball-Nancy Altman Social Security Reform Proposal,’’ http:// www.thebattleforsocialsecurity.com/media/pdf/ball-altman-plan.pdf. 2. The Board of Trustees, Federal Old-Age and Survivors Insurance and Federal Disability Insurance Trust Funds, The 2009 OASDI Trustees Report, Test of Long-Range Close Actuarial Balance, Washington, DC: GPO, 2009, p. 64-65. 3. Social Security Administration, Stephen C. Goss, Chief Actuary, ‘‘Estimated OASDI Financial Effects for a Proposal With Six Provisions That Would Improve Social Security Financing,’’ April 14, 2005, p.2, http://www.ssa.gov/OACT/solvency/RBall_20050414.pdf. 4. Ibid. 5. Ibid. 6. Thomas N. Bethell, ‘‘Future Shock: Some serious flaws found in latest proposals to fund Social Security,’’ AARP Bulletin Today, May 2005, http://bulletin.aarp.org/yourmoney/socialsecurity/articles/future_shock.html. 7. Federal Reserve Chairman Alan Greenspan, ‘‘On Investing the Social Security Trust Fund in Equities,’’ Testimony before the House Committee on Commerce, Subcommittee on Finance and Hazardous Materials, 106th Cong., 1st sess., March 3, 1999, http://www .federalreserve.gov/BOARDDOCS/TESTIMONY/1999/19990303.htm. 8. Social Security Administration, ‘‘Estimated OASDI Financial Effects for a Proposal With Six Provisions,’’ April 14, 2005, Sensitivity Analysis, p.5–6, http://www.ssa.gov/ OACT/solvency/RBall_20050414.pdf.
182
NOTES
9. The Office of the Chief Actuary had in 2008 estimated that a tax on all wages up to $213,600 would be needed to cover 90 percent of wages in 2009. The recession has brought that estimate down to $171,900. But because the reduction in high-end wages is seen as temporary, the author uses a figure near the mid-point as a rough gauge of future tax increases under this provision. 10. The Board of Trustees, The 2009 OASDI Trustees Report, Taxable Payroll and Payroll Tax Revenue, Assumptions & Methods, Washington, DC: GPO, 2009, p.114. 11. The net increase in benefits would be smaller because much of the benefits would be subject to income taxes. 12. Social Security Administration, Office of the Chief Actuary, ‘‘Long Range Solvency Provisions: Provisions Affecting OASDI Contribution and Benefit Base,’’ Summary Measures, E4 and E5, Long-range actuarial balance, http://www.ssa.gov/OACT/solvency/provisions/ wagebase_summary.html. Under Ball’s proposal, the increase in the taxable maximum would phase in more slowly, meaning somewhat less would be paid out in extra benefits to the highest earners by 2083. 13. Congressional Budget Office, ‘‘The Long-Term Budget Outlook,’’ December 2005, Option 3: Reduce Cost-of-Living Adjustments, p. 24, http://www.cbo.gov/ftpdocs/69xx/ doc6982/12-15-LongTermOutlook.pdf. 14. Kenneth J. Stewart, ‘‘The Experimental Consumer Price Index for Elderly Americans (CPI-E): 1982-2007,’’ Monthly Labor Review, Bureau of Labor Statistics, April 2008, p. 19, http://www.bls.gov/opub/mlr/2008/04/art2full.pdf.
CHAPTER 10 1. Social Security Administration, Stephen C. Goss, Chief Actuary, ‘‘Estimates of the Financial Effects for a Proposal to Restore Solvency to the Social Security Program,’’ October 8, 2003, Provisions 2 and 5, http://www.ssa.gov/OACT/solvency/DiamondOrszag _20031008.pdf. 2. As specified in their Social Security reform proposal, continued increases in benefit cuts and tax increases in the Diamond-Orszag plan were scheduled to continue indefinitely. An analysis of the proposal based on The 2009 OASDI Trustees Report shows that these measures could be frozen by 2065, while still leaving Social Security cash-flow positive through 2083. 3. Social Security Administration, Office of the Chief Actuary, ‘‘Long Range Solvency Provisions: Provisions Affecting Payroll Tax Rates,’’ Summary Measures, D1 and D2, http:// www.ssa.gov/OACT/solvency/provisions/payrolltax_summary.html. 4. Office of the Chief Actuary, Estimates of the Diamond-Orszag Proposal, Table 1: Projected OASDI Financial Status under Diamond-Orszag Proposal, October 8, 2003, http:// www.ssa.gov/OACT/solvency/DiamondOrszag_20031008.pdf, updated by author to reflect The 2009 OASDI Trustees Report. 5. Office of the Chief Actuary, Estimates of the Diamond-Orszag Proposal, October 8, 2003, Provisions 2 and 5, p. 2,3.
CHAPTER 11 1. Social Security Administration, ‘‘Annual Statistical Supplement, 2008,’’ Table 6.B5: Number, average age, and percentage distribution, by sex and age, selected years 1940–2007, http://www.ssa.gov/policy/docs/statcomps/supplement/2008/6b.html#table6.b5.
NOTES
183
2. Social Security Administration, Office of the Chief Actuary, ‘‘Unisex Life Expectancies at Birth and Age 65,’’ June 2008, Table 1—Period Life Expectancies, http://www.ssa.gov/ OACT/NOTES/ran2/index.html. 3. President’s Committee on Economic Security, The Program for Economic Security, ‘‘Need for Security,’’ January 1935, https://www.socialsecurity.gov/history/reports/ces/cesvol9 program.html. 4. C. Eugene Steuerle, ‘‘Alternatives to Strengthen Social Security,’’ Testimony before the House Committee on Ways and Means, 109th Cong., 1st sess., May 12, 2005, http:// taxpolicycenter.org/UploadedPDF/900806_Steuerle_051205.pdf. 5. Social Security Administration, Office of the Chief Actuary, The 2009 OASDI Trustees Report, Supplemental Single-Year Tables, Table IV, B1: Annual Income Rates, Cost Rates, and Balances, Calendar Years, 1970-2085, http://www.ssa.gov/OACT/TR/2009/lr4b1.html. 6. Federal Reserve Board Vice Chair Alice M. Rivlin, ‘‘Social Security,’’ Business Roundtable Social Security Symposium, April 6, 1999, http://www.federalreserve.gov/Boarddocs/ Speeches/1999/19990406.htm. 7. Social Security Administration, Office of Retirement and Disability Policy, Annual Statistical Supplement, 2008, Benefits in Current-Payment Status, Table 5A1.1: Number and average monthly benefit for retired workers, by sex, age, and race, December 2007, http://www .ssa.gov/policy/docs/statcomps/supplement/2008/5a.html#table5.a1.1. 8. David M. Cutler, Jeffrey B. Liebman, and Seamus Smyth, ‘‘How Fast Should the Social Security Retirement Age Rise?’’ Harvard University and the National Bureau of Economic Research, April 28, 2006, p. 12, 13, http://www.nber.org/programs/ag/rrc/04-05 CutlerLiebmanSummary.pdf. 9. Richard V. Burkhauser, Kenneth A. Couch, and John W. Phillips, ‘‘Who Takes Early Social Security Benefits? The Economic and Health Characteristics of Early Beneficiaries,’’ Gerontologist, vol. 36, no. 6 (1996), pp. 789–799. 10. Government Accountability Office, ‘‘Older Workers: Demographic Trends Pose Challenges for Employers and Workers,’’ November 2001, p. 34, http://www.gao.gov/new.items/ d0285.pdf. 11. Social Security Administration, Office of Retirement and Disability Policy, ‘‘Income of the Population 55 or Older, 2006,’’ released in 2009, Table 9.A1: Relative Importance of Social Security for Beneficiary Aged Units, http://www.ssa.gov/policy/docs/statcomps/ income_pop55/2006/sect09.html#table9.a1. 12. Alicia H. Munnell and Steven A. Sass, ‘‘The Labor Supply of Older Americans,’’ Center for Retirement Research at Boston College, June 2007, p. 4, http://escholarship.bc .edu/cgi/viewcontent.cgi?article=1156&context=retirement_papers. 13. Barbara A. Butrica, Karen E. Smith, C. Eugene Steuerle, ‘‘Working for a Good Retirement,’’ The Urban Institute, May 2006, p. 26, Table 4, http://www.urban.org/UploadedPDF/ 311333_good_retirement.pdf. Findings adapted based on comparison to The 2009 Trustees Report. 14. Congressional Budget Office, ‘‘Growing Disparities in Life Expectancy,’’ April 17, 2008, http://www.cbo.gov/ftpdocs/91xx/doc9104/04-17-LifeExpectancy_Brief.pdf.
CHAPTER 12 1. Andrew J. Rettenmaier and Thomas R. Saving, National Center for Policy Analysis, ‘‘Social Security Reform without Illusion: The Five Percent Solution,’’ Table 2, p. 9, December
184
NOTES
17, 2004, http://www.ncpa.org/pdfs/st272.pdf. The authors proposed a 3.5-percentage-point increase in the payroll-tax rate—the equivalent of nearly $200 billion in 2009—but their plan still counted on depleting the multi-trillion-dollar trust fund. 2. Federal Reserve Board Vice Chair Alice M. Rivlin, ‘‘Social Security,’’ Business Roundtable Social Security Symposium, April 6, 1999, http://www.federalreserve.gov/Boarddocs/ Speeches/1999/19990406.htm. 3. Ibid. 4. Ibid. 5. Martin Neil Baily and Jacob Funk Kirkegaard, U.S.Pension Reform: Lessons from Other Countries, Washington, DC, Peterson Institute for International Economics: 2009, p. 450. 6. Board of Trustees of the Federal Old-Age and Survivors Insurance and Federal Disability Insurance Trust Funds, Report of the 1994–1996 Advisory Council on Social Security, January 1997, http://www.ssa.gov/history/reports/adcouncil/report/findings.htm#overview. 7. U.S. Senate Roll Call Votes, 106th Cong., 1st sess., Vote Summary on Ashcroft Amendment 145 to S.Con.Res. 20, ‘‘To express the sense of the Senate that the Federal Government should not directly invest the Social Security trust funds in private financial markets,’’ March 24, 1999, http://www.senate.gov/legislative/LIS/roll_call_lists/roll_call_vote_cfm.cfm? congress=106&session=1&vote=00060. 8. Brian K. Bucks and others, ‘‘Changes in U.S. Family Finances from 2004 to 2007: Evidence from the Survey of Consumer Finances,’’Federal Reserve Bulletin, vol. 95, February 2009, Table 6.B., http://www.federalreserve.gov/pubs/bulletin/2009/pdf/scf09.pdf. 9. See, for example, Michael Sherradden PhD, director of the Center for Social Development at Washington University in St. Louis, ‘‘Assets and the Poor: Implications for Individual Accounts in Social Security,’’ testimony to the President’s Commission to Strengthen Social Security, October 18, 2001, p.144, http://govinfo.library.unt.edu/csss/meetings/transcripts/ October_18_transcript.pdf. 10. See, for example, Department of the Treasury, General Explanations of the Administration’s Fiscal Year 2010 Revenue Proposals, ‘‘Expand the Saver’s Credit and Provide for Automatic Enrollment in IRAs,’’ May 2009, p. 6–8 and Table 1: Revenue Estimates of FY 2010 Budget Proposals, http://www.treas.gov/offices/tax-policy/library/grnbk09.pdf. 11. Baily and Kirkegaard, U.S.Pension Reform: Lessons from Other Countries, p. 277. 12. See, for example, Bootie Cosgrove-Mather, ‘‘Poll: Support Sags for SS Reform,’’ CBS News, March 2, 2005, http://www.cbsnews.com/stories/2005/03/02/opinion/polls/ main677680.shtml.
CHAPTER 13 1. Jeffrey Liebman, Maya MacGuineas, and Andrew Samwick, ‘‘Nonpartisan Social Security Reform Plan,’’ December 14, 2005, http://www.hks.harvard.edu/jeffreyliebman/ lms_nonpartisan_plan_description.pdf. 2. The Office of the Chief Actuary had in 2008 estimated that a tax on all wages up to $213,600 would be needed to cover 90 percent of wages in 2009. The recession has brought that estimate down to $171,900. But because the reduction in high-end wages is seen as temporary, the author uses a figure near the mid-point as a rough gauge of future tax increases under this provision. 3. Social Security Administration, Office of the Chief Actuary, ‘‘Estimated Financial Effects of ‘A Nonpartisan Approach to Reforming Social Security–A Proposal Developed by
NOTES
185
Jeffrey Liebman, Maya MacGuineas and Andrew Samwick,’’’ November 17, 2005, p.3, http:// www.ssa.gov/OACT/solvency/Liebman_20051117.pdf. 4. The Office of the Chief Actuary’s analysis put the revenue from raising the maximum taxable earnings level at 1 percent of taxable payroll. Add in the 1.5 percent additional tax on all workers, and the combined revenue would equal 2.5 percent of payroll, or about 92 percent of the full shortfall equal to 2.7 percent. 5. Jeffrey Liebman, Maya MacGuineas, and Andrew Samwick, ‘‘Nonpartisan Social Security Reform Plan,’’ December 14, 2005, p. 2. 6. House Committee on Ways and Means Chairman Charles Rangel, opening remarks, hearing on the President’s fiscal year 2008 budget, February 6, 2007, 11th Cong., 1st sess., http://waysandmeans.house.gov/news.asp?formmode=release&id=475. 7. Social Security Administration, Office of the Chief Actuary, The 2009 OASDI Trustees Report, Supplemental Single-Year Tables, Components of Annual Income Rates, Calendar Years 2009-85, OASDI Taxation of benefits, http://www.ssa.gov/OACT/TR/2009/lr4b10 .html. 8. ‘‘Social Security Reform: A Bipartisan Proposal,’’ American Enterprise Institute for Public Policy Research, June 19, 2006, .http://www.aei.org/video/100584. 9. Jeffrey Liebman, Maya MacGuineas, and Andrew Samwick, ‘‘Nonpartisan Social Security Reform Plan,’’ December 14, 2005, p. 6. 10. Ibid.
CHAPTER 14 1. ‘‘Social Security Reform: A Bipartisan Proposal,’’ American Enterprise Institute for Public Policy Research, June 19, 2006, http://www.aei.org/video/100584.
CHAPTER 15 1. Depending on how fast the change occurred, raising the retirement age to 70 could come well short of closing the other half of Social Security’s financing gap. Increasing the Normal Retirement Age by 1 month every 2 years until it hits 70 could solve less than one-fourth of the full shortfall, according to this analysis: Social Security Administration, Office of the Chief Actuary, ‘‘Long Range Solvency Provisions: Provisions Affecting Retirement Age,’’ C3, Summary Measures and Graphs, http://www.ssa.gov/OACT/solvency/provisions/charts/ chart_run213.html. 2. Social Security Administration, Office of the Chief Actuary, ‘‘Long Range Solvency Provisions: Provisions Affecting Retirement Age,’’ http://www.ssa.gov/OACT/solvency/ provisions/retireage.html. 3. Congressional Research Service, Geoffrey Kollman, ‘‘Social Security: Raising the Retirement Age Background and Issues,’’ updated June 7, 2000, Table 2, http://digital .library.unt.edu/govdocs/crs/permalink/meta-crs-1299:1. 4. Congressional Research Service, ‘‘Raising the Retirement Age Background and Issues,’’ updated June 7, 2000, Legislative Proposals in the 104th Congress. 5. Social Security Personal Savings Guarantee and Prosperity Act of 2008,’’’ May 21, 2008, Plan Specification, p. 4, http://www.ssa.gov/OACT/solvency/PRyan_20080521.pdf. 6. In practice, there can be a substantive difference between indexing for longevity and explicitly raising the Normal Retirement Age. That is because benefits for the disabled aren’t impacted by increases in the NRA but could be by longevity indexing.
186
NOTES
7. Barbara A. Butrica, Karen E. Smith, C. Eugene Steuerle, ‘‘Working for a Good Retirement,’’ The Urban Institute, May 2006, p. 26, Table 4, http://www.urban.org/ UploadedPDF/311333_good_retirement.pdf. 8. Years of highest earnings aren’t determined by the nominal level of wages, but by a worker’s wage level in relation to an index of national wage levels. For further explanation, see Social Security Administration, Office of the Chief Actuary, ‘‘Automatic Increases: Average Indexed Monthly Earnings,’’ http://www.ssa.gov/OACT/COLA/Benefits.html#aime. 9. Social Security Administration, Office of the Chief Actuary, ‘‘Long Range Solvency Provisions: Provisions Affecting Level of Monthly Benefits,’’ B2, http://www.ssa.gov/OACT/ solvency/provisions/benefitlevel.html. 10. C. Eugene Steuerle, ‘‘Alternatives to Strengthen Social Security,’’ Testimony before the House Committee on Ways and Means, 109th Cong., 1st sess., May 12, 2005, http:// taxpolicycenter.org/UploadedPDF/900806_Steuerle_051205.pdf. 11. Social Security Administration, Office of the Chief Actuary, ‘‘Estimated OASDI Financial Effects of the ‘Retirement Security Act,’’ February 11, 2004, p. 8-9, http://www.ssa.gov/ OACT/solvency/Kolbe_20040211.pdf. 12. Ibid. 13. It would be possible to avoid any weakening of the safety net in very old age in tandem with an increase in the number of earnings years factored into the benefit formula. In his 2005 testimony, Steuerle suggested: ‘‘Whatever the level of lifetime benefit that is settled upon in a final reform package, actuarial adjustments can provide more benefits later and fewer earlier.’’ 14. Social Security Administration, Office of the Chief Actuary, ‘‘Estimates of Financial Effects for Three Models Developed by the President’s Commission to Strengthen Social Security,’’ Model 3 Specifications, p. 11, January 31, 2002, http://www.ssa.gov/OACT/solvency/ PresComm_20020131.pdf. 15. Half of the delayed retirement credit, 5 percentage points, would be awarded based on the number of years worked in five-year period ending with the year of delayed retirement.
CHAPTER 16 1. Social Security Administration, Office of the Chief Actuary, The 2009 OASDI Trustees Report, Supplemental Single-Year Tables, Table VI.F10—Annual Scheduled Benefit Amounts for Retired Workers with Various Pre-Retirement Earnings Patterns, http://www.ssa.gov/ OACT/TR/2009/lr6f10.html. 2. Congressional Budget Office, ‘‘Is Social Security Progressive?’’ December 15, 2006, p. 6, http://www.cbo.gov/ftpdocs/77xx/doc7705/12-15-Progressivity-SS.pdf. 3. Social Security Administration, Office of the Chief Actuary, ‘‘Internal Real Rates of Return under the OASDI Program for Hypothetical Workers,’’ April 2009, Table 3: Payable Benefits Scenario, http://www.ssa.gov/OACT/NOTES/ran5/index.html. 4. This standard is derived from Social Security Administration, Office of the Chief Actuary, ‘‘Estimated Financial Effects of ‘A Nonpartisan Approach to Reforming Social Security,’ ’’ November 17, 2005, Table B1a, http://www.ssa.gov/OACT/solvency/Liebman_20051117 .pdf. The table shows, for example, that an $18,900-earner who saves 1 percent of earnings over a full career could overcome a 7.6 percent benefit cut, while a $67,200-earner could overcome a 12.4 percent benefit cut by saving 1 percent of income.
NOTES
187
5. Author’s calculation 6. The Board of Trustees, Federal Old-Age and Survivors Insurance and Federal Disability Insurance Trust Funds, The 2009 OASDI Trustees Report, Average Earnings Assumptions, Washington, DC: GPO, 2009, p. 95, http://www.ssa.gov/OACT/TR/2009/V_economic .html#131323. 7. The Henry J. Kaiser Family Foundation, ‘‘Health Care Costs: A Primer,’’ March 2009, p. 4, 13, http://www.kff.org/insurance/upload/7670_02.pdf. 8. Joint Committee on Taxation,Tax Expenditures for Health Care (JCX-66-08), Senate Committee on Finance, 110th Cong., 2nd sess., July 31, 2008, p. 12, http://finance.senate .gov/hearings/testimony/2008test/073108ektest.pdf 9. Jason Furman, ‘‘Health Reform through Tax Reform: A Primer,’’ Health Affairs, Volume 27, Number 3, May/June 2008, p. 624, http://www.brookings.edu/~/media/Files/rc/articles/ 2008/05_health_reform_furman/05_health_reform_furman.pdf 10. Social Security Administration, Office of the Chief Actuary, ‘‘Long Range Solvency Provisions: Provisions Affecting Coverage of Employment or Earnings,’’ F2, Summary Measures, http://www.ssa.gov/OACT/solvency/provisions_tr2008/charts/chart_run199 .html#graph. 11. Office of the Chief Actuary, ‘‘Effects of the ‘Social Security Personal Savings Guarantee Act of 2008,’ ’’ May 21, 2008, Effects on Trust Fund Assets and Unfunded Obligations, Table 1e: Revenue and Benefits from Coverage of Group Health Premiums, http://www.ssa.gov/ OACT/solvency/PRyan_20080521.pdf. 12. Ibid. 13. For a partial discussion on how limits can be placed on tax-free, employer-provided health care, see: Len Burman, ‘‘Caps and ‘Gold Plated’ Health Plans, TaxVox: The Tax Policy Center blog, July 31, 2009, http://taxvox.taxpolicycenter.org/blog/_archives/2009/7/31/ 4272789.html. 14. Author’s calculation based on Social Security Administration, Office of the Chief Actuary, ’’Long Range Solvency Provisions: Provisions Affecting Coverage of Employment or Earnings,’’ F2, Detailed Single Year Tables, http://www.ssa.gov/OACT/solvency/provisions _tr2008/tables/table_run199.html. 15. Based on author’s calculations. 16. Author’s calculations based on Social Security Administration, Office of the Chief Actuary, ‘‘Provisions Affecting OASDI Contribution and Benefit Base,’’ E6: Detailed single year tables, updated based on The 2009 OASDI Trustees Report, http://www.ssa.gov/OACT/ solvency/provisions/tables/table_run249.html.
CHAPTER 17 1. Butrica, Smith, and Steuerle, ‘‘Working for a Good Retirement,’’ The Urban Institute, May 2006, p. 3, http://www.urban.org/UploadedPDF/311333_good_retirement.pdf. 2. Under Social Security’s retirement earnings test, workers who remain in the workforce between 62 and the Normal Retirement Age can receive their full benefit if they make no more than $14,160 in 2009. 3. Butrica, Smith, and Steuerle, ‘‘Working for a Good Retirement,’’ The Urban Institute, May 2006, p. 19.
188
NOTES
CHAPTER 18 1. Under this original design of A Well-Tailored Safety Net, the carve-out portion would have been put back, so to speak, meaning it would have been cashed in for Social Security’s traditional annuity upon retirement. 2. Social Security Administration, Office of the Chief Actuary, The 2009 OASDI Trustees Report, Supplemental Single-Year Tables, OASDI and HI Annual Income Excluding Interest, Cost, and Balance in current dollars, http://www.ssa.gov/OACT/TR/2009/lr6f9.html. 3. Social Security Administration, ‘‘Survivors Benefits,’’ SSA Publication No. 05–10084, January 2009, p. 5, http://www.ssa.gov/pubs/10084.pdf. 4. Weekly Compilation of Presidential Documents, Volume 41, Number 2, ‘‘Remarks in a Discussion on Social Security Reform,’’ January 11, 2005, Washington, DC: GPO, 2005, p. 42, http://www.gpo.gov/fdsys/pkg/WCPD-2005-01-17/pdf/WCPD-2005-01-17.pdf. 5. David Kamin and Jason Furman, Center on Budget and Policy Priorities, ‘‘Social Security and Inheritance: The Dubious Promise of Private Accounts,’’ May 3, 2006, http:// www.cbpp.org/files/5-3-06socsec.pdf. 6. See, for example, ‘‘EPI Issue Guide: Social Security,’’ Economic Policy Institute, Racial differences in life expectancy, last updated May 2005, http://epi.3cdn.net/4f894bf3eec6822 dec_a5m6brb5d.pdf. 7. See, for example, David Walker, interview for ‘‘Ten Trillion and Counting,’’ Frontline, PBS, March 24, 2009, http://www.pbs.org/wgbh/pages/frontline/tentrillion/interviews/ walker.html#3. 8. Peter A. Diamond and Peter R. Orszag, Saving Social Security: A Balanced Approach,’’ Washington, DC, Brookings Institution Press: 2004, p. 154. 9. If these accounts were linked to a new universal payroll-deduction retirement savings program that has bipartisan support, then efficient investment in a broader range of investments might be possible, despite the small size of the add-on accounts in A Well-Tailored Safety Net. 10. Social Security Administration, Felicitie C. Bell and Michael L. Miller, Life Tables for the United States Social Security Area 1900-2100, August 2005, Figure 5—Survival Function for SSA Population, http://www.ssa.gov/OACT/NOTES/s2000s.html. 11. Social Security Administration, Office of Retirement and Disability Policy, Annual Statistical Supplement, 2008, Benefits Awarded, Retired Workers, Table 6B5: Number, average age, and percentage distribution, by sex and age, selected years 1940-2007, http://www.ssa .gov/policy/docs/statcomps/supplement/2008/6b.html#table6.b5. 12. Social Security Administration, Office of the Chief Actuary, ‘‘Estimates of Financial Effects for Three Models Developed by the President’s Commission to Strengthen Social Security,’’ Model 3 Specifications, p. 11, January 31, 2002, http://www.ssa.gov/OACT/solvency/ PresComm_20020131.pdf. 13. A retiree who works all five years from age 65 through 69 would receive a 10 percent delayed retirement credit, with 5 percent going toward the lump sum and 5 percent going toward a bigger traditional benefit. A retiree who only works four of those years would still get a 5 percent bigger traditional benefit, but a proportionally smaller lump-sum payment.
NOTES
189
CHAPTER 19 1. House Majority Leader Steny Hoyer, keynote address at Bipartisan Policy Center forum, ‘‘Unprecedented Federal Debt: Putting Our Fiscal House in Order,’’ May 6, 2009, 111th Cong., 1st sess., http://www.hoyer.house.gov/newsroom/index.asp?ID=1383. 2. Estimate based on author’s calculations reflecting The 2009 OASDI Trustees Report to update Office of the Chief Actuary solvency analysis of Bennett proposal from February 12, 2009. 3. Estimate based on author’s calculations reflecting The 2009 OASDI Trustees Report to update Office of the Chief Actuary solvency analysis of Diamond-Orszag proposal from October 8, 2003. 4. The 2009 OASDI Trustees Report, Table VI, F4, OASDI and HI Annual and Summarized Income, Cost, and Balance as a Percentage of GDP, p. 184, http://www.ssa.gov/ OACT/TR/2009/VI_OASDHI_GDP.html#147880. 5. Estimate based on author’s calculations reflecting The 2009 OASDI Trustees Report to update Office of the Chief Actuary solvency analysis of Bennett proposal from February 12, 2009. 6. Estimate based on author’s calculations reflecting The 2009 OASDI Trustees Report to update Office of the Chief Actuary solvency analysis of LMS proposal from November 17, 2005. 7. Social Security Administration, Office of the Chief Actuary, ‘‘Estimated Financial Effects of ‘A Nonpartisan Approach to Reforming Social Security,’ ’’ November 17, 2005, Table B1c., http://www.ssa.gov/OACT/solvency/Liebman_20051117.pdf. 8. For a discussion, see Prieur du Plessis, ‘‘Words from the (investment wise) for the week that was.’’ Investment Postcards from Cape Town, May 24, 2009, http:// www.investmentpostcards.com/2009/05/24/words-from-the-investment-wise-for-the-weekthat-was-may-18-%E2%80%93-24-2009/. 9. Author’s estimate. 10. U.S. Senate, introduction of S.2427, ‘‘The Sustainable Solvency for Social Security Act,’’ 109th Cong., 2nd sess., Congressional Record (March 16, 2006): S2318. 11. Diamond and Orszag, Saving Social Security: A Balanced Approach, p. 110. 12. Office of the Chief Actuary solvency analysis of Diamond-Orszag proposal from October 8, 2003, p. 6. 13. Jason Furman, ‘‘An Analysis of Using ‘Progressive Price Indexing’ to Set Social Security Benefits,’’ Center on Budget and Policy Priorities, May 2, 2005. 14. Elmendorf, Liebman, Wilcox, ‘‘Fiscal Policy and Social Security Policy During the 1990s,’’ 2002, p. 91; Peter Wallsten and Joel Havemann, ‘‘Bush Shifts Pension Stance,’’ Los Angeles Times, February 17, 2005, http://articles.latimes.com/2005/feb/17/nation/nasocial17. 15. Author’s estimate.
CONCLUSION 1. See, for example, Felix Salmon, ‘‘Replacing Social Security with Carbon Taxes,’’ Reuters Blogs, March 16, 2009, http://blogs.reuters.com/felix-salmon/2009/03/16/replacing-socialsecurity-with-carbon-taxes/.
190
NOTES
2. Kim Chipman and Catherine Dodge, ‘‘Obama Plan Has $79 Billion From Cap-andTrade in 2012,’’ Bloomberg, http://www.bloomberg.com/apps/news?pid=20601130&sid= aAO_KEIgeOOc. 3. Bootie Cosgrove-Mather, ‘‘Poll: Support Sags for SS Reform,’’ CBS News, March 2, 2005, http://www.cbsnews.com/stories/2005/03/02/opinion/polls/main677680.shtml.
Index
AARP, 20 add-on accounts, 13, 133–34, 136–38, 159–60, 166t, 165t. See also carve-out plans; notional accounts; personal accounts aging population: under Ball’s reform plan, 64–65, 65t; benefits cuts, 94; benefits of working longer, 76–78; better health of, 75; under Bush’s reform plans, 47; dependency ratio, 5; employer incentives to employ older workers, 128–29, 128t, 165–66; income security for, 7–8, 27, 43, 65, 100, 105, 160–61; life expectancies, 59f, 73–74, 74t119–20, 1560–61; and longevity indexing, 49– 52, 75, 100–101, 157; poverty rates, 9, 59; projected growth of, 5, 7; safety net for, 7–12, 40, 57–60, 64, 75–76, 93–96, 100. See also Old-Age Risk-Sharing Altman, Nancy J., 61 American Association of Retired People (AARP), 20 baby boomers, retirement of, 5, 73, 76, 77f Baily, Martin N., 82 Ball, Robert, 61, 118
Ball’s Social Security reform plan, 61–65, 65t, 118, 158 base benefit (Primary Insurance Amount), 89, 104, 139, 158, 175 n.14 The Battle for Social Security (Altman),61 benefits cuts: for the aging population, 94; under Bennett’s reform plan, 50t, 75; under Diamond-Orszag reform plan, 69–72, 70t, 180n2; under LMS reform plan, 85–86, 90–91, 116t; phasing out of, 105–6, 105t; progressive approach, 95 benefits formulas: diminishing rate of return,112f; and employer-provided health care benefits,118–19; inequitable distribution under LMS reform plan, 90–91, 90t; progressive nature of, 111– 15, 113f; shift to 40 year formula, 101– 3, 102t; in Social Security reform plans,157 Bennett, Robert, 49, 113–14 Bennett’s Social Security reform plan: benefits cuts, 50t, 75; and burdens of sacrifice, 114, 114t; disability benefits, 150–51, 150t; effect on government debt, 146–47, 147f, 157–58; effect on
192
INDEX
Social Security Trust Fund, 143–45; and income security, 136–38; longevity indexing, 49–52;safety net, 50–51, 147, 148t; taxes on employer-provided health benefits, 51 Boxer, Barbara, 15 budget. See debt, government;deficits, budget;surpluses, budget burdens of sacrifice: balancing of, 115–16; under Bennett’s reform plan, 114, 114t; under Bush’s reform plans, 46, 114; covered workers, increasing numbers of, 107; under Diamond-Orszag reform plan, 70–71; and fair Social Security reform, 8; on highest earners, 94; increasing over the years, 144; under LMS reform plan, 115–16, 115t, 116t; and progressive saving offset, 121t; under Ryan’s 2008 reform plan, 52; under A Well-Tailored Safety Net, 121–23, 121t, 153–54, 153t Bush, George W.: and personal accounts, 41; Social Security reform plans, 6, 43– 47, 46t, 103, 113–14; on Social Security Trust Fund, 23, 27 cap-and-trade revenues, 159 carbon tax revenues, 159 carve-out plans, 87–88, 131–36, 133t, 160. See also Delayed Retirement Accounts cash flow deficit, 25 CBO (Congressional Budget Office): on government debt, 4–5, 19, 28; on growth of Social Security costs, 28; long-term budget outlook, 29, 30f Certner, David, 90 Clinton, Bill, 19, 24, 31 Clinton, Hillary, 54, 56 COLA (Cost of Living Adjustment), 62, 64–65, 65t Congressional Budget Office. See CBO Conrad, Kent, 22 Cooper, Jim, 19–20 corporate taxes, 38–39 Cost of Living Adjustment (COLA), 62, 64–65, 65t
covered workers,increasing numbers of, 107, 159, 167 debt, government, 4–5, 19, 25, 28, 29f, 145–46, 147f, 157–58. See also interest, on government debt; Social Security Trust Fund deficits, budget, 23–25 delayed retirement: incentives for, 8, 82, 101–5, 138–41, 140t, 186n13; lump-sum transfers, 13, 135, 139–41, 160, 167; under Old-Age Risk-Sharing, 108, 126–27; and progressive saving offset, 140t;as route to ownership, 13; under A Well-Tailored Safety Net,13, 135, 139–41, 165–67, 166t, 167t Delayed Retirement Accounts, 13, 135–41, 137t, 140t, 166–67, 167t DeMint, Jim, 27, 37 Democrats: and LMS reform plan, 85–89; and Medicare, 31; modified LMS reform plan, 153–54, 154t; and personal accounts, 49, 82–83, 87–88; and planned tax cuts, 114; repeal of Bush’s tax cuts, 53–54; and tax increases, 6, 19, 21–22, 53; and A Well-Tailored Safety Net, 23, 27 dependency ratio, 5 Diamond, Peter, 69 Diamond-Orszag Social Security reform plan: benefits cuts, 69–72, 70t, 180n2; and burden of sacrifice, 70–71; and disability benefits, 149–51, 151t; and early retirement, 71; effect on government debt, 145–46, 147f; effect on Social Security Trust Fund, 144–45; payroll tax increases, 69–72, 74, 80, 144, 147–48, 156; safety net, 71, 148–49, 149t disability benefits: Disability Insurance Trust Fund, 20, 40, 134; under Social Security reform plans, 50–51, 50t, 150– 51, 151t, 168, 168t early-retirement: under Diamond-Orszag reform plan, 71; discouragement of, 11, 76–77; encouraged by raising payroll taxes, 82; and income security in old age, 26, 65, 65t, 100, 104t, 127; minimum
INDEX
age increase, 11, 86, 94–95; not sustainable, 125; penalties for, 7–8, 51, 57–60, 60t, 95, 100–104, 104t; voluntary delay of, 101; under A Well-Tailored Safety Net, 126–27, 127t, 165, 166t economic growth and Social Security reform, 76, 77f, 79–80, 85 Edwards, John, 54, 56 Elmendorf, Douglas, 32 employer incentives to employ older workers, 128–29, 128t, 165 employer-provided health benefits, 8, 12, 51–52, 117–22, 159, 164 estate taxes, 32, 61–63 Ferrara, Peter, 39 fiscal responsibility, 28 flexible hours for older workers, 128–29, 128t, 165–66 40 year formula, 101–3, 102t Furman, Jason, 117, 152 GAO (Government Accountability Office), 4, 54, 169nn3–4 GDP (Gross Domestic Product): percent of Social Security gap, 174n19; percent spent on debt, 28, 144; percent spent on entitlement programs, 28–29, 31–32, 55f; taxes as share of, 29–30 Government Accountability Office (GAO), 4, 54, 169nn3–4 government bonds, 7, 13, 41, 135–37, 157, 160, 166 government debt. See debt, government Gramlich, Edward, 21 Greenspan, Alan, 39–40 Greenstein, Robert, 5, 54 Gross Domestic Product. See GDP health care reform, 4, 47, 51, 117–22. See also employer-provided health benefits Heinz, John, 21 highest earners, 64, 94, 107–9, 108t, 111–15 Hilda and David test, 93–94 Hoyer, Steny, 6, 143 Hunter, Larry, 40
193
incentives: for delayed retirement, 8, 82, 101–5, 138–41, 140t, 186n13; for employers to employ older workers, 128–29, 128t, 165–66 income security: under Bennett’s reform plan, 136–38; and early retirement, 26, 65, 65f, 100, 104t, 127; and increasing life expectancies, 160–61; Social Security promise of, 75, 83, 125–26; for very old age, 7–8, 27, 43, 65, 100, 105; under A Well-Tailored Safety Net, 136–38, 160– 61. See also safety net income taxes, 71. See also payroll taxes inflation, 45, 47, 64 inheritibility of personal accounts: to encourage delayed retirement, 101; under LMS reform plan, 87; and lower-earners, 9–10; under A WellTailored Safety Net, 13, 134–37, 160 interest: on government debt, 3–5, 23, 28, 29f, 32, 37; on personal accounts, 135, 140; on the Social Security Trust Fund IOUs, 7, 16–17, 18t, 32 ‘‘Irene and Bernie’’ tests, 93–94 Kerry, John, 53–54 Kirkegaard, Jacob F., 82 Kogan, Richard, 54 Kolbe, Jim, 102 Kolbe-Stenholm plan, 102–4 Leonard, Herman B., 21 Liebman, Jeffrey, 32, 85, 91, 95–96, 118 Liebman-MacGuineas-Senwick Social Security reform plan (LMS): benefits cuts, 85–86, 90–91, 116t; and burdens of sacrifice, 115–16, 115t, 116t; and carve-out plans, 86–88; and disability benefits, 150–51; effect on government debt, 145–46, 147f; effect on Social Security Trust Fund, 145–46; modification of, 153–54, 154t; and Normal Retirement Age, 85–86; and payroll taxes, 86, 91, 91t, 115t, 147–48, 158; and personal accounts, 86–91, 88t, 90t; safety net, 148–49, 149t; and taxes on benefits, 88
194
INDEX
life expectancies: and income-security, 160–61; increases in, 59f, 160–61; and later Normal Retirement Age, 119–20; of lower-income groups, 78, 95; and Old-Age Risk-Sharing, 104–9; and retirement age, 59f, 73–74, 74t, 119–20 LMS reform plan. See Liebman-MacGuineas-Senwick Social Security reform plan longevity indexing, 49–52, 75, 100–101, 156 lower-earners: hurt by 40-year benefits formula, 102; life expectancies of, 78, 95; need for safety net, 94; under Old-Age Risk-Sharing, 105–6, 106t; and personal accounts, 9–10, 81–83, 137t; progressive nature of benefits, 111–15, 113f lower-intensity work for older workers, 128–29, 165–66 lump-sum transfers, 13, 135, 139–41, 160, 167 MacGuineas, Maya, 85–86 Making Work Pay tax credit, 159 Mankiw, Greg, 47 Medicaid, 5–6, 28, 31 medical insurance. See health care reform Medicare, 5–6, 28, 31 middle class, 64, 94, 118, 152, 152t minimum benefits, 102, 107t, 126 minor children, benefits for, 9, 20, 40, 87, 134 Moore, Stephen, 39–40 Moynihan, Daniel Patrick, 21–22 Munnell, Alicia, 77 Normal Retirement Age (NRA): increase in, 11, 85–86, 184n1; and increasing life expectancies, 59f, 73–74, 74t, 119– 20; under LMS reform plan, 85–86; under Old-Age Risk-Sharing, 109, 164– 65, 165t; Social Security benefits levels, 57–60, 58t; under A Well-Tailored Safety Net, 119–20, 126–27, 127t, 165, 166t notional accounts, 135, 160 NRA. See Normal Retirement Age
Obama, Barack: 2009 budget, 19, 28; on repeal of Bush’s tax cuts, 54; on Social Security reform, 3, 6; on tax increases for the wealthy, 56; on taxing over the current ceiling, 56 Old-Age Risk-Sharing: and delayed retirement, 108, 126–27; and life expectancies, 104–9; and lower-earners, 105–6, 106t; and Normal Retirement Age (NRA), 109, 164–65, 165t; and progressive saving offset, 103; and safety net preservation, 157; under Social Security reform plans, 11, 104–9, 126– 27, 145, 157, 164, 165t; summary of, 164, 165t Orszag, Peter, 31, 54, 69, 118 ownership incentives, 9–10, 13, 87, 133, 138, 159, 167 Palmer, John, 19 pay-as-you-go systems, 21–22, 40–41, 44 payroll taxes: and benefits of working longer, 77–78; to cover funding gap, 33– 34, 34f; under Diamond-Orszag reform plan, 69–72, 74, 80, 144, 147–48, 158; increases unavoidable, 81; under LMS reform plan, 86, 91, 91t, 115f, 147, 158; as percent of scheduled benefits, 74t; and personal accounts, 134; progressive approach to increases, 95; reductions for older workers, 12; regressive nature of, 111, 115t, 116t; taxing over the current ceiling, 12, 55–56, 62–64, 82, 86, 158; under A Well-Tailored Safety Net, 120– 21, 120t, 147 Pelosi, Nancy, 17 personal accounts: benefits of, 87; Bush on, 41; and delayed retirement, 101; Delayed Retirement Accounts, 13, 135– 41, 137t, 140t, 166–675, 167t; and government bonds, 7, 13, 41, 135–37, 157, 160, 166; under LMS reform plan, 86–91, 88t, 90t; and lower earners, 9– 10, 81–83, 137t; and payroll taxes, 134; progressive saving offset, 127–28, 128t; under the Ryan-Sununu plan, 37–42; security of, 136–38; and Social Security
INDEX
reform plans, 159–60; and stock market instability, 47; summary of, 166–67, 167t; under A Well-Tailored Safety Net, 120–21, 121t phased-in benefits, 12, 163, 164t poverty rate among older Americans, 9, 59 poverty-related minimum benefits, 12 pre-funding of future benefits, 86 Primary Insurance Amount (base benefit), 89, 104, 139, 158, 175n14 progressive ownership rationale for add-on accounts, 133–34 progressive price indexing, 113, 113t, 157 progressive saving offset: burden of sacrifice, 121t; and delayed retirement, 140t; and disability benefits, 168t; modification of, 122–23, 122t; and Old-Age Risk-Sharing, 103; under A Well-Tailored Safety Net, 12, 121–22, 127–28, 128t, 163, 164t. See also personal accounts public debt. See debt, government Rangel, Charles, 88 reform plans. See Social Security reform plans Republicans: on LMS reform plan, 85–86, 92; on personal accounts, 39–40, 88; on tax increases, 6, 92 retirement age: benefits of working longer, 76–78; increase in, 11, 100; and life expectancies, 59f, 73–74, 74t, 119–20; need to encourage delaying, 85. See also delayed retirement; early-retirement; Normal Retirement Age (NRA) risks: government bonds to reduce, 137; on investments under LMS reform plan, 89–90, 90t; or investments, 160; of spending down personal accounts, 91 Rivlin, Alice, 75 Rother, John, 20 Ryan, Paul, 37, 51–52 Ryan-Sununu plan, 37–42 safety net: and an aging population, 57–60, 75–76; under Ball’s reform plan, 64–65, 65t; under Bennett’s reform plan, 50–51,
195
147, 148t; under Diamond-Orszag reform plan, 71, 148–49, 149t; under LMS reform plan, 148–49, 149t; Old-Age Risk-Sharing to preserve, 157; phasing out of benefit cuts, 105–6, 105t; as point of compromise, 157; for very old age, 7–12, 40, 57–60, 64, 93–96, 100; under A Well-Tailored Safety Net, 147–49, 148t, 149t. See also income security; Old-Age Risk-Sharing Samwick, Andrew, 85–86 Sass, Steven, 77 Saving, Thomas, 19, 44 savings, personal, 80, 82 Shaw, Clay, 23 social insurance function, 40 Social Security: benefits linked to inflation vs. wages, 45–47, 49; challenges facing Social Security, 3–7, 157, 159; cost projections, 4–5, 25; costs to fix, 34, 34f; covered workers, increasing numbers of, 107, 159, 167; earnings calculations, 101–2, 111, 168, 175n14, 185n8; government investment of revenues, 81– 82, 116, 156; low rate of return on contributions, 40–41; as pay-as-you-go system, 22, 40–41; Primary Insurance Amount, 89, 104, 139, 158, 175n14; promise of income security, 75, 83, 125– 26; social insurance function, 40; as third rail of American politics, 6; trustees, 19, 44 Social Security Disability Insurance Trust Fund: under Bennett’s reform plan, 50–51, 50t; projected depletion of, 20 Social Security reform plans: anti-reform arguments, 31–32; benefits formulas, 157; debt-financed privatization, 39–40; and disability benefits, 50–51, 50t,150– 51, 150t, 168, 168t; and economic growth, 79–80; effect on government debt, 25, 99, 145–46, 147f; effect on Social Security Trust Fund, 143–45; and fair burdens of sacrifice, 8; linked to health care reform, 117–22; Old-Age Risk-Sharing, 11, 104–9, 126–27, 144,
196
INDEX
157, 164, 165t; pay-as-you go proposals, 21–22, 44; and progressive benefits, 112–13; shifting safety net to older population, 75–76; tax increases, 25. See also payroll taxes; personal accounts; specific Social Security reform plans Social Security Trust Fund: backed by government IOUs, 7, 16–17, 18t, 25, 32; Bush on, 23, 27; effect of various reform plans, 44, 143–45; and government debt, 4–5, 7, 18t; government investment of, 81–82; investment in the stock market, 24, 62–63, 81–82; overview of, 15–20; and personal accounts claim on bonds, 135; projected exhaustion of, 7, 16, 33f, 99, 143, 159; under A Well-Tailored Safety Net, 116–17, 122 spousal benefits, 20, 87, 134, 167–68 Stabenow, Debbie, 27 standards of living, 45–47 Stenholm, Charlie, 102 Steuerle, Eugene, 31, 73, 102 stock market: instability of, 47; investment of Social Security funds in, 24, 62–63, 81–82; investment under LMS reform plan, 89–90, 90t. See also personal accounts Sununu, John, 37 supplemental savings, 13 surpluses, budget, 22 surviving spouses. See spousal benefits Sweden, 135 taxes: corporate, 38–39; Democrats and, 6, 19, 21–22, 53, 114; on employer-provided health benefits, 8, 12, 51–52, 117–22, 144, 159, 164; estate taxes, 32, 61–63; income taxes, 71; need to increase, 44; repeal of Bush’s tax cuts, 53–54; as share of GDP, 29–30; on Social Security benefits, 88; tax increases and Social Security reforms, 25
very old age. See aging population; A Well-Tailored Safety Net;income security; safety net wages vs. inflation,45 Walker, David, 16 Wehner, Peter, 39 A Well-Tailored Safety Net: accuracy of author’s figures, 145; benefits formula, 157; burdens of sacrifice under, 121–23, 121t, 153–54, 153t; carve-out plans, 131–36, 133t; and Delayed Retirement Accounts, 13, 135–41, 137t, 140t, 166– 67, 167t; and disability benefits, 150– 51, 150t; and early retirement, 126–27, 127t, 165, 166t; effect on government debt, 145–46f, 146f; effect on Social Security Trust Fund, 145–46; and employer-provided health benefits, 122; incentives for employers to employ older workers, 128–29, 128t; income security, 136–38, 160–61; and inheritibility of personal accounts, 13, 134–37, 160; middle class under, 152, 152t; and modified LMS reform plan, 154t; and Normal Retirement Age, 119–20, 126– 27, 127t, 165, 166t; and Old-Age RiskSharing, 11, 104–9, 145; payroll tax increases, 120–21, 120t, 147–48; and personal accounts, 120–21, 121t; phased-in benefits, 165, 166t; Plan B, 8, 122–23, 122t, 145–56; progressive saving offset, 12, 121–22, 127–28, 128t, 163, 164t; and progressive saving offset, 163, 164t; safety net, 147–49, 148t, 149t; and the Social Security Trust Fund, 116–17, 122; solvency impact of, 169 Wilcox, David, 32 women: poverty rate, 9; spousal benefits, 20, 87, 134, 167–68 workers per beneficiary, 5, 74t Wyden, Ron, 51
About the Author JED GRAHAM writes about economic policy for Investor’s Business Daily. He set about devising a new approach to saving Social Security after his reporting on the 2005 battle over Social Security reform led him to conclude that none of the solutions being put forward could deliver both an affordable and an effective safety net. Graham’s commentary also appears at the economics blog RGE Monitor. This work does not reflect the editorial view of IBD.