Canadian Policy Debates and Case Studies in Honour of David Laidler
This page intentionally left blank
Canadian Pol...
14 downloads
389 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
Canadian Policy Debates and Case Studies in Honour of David Laidler
This page intentionally left blank
Canadian Policy Debates and Case Studies in Honour of David Laidler Edited by
Robert Leeson
Selection and editorial matter © Robert Leeson 2010 Preface © James Davies and Robert Leeson 2010 Foreword (1) © David Dodge 2010 Foreword (2) © Gordon Thiessen 2010 Individual chapters © their authors 2010 All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No portion of this publication may be reproduced, copied or transmitted save with written permission or in accordance with the provisions of the Copyright, Designs and Patents Act 1988, or under the terms of any licence permitting limited copying issued by the Copyright Licensing Agency, Saffron House, 6-10 Kirby Street, London EC1N 8TS. Any person who does any unauthorized act in relation to this publication may be liable to criminal prosecution and civil claims for damages. The authors have asserted their rights to be identified as the authors of this work in accordance with the Copyright, Designs and Patents Act 1988. First published 2010 by PALGRAVE MACMILLAN Palgrave Macmillan in the UK is an imprint of Macmillan Publishers Limited, registered in England, company number 785998, of Houndmills, Basingstoke, Hampshire RG21 6XS. Palgrave Macmillan in the US is a division of St Martin’s Press LLC, 175 Fifth Avenue, New York, NY 10010. Palgrave Macmillan is the global academic imprint of the above companies and has companies and representatives throughout the world. Palgrave® and Macmillan® are registered trademarks in the United States, the United Kingdom, Europe and other countries. ISBN: 978-0-230-23734-6
hardback
This book is printed on paper suitable for recycling and made from fully managed and sustained forest sources. Logging, pulping and manufacturing processes are expected to conform to the environmental regulations of the country of origin. A catalogue record for this book is available from the British Library. A catalog record for this book is available from the Library of Congress. 10 9 8 7 6 5 4 3 2 1 19 18 17 16 15 14 13 12 11 10 Printed and bound in Great Britain by CPI Antony Rowe, Chippenham and Eastbourne
Contents List of Tables and Figures
vi
Foreword – David Dodge
viii
Foreword – Gordon Thiessen
x
Preface
xii
Notes on the Contributors
xiii
1
The Canadian Monetary Order: Past, Present, and Prospects William Robson Discussion: Peter Howitt
2
Canadian Monetary Policy Peter Howitt Discussion: John Crow
41
3
The Lender of Last Resort: Lessons from Canadian History Angela Redish Discussion: Neil Skaggs
80
4
Canadian Policy and the Economic Revolution in Asia John Whalley
100
5
Reforming Canadian Medicare: Can an Icon be Redesigned? Åke Blomqvist
122
6
Tax Incentives for Owner-Occupied Housing – Then and Now Finn Poschmann Discussion: John Palmer
7
8
The Political Economy of Inflation Targets: New Zealand and the UK C. A. E. Goodhart Discussion: Charles Freedman Monetary Union Proposals in North America and Southern Africa: Do the Same Q&A Apply? J. Clark Leith
1
156
171
215
Name Index
231
Subject Index
233 v
List of Tables and Figures Tables 1.1 Canada’s post-war monetary regimes: Key characteristics
26
3.1
86
Losses on deposits
3.2 Canadian bank failures, 1868–1934
87
3.3 Canadian bank amalgamations, 1868–1920
89
4.1 Annual growth rates in China, India, and ASEAN(5) compared to OECD
107
4.2 Canada’s imports and exports with selected countries from 2001 to 2005
111
6.1 Pros and cons of retaining current US tax concessions
163
7.1 Annual inflation rate (consumer prices), 1970–2004
200
7.2 Annual growth rate (GDP), 1970–2004
200
8.1 SADC member GDP and PPP GDP per capita
220
Figures 1.1
Short- and long-term interest rates: 1950–2005
8
1.2
Growth in M1 and GDP: 1950–2005
8
1.3
US$/C$ nominal and GDP-price-based real exchange rates: 1950–2005
10
1.4
Consumer price index (CPI) and GDP inflation: 1950–2005
12
1.5
Realized real short-term interest rates: 1950–2005
13
1.6
CPI inflation and target band: 1990–2005
16
1.7
Total and core CPI inflation: 1990–2003
19
1.8
Nominal and real-return 30-year bond yields
22
1.9
Twelve-month CPI inflation rates: December 1995 to December 2005
24
1.10 Standard deviation of quarterly change in real GDP (12-quarter window) vi
25
List of Tables and Figures
vii
3.1
Seasonal contradiction of notes: October to January
88
7.1
Consumer prices
201
7.2
GDP at constant prices
201
8.1
Structure of Botswana and South African economies
221
Foreword – David Dodge Remarks on the occasion of David Laidler’s retirement dinner, Thursday, May 18, 2006 David, Welcome to you and Antje. Sorry for the delay in having you both here. This is a social evening. We all look forward to a great dinner, but I would note that there will be an extra course of “Roast David” following dessert! It is a real pleasure to have you here as a friend of the Bank. As a true friend you have supported us, created interest in what we do and most importantly provided constructive criticism – criticism that has helped us to improve. Over the years, you never let the Bank forget that “monetary policy” begins and ends with “m – o – n – e – y”. You convinced Chuck Freedman to institute the “Blue Book” – formalizing the way in which money and credit are considered as we set monetary policy. Likewise, when Chuck retired three years ago, you reminded us that monetary policy without money is like “Hamlet without the ghost”. But David, your interest in central banking was not confined to monetary issues alone. Indeed, it was your wide-ranging interest in the Bank’s work that made you the ideal choice as the first Special Adviser to the Bank in 1998. In this capacity, you were active in all policy debates, and set the “gold standard” for those who would follow in your footsteps. David, the Bank has always appreciated your support and thoughtful criticism on crucial issues over the years. Not only as a professor at Western, but also through your work as fellow-in-residence at the C. D. Howe Institute, you have popularized monetary policy issues and made these issues understandable and relevant to Canadians. Your long list of publications with the Institute forms a lasting legacy, for us at the Bank, and for those who follow what we do. More recently, this has included your recommendations as a member of the Institute’s shadow monetary policy committee. Indeed, before you viii
Foreword – David Dodge ix
leave tonight we want to know how you would have voted in today’s close 6–5 recommendation on what we should do next week. On behalf of the Bank of Canada, I simply want to say “Thank You”. It’s great to have you and Antje here. Bon appetit!
Foreword – Gordon Thiessen Foreword for David Laidler’s Festschrift Volume The news in 1975 that the University of Western Ontario had succeeded in attracting David Laidler to move to Canada was greeted with great enthusiasm among economists at the Bank of Canada. We were familiar with his work on the demand for money; indeed, we were at the time seeking a stable demand function to use as a basis for an M1 target for Canadian monetary policy. What we could not imagine was the immense contribution he would subsequently make to the analysis, understanding and evolution of Canadian monetary policy and the operations of the Bank of Canada. David’s prolific association with the C. D. Howe Institute has been the source of some of the best analysis and practical commentary on monetary policy issues in Canada. The Bank decided in the late 1990s that it would benefit from inviting a series of outside experts to participate in research and decision-making on monetary policy inside the Bank. Not surprisingly, David Laidler was our choice for the first Visiting Special Adviser in 1998–9. It was a remarkably successful year. David’s active and helpful participation in the internal policy decision-making process was almost instantaneous. He understood the issues, and he quickly absorbed the flow of information and analysis that went into those decisions. As well, he led a group of Bank economists in a major research project during that year on a more active role for money in monetary policy decisions. David’s contributions to Canadian monetary policy have been many; however, I would like to emphasize three areas in particular. The first is his contribution to the development of inflation targeting in Canada, beginning with his analysis of price stability as the appropriate focus for monetary policy, to his work on issues relating to the implementation of inflation-control targets. A second area of contribution was his clear articulation of what constitutes a coherent monetary policy framework. He helped us to understand, improve and explain the policy framework we had put in place in Canada following our adoption of inflation-control targets in 1991. He also defended the crucial requirement of a floating exchange rate in that framework in public discussions and debates with advocates of greater monetary integration with the US. The third area x
Foreword – Gordon Thiessen xi
I want to mention is the role of money in monetary policy. David persistently criticised the Bank for the passive role that money played in our view of the policy transmission mechanism. Despite the success of his eloquent arguments in making us feel guilty about the lack of active money in our policy process and encouraging us to work harder at our research on this issue, neither his work nor ours managed to persuade us to move away from our focus on interest rates as the transmission mechanism for monetary policy. But it was not for lack of trying. David’s participation in the analysis and public discussion of monetary policy issues in Canada continues to this day. He has recently contributed to the discussion around the reconsideration of the Bank’s agreement on inflation-control targets with the Minister of Finance last year. The research on issues for the next reconsideration of the targets in five years’ time has already begun. I will not be surprised if once again David is an active contributor to the debate, helping the Bank to make the right decisions in 2011. Gordon Thiessen March 2007
Preface For over three decades, David Laidler has been a prolific contributor to Canadian public policy research. David migrated to Canada in 1975, and within six years was elected a Fellow of the Royal Society of Canada and, shortly afterwards, President of the Canadian Economics Association/ Association Canadienne d’Economique. In 1999, he received University of Western Ontario’s highest award for research, the Hellmuth Prize. David has long been closely associated with the C. D. Howe Institute. With William Robson (the Institute’s current President and CEO), he was the joint recipient of two of Canada’s most prestigious social science prizes. In 1994, they won the Douglas Purvis Memorial Prize awarded by the Canadian Economics Association for a significant work on Canadian Economic Policy, recognizing The Great Canadian Disinflation. Their Two Percent Target: Canadian Monetary Policy Since 1991 won the Donner Prize for the best book of the year on Canadian public policy in 2004. David has also made significant direct contributions to Canadian governance: as member of the Economic Advisory Panel to the Hon. Marc Lalonde, Minister of Finance (1982–4), as a research coordinator for the Royal Commission on the Economic Union and Development Prospects for Canada (the Macdonald Commission, 1983–5) and as a Special Advisor to the Bank of Canada (1998–9). In 2005, the University of Western Ontario hosted a conference to mark the occasion of David’s retirement from full-time teaching. The following year, the Economic Policy Research Institute, the Department of Economics and the Faculty of Social Science at U.W.O hosted a Festschrift conference in David’s honour. This two-volume series is the product of that conference. The essays in the first Festschrift volume focus on issues raised by David’s contribution to macroeconomics; the essays in this second volume are focused primarily on Canadian policy issues. Two former Bank of Canada Governors, David Dodge and Gordon Thiessen have provided Forewords to this volume. James Davies and Robert Leeson May 2009 xii
Notes on the Contributors Ake Blomqvist is Professor of Economics at the China Center for Human Capital and Labor Market Research, Central University of Finance and Economics, Beijing John Crow is Advisor at Lawrence Asset Management, Canada and was formerly Governor of the Bank of Canada James B. Davies is Professor in the Department of Economics at the University of Western Ontario, Canada Charles Freedman is Scholar in Residence in the Economics Department at Carleton University, Canada Charles Goodhart is Emeritus Professor of Banking and Finance at the London School of Economics, UK Peter Howitt is Lyn Crost Professor of Social Sciences at Brown University, US Robert Leeson is Adjunct Professor of Economics at Notre Dame Australia University J. Clark Leith is Professor Emeritus in the Department of Economics, University of Western Ontario, Canada John Palmer is Professor in the Department of Economics at the University of Western Ontario, Canada Finn Poschmann is Vice President, Research at the C. D. Howe Institute, Canada Angela Redish is Professor in the Department of Economics at the University of British Columbia, Canada William B. P. Robson is President Chief and Executive Officer at the C. D. Howe Institute, Canada Neil T. Skaggs is Professor in the Economics Department at the Illinois State University, USA John Whalley is Professor of Economics at the University of Warwick, UK
xiii
This page intentionally left blank
1 The Canadian Monetary Order: Past, Present, and Prospects William Robson
The troubled history of fiat money suggests that a durable monetary order needs a number of characteristics: a unique, logical goal for monetary policy; technical power to influence monetary conditions decisively; tactical skill to use that power effectively; private expectations and behaviour that conform to the goal; democratic support and accountability; and resilience in the face of foreseeable shocks. Canada’s post-war history features a number of monetary regimes that lacked several of these features and proved brittle. The period of 2-percent inflation targeting that has run from the end of 1995 to the present, however, exhibits these characteristics to an unusual degree. While minor modifications to the regime are conceivable, and some are desirable, Canada’s current order appears robust enough to warrant confidence in its durability.
Introduction and overview1 The idea of a monetary order – a coherent and robust set of rules, institutions, expectations, and behaviours – deserves attention in a world still struggling to master that extraordinarily powerful and problematic invention, fiat money. My first encounter with the term was in a 1991 paper by David Laidler in which he drew on the exposition of Brunner (1984) to stress the need for considering the broader environment – the rules about the rules – in looking for alternatives when a given operational rule for monetary policy failed. 1
2
The Canadian Monetary Order
Attention to many features – the type of goal monetary policy has, the central bank’s technical powers and tactical skill in pursuing it, democratic support for the goal and the central bank’s accountability for its performance, the degree to which people believe the central bank will achieve its goal, and the resilience of the regime in the face of shocks – is always necessary to understand a particular monetary regime well enough to offer policy advice. And David, in whose work useful policy advice has been a priority, has devoted attention to all of them. This paper describes the development of a monetary regime in Canada that, since the early 1990s, appears to have acquired a logical goal, power to achieve it, democratic support, and enough resilience that agents can reasonably expect it to endure – in short, a monetary order worthy of the name. Inflation targeting in Canada is not a unique accomplishment; nor, as I discuss in the final section of the paper, is it obviously a kind of ‘end of monetary history’. The troubled story of fiat money in Canada and elsewhere, however, does make it notable. Inflation targets in Canada may evolve in the next five years. The numeric target, the price index used, or the time-period over which it is measured might change. Even so, I expect the roots of Canada’s future monetary order in today’s will still be clear enough in a decade or so to justify treating them in one narrative. A reasonable person – or, to express the idea in terms more in keeping with the topic, a rational agent – can therefore make plans sensitive to the Canadian dollar’s domestic purchasing power over time with a degree of confidence that has not existed for half a century. David’s central role in the development of Canada’s current monetary order, one that has given me the pleasure of regular co-authorship and collaboration, makes this exploration an apt contribution to a festschrift in his honour. It will also provide, I hope, an account of interest to Canadians, and to those elsewhere who wonder if Canada’s experience contains useful lessons for them.
Defining a monetary order The context, often implicit, for much exploration of approaches to stabilizing prices around some trend, and potential trade-offs between that goal and others, is one where the tools of monetary policy work, the democratic process supports the goal, and agents respond as the modeler him/herself would. Since real life contexts often feature illogical or contradictory goals, inept implementation, expectationsconfounding shifts, and high-level political conflicts over monetary
William Robson 3
policy, an explicit list of the elements in a benign monetary order is a good place to start this account. Clear goal Like David (Laidler 1991b), I would give a well-defined goal for the monetary authority pride of place among the features of anything meriting the term ‘monetary order’. A monetary regime to which individuals and businesses will adapt their expectations and behaviour, and that commands enough support to be resilient in the face of shocks, must have a central focus that is logical and that allows measurement of success. This feature might appear obvious to many monetary economists. Monetary economists can differ about the virtues of particular goals or their compatibility with others, however, and confusion on this point is common among policymakers and the public. The idea that monetary policy can pursue many goals at once or should switch with circumstances has prevailed in the past, arises regularly, and features in the mandates and/or enabling legislation for many central banks. The Bank of Canada’s mandate, virtually unchanged since the 1930s, is a fine example. It directs the Bank to regulate credit and currency in the best interests of the economic life of the nation, to control and protect the external value of the national monetary unit and to mitigate by its influence fluctuations in the general level of production, prices and employment, so far as may be possible within the scope of monetary action, and generally to promote the economic and financial welfare of the Dominion.2 Wide acceptance nowadays among non-economists of the proposition that monetary policy is one tool that can logically achieve only one goal among a small set of choices is a tribute to painful experience and to persistent, clear argument.3 Without fresh memory of that experience and continued cogent argument supporting it, that acceptance could evaporate. Technical power and tactical skill The logical necessity for the monetary authority to be able to achieve its goal has two elements, usefully considered separately. One is a technical requirement. It must have enough leverage over the financial system and monetary conditions to achieve the goal. In practice, this means controlling the terms on which financial intermediaries
4
The Canadian Monetary Order
access a means of payment of unique soundness and liquidity. Observing the evolution of this control from the days of legally mandated reserves and open market operations to today’s participation in ‘real time’ clearing and settlement systems, one might wonder, like Friedman (1999), if technological innovation will undermine this control. For now, however, such systems let central banks exert decisive influence by standing ready to borrow practically unlimited amounts or lend literally unlimited amounts of high-powered money at the floor and ceiling of a range they set for a very short-term interest rate. The second requirement is tactical. The monetary authority must possess the skill to deploy its leverage effectively. The need to understand the monetary transmission mechanism and the consequences of policy actions may also seem obvious but should not be taken for granted. Inflation-targeting central banks, for example, have had to overcome reservations about their ability to keep inflation in a band narrow enough to be meaningful to private agents. Many non-economists would be astonished (and policymakers sometimes are) to learn how deep the uncertainty inside central banks is about such questions as the significance of money, the role of the exchange rate, and the level of a medium-term ‘neutral’ policy rate. Every debate over an interest-rate decision by a central bank reveals how many conditions we must satisfy to assert that a central bank has the tactical ability to achieve its goal. Conforming expectations and behaviour A further characteristic of a monetary order, which following David (Laidler 1999b, 1–2) – who there and elsewhere acknowledges his debt to Axel Leijonhufvud – I will call coherence, is that private agents understand the goal, expect its successful pursuit, and act accordingly.4 Such understanding and expectation need not be universal, but one would want them to prevail sufficiently to affect the relevant prices and quantities on the relevant margins. Despite modern preoccupation with formal goals, accountability and policy rules, considerable coherence can exist without them. Private expectations that are decisively inconsistent with a given set of goals and rules will cause an order to break down.5 Smaller deviations, however, need not. The gold standard did not decisively constrain all its adherents all the time, but nevertheless lasted for generations, sustained by conventions and shared assumptions that aligned behaviour tolerably well with its formal goal. This type of coherence, which gives a monetary order characteristics akin to common law or a language, creates considerable robustness under stress.6
William Robson 5
Political commitment and accountability The controversial questions surrounding the appropriate relationship between the monetary authority and other executive and legislative organs of government – a set of issues David (Laidler 1999b) addresses under the heading of a ‘liberal’ monetary order, on the assumption that the government in question is a democratically elected one – also has two elements worth separate consideration. The first of these – the role of those who determine the goals of monetary policy and supervise the people charged with pursuing them – is usually framed as a kind of negative option connected with the degree of ‘goal independence’ and ‘instrument independence’ of the central bank. The government cannot be so unhappy with the goal chosen by the monetary authority that it formally overturns it. When an elected government has endorsed a goal such as a price-level-related target or a fixed exchange rate, one can reasonably argue that lack of goal independence is a positive feature of the regime, since a goal with the electorate’s endorsement – even if only in the negative sense that they are not upset enough to choose something else – is presumably more reliable than one without. With regard to instrument independence, the government cannot be so unhappy that it interferes with the management of monetary policy so as to effectively overturn the goal. The key issue here is one of the government’s committing to the goal strongly enough to overcome any short-term temptation to undermine it. The second element is a requirement for contingent positive action. Other branches of government must be able to reward or punish a central bank for its determination and adeptness in pursuing its goal. Without this ability, irresolute or inept behaviour could undermine faith in the medium-term achievement of the goal and the long-term durability of the regime, precluding the conforming private expectations and actions that underlie many of the key benefits of a benign monetary order. Resilience While the challenges of time inconsistency and agency problems just alluded to have generated much thoughtful commentary, the theme of a monetary regime’s durability in the face of larger changes in the economic or political environment is, as far as I know, less well explored. But a monetary order to which reasonable people adapt their expectations and behaviour must logically be, and appear to be, resilient to many types of shocks. A central bank might receive a mandate from a democratically elected and accountable government to change its goal for a variety of
6
The Canadian Monetary Order
reasons: distress over exchange-rate volatility, a change in the economy or the financial system that appeared to make pursuit of a previous goal damaging or impossible, or convincing evidence that monetary policy could enhance well-being in ways previously considered out of its scope. The possibility of such changes always exists, and even very long-lived regimes have ultimately given way in the face of war, disaster or revolution. For present purposes, the key question is the reasonableness of private agents’ conforming their expectations and behaviour to the central bank’s goal and its tactics in pursuing it. The world’s monetary regimes would likely not survive the impact of a mass-extinction-scale meteor. Short of anything so extreme, one might judge an order fragile if a foreseeable shock on a less cosmic scale might lead the electorate to vote for a change. If a population that understands the goal of policy and the means by which the central bank pursues it appears likely to exercise its power to change matters only to bring a central bank that has strayed from its goal back on course, the regime is a resilient one.
The evolution of Canada’s monetary order History contains many examples of monetary regimes that featured the elements just described – perhaps in varying degrees, but sufficiently to constitute a recognizable order. In Canada’s case, the 60 years of integration with the gold standard from 1854 to 1914 (Powell 1999, 14–18) would constitute a monetary order. This paper can usefully begin in the post–World War II era, when the Bank of Canada, established in 1934 and reconstituted as a government-owned enterprise in 1938, first began operating in ‘normal’ times. This section describes Canada’s monetary history through the floating exchange rate and conflicts between the Bank and the elected government of the 1950s, the fixed exchange rate of the 1960s, the high and variable inflation from the early 1970s through the end of the 1980s, the advent of inflation-reduction targets in the early 1990s, and the two-percent inflationtargeting regime that has run from the end of 1995 to the present. The presence or absence of key features of a coherent and resilient monetary order provides some notable contrasts. An assessment of Canada’s current order in their light is the subject of the following section. The floating exchange rate and conflicts of the 1950s While noting the risk that hindsight promotes facile decisive judgements, Canada’s monetary history from 1950 to 1961 has features that would
William Robson 7
justify withholding the term ‘monetary order’ from the arrangements of the period. One key criterion – technical power – was in place. The Bank could influence monetary conditions and the price level. It controlled access by the chartered bank to reserves they had to hold against demand and saving deposits. There was also a major effort in the 1950s to foster the growth of a deep, liquid money market that, among other things, would help the Bank conduct efficient open-market operations. Whether the tactics of monetary policy were adept is hard to say, however, because two key criteria – a clear goal and democratic support – were lacking. As a result, the regime proved brittle. As noted above, the Bank of Canada’s mandate – set out in the preamble to the legislation that created it – mentioned a profusion of ways by which monetary policy was to enhance economic well-being. Such a multiplicity of objectives need not be fatal to a monetary order, if the common understanding is that a particular goal, such as domestic price stability or the maintenance of a given bilateral exchange rate, has primacy. In the 1950s, however, no such common understanding existed. The wording of the mandate gave the external value of the currency pride of place, reflecting the hopeful expectation of a restored gold standard at the time of its drafting in the 1930s. But attempts to fix the value of the exchange rate in 1946 and in 1949 had proved unsatisfactory, and in 1950, Canada reacted to an acceleration of inflation generated by favourable terms-of-trade shifts, buoyant net exports, and inflows of capital by floating. Because this arrangement was declared to be temporary, and was self-consciously inconsistent with international practice and Canada’s Bretton Woods commitments (Powell 1999, 42), no clear statement of another goal for monetary policy replaced the requirement for a fixed exchange rate. And when Graham Towers, the governor who had presided over the Bank of Canada since its creation, was succeeded by James Coyne in 1954, this lack of clarity combined with a problem of political commitment – high-level conflict over monetary policy – to undermine the regime. The new governor paid considerable attention to the country’s balance of payments. Canada’s tendency to attract large amounts of saving on occasion meant that this focus yielded somewhat erratic policy – evident in a saw-tooth pattern of short-term interest rates (Figure 1.1) and in growth of M1 and nominal GDP (Figure 1.2). In 1957, the Liberal government that had won election after election since the mid-1930s lost to a minority Progressive Conservative government that had criticized
8
The Canadian Monetary Order 20
Percent
15
10
5
0
–5 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 3-Month T-bills
Spread
Short- and long-term interest rates: 1950–2005
Year-over-year change (%)
20
20
15 15 10 10 5 5 0
–5 0 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 Net M1 (left scale) Figure 1.2
Nominal GDP (right scale)
Growth in M1 and GDP: 1950–2005
Year-over-year change (%)
Figure 1.1
10-Year-Plus Bonds
William Robson 9
monetary policy in opposition, and a huge Conservative majority followed in 1958. A refinancing of war-time debts which put upward pressure on long-term interest rates (Bothwell, Dummond, and English 1981, 220–2), and economic weakness and rising unemployment at the end of the 1950s, brought matters to a head. On one level, the governor of the Bank during those years might have appeared to enjoy a high degree of independence with regard to goal and instruments, serving during good behaviour for a seven-year term. The events that followed, however, showed the fragility of this arrangement in the face of profound disagreements with the elected government. In 1960, Coyne began making speeches expressing strong opinions on trade policy, inbound investment, fiscal policy, and the exchange rate. A procedural squabble related to the governor’s pension further inflamed the situation. Determined to rein him in, the government passed legislation in the House of Commons declaring the post of governor vacant. The bill failed in the Senate, but nevertheless achieved its goal: governor James Coyne resigned. Monetary policy promptly turned expansionary, as is evident in the wider spread between long- and short-term interest rates and faster growth of money and spending. Easier money and unscripted statements about the government’s desire for a lower exchange rate put the Canadian dollar under strong downward pressure. In May 1962, the government abandoned its efforts to stem the pressure through foreign-exchange interventions alone, and pegged the Canada–US dollar exchange rate at one Canadian dollar to 92.5 US cents (Figure 1.3), closing out Canada’s messy first major post-war monetary regime. A pegged exchange rate and rising inflation in the 1960s By comparison with what came before it, Canada’s monetary regime in the 1960s had some positive features. Monetary policy had a clear goal: maintaining the exchange rate within one percent of the specified value. From a technical point of view and – with more nuance – from a tactical point of view, power to achieve the goal was also evident. In the matter of central bank autonomy and accountability, moreover, the aftermath of what came to be known as ‘the Coyne affair’ yielded a new approach to the problem of political commitment that moved Canada towards a more resilient regime. Coyne’s successor as governor, Louis Rasminsky, made clarification of the respective monetary policy responsibilities of the elected government and the Bank of Canada a condition of accepting appointment. This clarification, later dubbed the doctrine of ‘dual responsibility’, vested the Bank of Canada with
The Canadian Monetary Order 110
1.40 1.30
100
US cents
1.20 90 1.10 80 1.00 70 60
Index: 1997 = 1.00
10
0.90 0.80 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 NER (left scale)
RER (right scale)
Figure 1.3 US$/C$ nominal and GDP-price-based real exchange rates: 1950–2005
authority to conduct policy, but noted the Minister of Finance’s ability to exercise final authority on behalf of the parliament. A 1967 amendment to the Bank of Canada Act gave this doctrine legislative effect, specifying that to exercise his authority, the Minister of Finance must issue a written directive with explicit instructions as to actions and their time frame. That a Governor who received such a directive would resign was made clear by Rasminsky at the time, and has been a key feature of Bank of Canada commentary on this provision since. In retrospect, requiring the government to override the governor only in this visible and potentially damaging way appears to have bolstered the Bank’s autonomy in implementing monetary policy. Pegging the exchange rate, however, made this aspect of the change less salient at the time. What did become clear as the 1960s progressed was that the peg – for reasons that require little elaboration to a modern audience likely to be sceptical of such arrangements – turned out not to be a resilient regime in the face of changes in US monetary policy and in Canada’s external balance. Historical accounts of this episode (Bothwell, Dummond, and English 1981, 319–20) tend to focus on Canada’s initial difficulty in staving off downward pressure on the peg, particularly in the face of American concerns about US current account deficits. Looking at interest rates and money growth through the period, however, the more notable feature
William Robson 11
is a gradual move to higher short-term rates in the face of rising rates of money and spending growth through mid-decade. A spell in the late 1960s when maintaining the peg and avoiding serious domestic inflation appeared less at odds – evident in the downward dip in an otherwise rising trend in the real exchange rate in Figure 1.3 – was followed by a surge in demand for Canadian exports, and the classic tension between incompatible goals for the exchange rate and the domestic economy became overwhelming. In May 1970, the government floated the dollar again, drawing a line under Canada’s second major post-war monetary regime. Disorder: A floating exchange rate and variable inflation in the 1970s and the 1980s Canada’s monetary history over the following two decades has much in common with that of other major democracies during the period. In principle, the decision to float the exchange rate opened the way to a domestic price-level-related target. But this float also was seen as temporary, with no lack of official and unofficial plans to restore a peg (Powell 1999, 48–9). The Bank of Canada, moreover, took the end of the need to sterilize excess money from interventions to hold the exchange rate down as an opportunity to lower short-term interest rates. A torrent of money growth ensued, followed by more rapid spending and much higher inflation. The removal of the constraint related to the external value of the currency thus ushered in, as elsewhere, a regime in which goals were unclear, political commitment uncertain, implementation erratic, and expectations incoherent. The drama and detail of monetary policy in Canada in the 1970s is beyond the scope of a brief account. The governors of the Bank of Canada – Rasminsky until 1973, Gerald Bouey from 1973 to 1987 and John Crow after 1987 – frequently spoke against inflation, and in the late 1980s, began mentioning price stability as the long-term goal for monetary policy. These references, however, were too vague and compromised by other articulated and implicit objectives of the Bank and the government to constitute a clear goal. Although the Bank continued to have the technical means to control monetary conditions, its tactics reflected confusion and disagreement about strategy and implementation. Under these circumstances, neither coherent private expectations nor resilience in the face of shocks was in evidence. The period of greatest disorder was from the float in 1970 to 1975. Although there are hints of restraint in the upward movement of shortterm interest rates during this period and some deceleration in money
12
The Canadian Monetary Order
growth and in inflation around mid-decade (Figure 1.4), the attempt to find an attractive trade-off between inflation and an unemployment rate rising for various structural and policy reasons gives the retrospective observer much the same impression as that of many people at the time: that things were out of control. High inflation had become enough of a political issue by 1975 to prompt two changes. Wage and price controls from 1975 to 1978 had no lasting significance, and merit mention only as evidence of confusion about inflation’s causes and remedies severe enough to preclude any real political commitment to a monetary solution or a tactically adept programme.7 More significant for its enduring impact on Canadian monetary policy was a particular tactic – a variation on rules tried elsewhere – of gradual reduction in the growth rate of M1. Even had concern to contain movements in the exchange rate not frequently compromised the programme, the implementation of the policy of ‘gradualism’ was, as David has stressed (as, for example, in Laidler 2005,4–5), inconsistent with its monetarist inspiration. Rather than treating M1 as an independent influence on output and prices, the Bank treated its stock as demand-determined, and focused on moving short-term interest rates to levels that, according to estimated short-run demand-for-money equations and projections for output and
Year-over-year changes, (% pts)
16 14 12 10 8 6 4 2 0 –2 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005
CPI Figure 1.4
GDP Price Index
Consumer price index (CPI) and GDP inflation: 1950–2005
William Robson 13
prices, would yield a stock consistent with the desired declining growth path. With inflation-boosted interest rates changing the relationship between M1 and broader monetary aggregates, and between M1 and nominal income (Howitt 1990a), a policy that effectively validated most ongoing inflation in this way had little chance of success. Spending growth and inflation stayed high even as M1 growth fell (Figures 1.2 and 1.4), and gradualism became defunct in fact before its formal abandonment in the early 1980s. Tactical uncertainty naturally affected the Bank’s policy-rate setting. An oblique testimony to the tensions of the time was the 1980 move from periodic announcements of the policy interest rate, the Bank Rate, to a system that set it 25 basis points above the yield on the federal government’s three-month treasury bills at their weekly auction. What effect this attempt to direct attention to money-market conditions rather than discrete policy changes had on private-sector understanding of the goals or tactics of monetary policy is hard to say.8 The erratic pattern of realized real short-term interest rates during the decade from the early 1970s to the early 1980s (Figure 1.5) certainly illustrates vividly the stormy seas private agents had to navigate during this period. The most striking volatility occurred around the turn of the decade, when the Bank of Canada reacted aggressively to protect the value of the Canadian against the US dollar during the double-barreled US tightening in 1980 and 1981. As a result, real output dropped, and consumer price
3-Mo T-Bill yield minus y/y CPI (%pts)
10 8 6 4 2 0 –2 –4 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 Figure 1.5
Realized real short-term interest rates: 1950–2005
14
The Canadian Monetary Order
index (CPI) inflation dropped from double digits in 1981 to 3–4 percent in 1984. Through the mid-1980s, the Bank might be said to have resumed its search for an attractive point on an inflation–unemployment frontier. Notwithstanding earlier mentions of price stability, it was not until Governor Crow explicitly emphasized it in his Eric W. Hanson Memorial Lecture in January 1988 (Crow 1988) that this potential goal attracted attention. As policy tightened in response to the inflationary pressures unleashed by exaggerated fears of recession after the 1987 stock-market crash and attempts to damp a cyclically fuelled exchange-rate rise, it began to achieve notoriety as well. The late 1980s were a particularly fraught period for Canadian monetary policy. Crow’s personal forcefulness impressed many and appalled others. But his designation of price stability as an objective did not define a goal definite enough for endorsement by the elected government or to anchor private expectations. Many observers accepted that inflation at the time was too high, and feared lest it move higher, but the attitude of many in the private sector was nicely summed up by a visitor from the Organization for Economic Cooperation and Development (OECD) to the C. D. Howe Institute who summarized his conversations with business leaders to me in this way: I ask them: ‘Do you think Governor Crow’s goal of price stability is credible?’ They say: ‘Oh yes, he’s very determined.’ So I ask them: ‘So you think he’ll drive inflation down all the way to zero?’ They say: ‘Well, the man’s not crazy!’ The hazards of this situation shortly showed in acute form. Many financial market participants saw an early-1990s downward move in the Bank Rate – too small to see at less-than-daily frequency – in response to flagging domestic demand as a sign that the Bank’s determination to drive inflation down was weakening. The exchange rate dipped and long-term interest rates spiked. The Bank responded by pulling short-term rates sharply higher again through mid-year. Money growth turned negative, so did nominal spending, and the economy went into its second severe recession in a decade. Inflation-reduction targets: 1991–5 In an awkward coincidence, this contraction occurred in the run-up to an important fiscal change: replacement of a federal wholesale sales tax with a (value-added) Goods and Services Tax (GST) at the beginning
William Robson 15
of 1991. The base for the old tax had included exports and many intermediate inputs but not imports. Since the new tax was closer to a consumption tax, its imposition would give consumer prices a one-time boost. Not wanting this effect to exacerbate the pain of reducing inflation, the Minister of Finance and the Governor of the Bank of Canada jointly announced inflation-reduction targets at the time of the 1991 federal budget. These first targets called for the year-over-year change in the CPI to fall gradually to two percent by the end of 1995. In what is now a standard feature of such targets, the announcement also set an error band of one-percentage point either side of the target. In what is also a common move, the announcement indicated that the CPI excluding food and energy would be the Bank of Canada’s operational target, and that Bank would ignore any first-round effects of changes in indirect taxes – of which the implementation of the GST was the salient instance – in pursuing the target. The inflation-reduction targets gave Canadian monetary policy a clear, logically consistent goal, and involved a fresh degree of political commitment. As for technical powers, between 1992 and 1994, the government phased out the fractional reserve requirements on demand and notice deposits that gave the Bank the familiar power to influence financial institutions’ willingness to expand their balance sheets by varying the size and composition of its own. While this change may have complicated attempts to interpret the growth rate of M1 in the first half of the decade (Boessenkool, Laidler, and Robson 1996), the Bank did not devote much attention to M1 in those years, and it is fair to put this under the heading of tactical adeptness rather than anything that fundamentally affected its ability to steer monetary conditions. Its control over the price of, and conditions of access to, the high-powered money financial institutions required to settle transactions with each other in the clearing and settlement system clearly continued. On the tactical front, however, the new regime had other problems than those created by the phase-out of reserve requirements. Falling inflation and short-term nominal interest rates spurred an increase in demand for transactions money – an increase that the Bank, neglectful of money and determined in any event to display no weakness, failed to accommodate fully. As a result, inflation dropped below target, and economic weakness persisted well into the 1990s. The disappearance of the GST-related boost from year-over-year CPI inflation in 1992 left inflation below the bottom of the target range. During the 37 months from then until the first quarter of 1995, CPI inflation
16
The Canadian Monetary Order
averaged 1.2 percent. For 29 of those months (12 of which were slightly affected by a cut in cigarette excise taxes early in 1994), it was below the bottom of the target band (Figure 1.6). Another problem was the stripped-down CPI used for operational purposes – a deficiency criticized at the time (Laidler 1991a) for its inadequacy should relative price changes cause significant differences between it and the total CPI. The prominence of the ‘core’ measure muddled private sector expectations and behaviour: some observers thought it was the Bank’s formal target, and a great many others thought – not without reason – that it was the Bank’s de facto target. A second obstacle to the adaptation of expectations to the new regime was that the Bank’s and the government’s intentions were vague after 1995: the announcement simply stated that year-over-year CPI inflation would fall to a rate ‘clearly below 2 percent’ at a later date (Bank of Canada 1991, 5). The new regime scored much higher on the political commitment scale. Not only did the elected government of the day endorse the targets, but in the federal election campaign of late 1993 – to the relief of those, including David, who feared the Finance Minister’s involvement might expose the targets to partisan attack (Laidler 1991a) – the targets were not an explicit issue. While continuing weakness in the economy
Year-over-year change (%)
8
6
4
2
0
2 19 93 19 94 19 95
91
19 9
19
19
90
–2 3 96 997 998 999 000 001 002 00 004 005 2 2 2 2 2 2 1 1 1
19
CPI Figure 1.6
Target Band
CPI inflation and target band: 1990–2005
William Robson 17
meant that monetary policy did figure in the campaign, the focus of criticism tended to be Governor Crow himself. When the opposition Liberals took power, they declined to appoint Crow to a second term, but the man who took his place was Gordon Thiessen, the previous Senior Deputy Governor, and part of the inflation-targeting team. A joint announcement by the new Minister of Finance, Paul Martin, and Governor Thiessen reaffirmed the targets to the end of 1995, and extended the two-percent target, with its 1–3 percent band, until 1998. Despite the disappearance at this point of the ultimate goal of inflation clearly below two percent, the post–1993 election outcome demonstrated that Canada’s monetary regime had acquired new resilience. The numerical targets for inflation survived. The focusing of dissatisfaction with the Governor personally did not result in the use of the directive by the Finance Minister, the Governor served out his term and his senior deputy succeeded him. Two-percent inflation targeting since 1995 The monetary order that now prevails in Canada dates from the end of 1995, the point when the original inflation-reduction targets specified two percent, and from which the two-percent target and its one-percentage point error band was extended. The same target was extended for a further three years in 1998, and again for five years in 2001, a renewal that stressed keeping inflation in the middle of the range more than its predecessors had (Bank of Canada 2001). In the light of the comments made above about the extent to which previous regimes contained elements of a coherent, liberal, resilient monetary order, the two-percent targeting regime deserves scrutiny under each heading. The goal From a world perspective, Canada now appears simply as the second in line to adopt a regime pioneered by New Zealand a few years earlier, and since taken on in two dozen countries and currency areas. For Canadians, however, establishing an unchanging numerical goal for inflation over a multi-year period was a fundamental innovation. For the first time, monetary policy had a domestic price-level goal consistent with the central bank’s powers. The explicit endorsement of the inflation target by the elected government also ensured that the official goal had undoubted primacy over other possible objectives, including maintaining a value for the exchange rate.
18
The Canadian Monetary Order
Technical powers The Bank of Canada’s subsequent success in hitting the inflation targets shows that its technical ability to control monetary conditions continued through this period, despite further evolution in the technology and practice of monetary control. At the beginning of the 1990s, the fulcrum for the Bank of Canada’s control of monetary conditions was a clearing and settlement system in which each financial institution was uncertain about the net position its customers’ activities today would require it to settle by noon tomorrow. This uncertainty created day-to-day demand for high-powered money, since the financial institutions that participated directly in the system wanted sufficient balances at the Bank of Canada to avoid borrowing at the Bank Rate – which continued to be set at a 25 basis-point margin above the yield on three-month treasury bills at weekly auction. In 1999, a new real-time, gross-equivalent scheme, the Large Value Transfer System (LVTS), started operating, and rapidly became the clearing system for most inter-bank activity.9 Because the LVTS provides immediate, final settlement of transactions, it eliminated much of the old system’s uncertainty about overnight net positions, and with it much of the need for precautionary balances.10 It did not eliminate the need for central-bank money entirely, however, because LVTS participants face uncertainty about their ability to meet whatever net demands their customers’ activities create, and this uncertainty rises with the volume of transactions in the network. By standing ready to borrow from or lend to LVTS participants – whose potential demand for high-powered money rises with the volume of transactions through the system – at 25 basis points below or above the target it sets for the overnight rate, the Bank still exerts satisfactory control over monetary conditions. Tactical skill On the tactical front, the decade of two-percent targeting has featured important nuances in at least three areas: the use of ‘core’ inflation measures as policy guides, the Bank of Canada’s reaction to fluctuations in the exchange rate, and the setting of the policy rate. Since 1995, the Bank has given varying amounts of emphasis to core inflation measures that strip out or give less weight to the more volatile components of the overall index. For a time, it gave core inflation pride of place in the upper right corner of the main page of its website, which further fed confusion – even among professional central-bank watchers – about whether core inflation was the actual target.
William Robson 19
For those who lean towards a money-based theory of price-level determination, stripped-down or re-weighted indices present obvious problems; for those who favour a Phillips curve, output-gap version of the inflation story, it is less troubling in theory.11 In practice, the core inflation’s ability to give a ‘cleaner’ read on the trend of price pressure matters. In Canada’s case, no sooner had some researchers marshaled statistical support for its value as a leading indicator of inflation (Macklem 2001), than a prolonged uptrend in energy prices created a persistent gap between the core measure and the total, targeted rate (Figure 1.7). While the core measure continues to appear on the front page of the Bank’s website at the time of writing, the total measure now gets pride of place. The emphasis the Bank of Canada gave to short-term movements in the exchange rate has also varied since 1995. Early in the period, the Bank tended to use a summary Monetary Conditions Index (MCI) – a weighted combination of short-term interest rates and the trade-weighted exchange rate – to gauge the impulse monetary policy was giving to demand. In the MCI, a one-percentage-point move in short-term interest rates is assumed to be equivalent in its effects to an exogenous threepercent move in the exchange rate in the other direction – assuming that nothing meanwhile has changed the ‘neutral’ value of the index.
8 Year-over-year change (%)
7 6 5 4 3 2 1 0 –1 3 96 997 998 999 000 001 002 00 2 2 2 2 1 1 1
19
CPI Figure 1.7
20 04
19 93 19 94 19 95
91 19 92
19
19 90
–2
"Core" CPI
Total and core CPI inflation: 1990–2005
05 20
20
The Canadian Monetary Order
The use of this formula for offsetting movements in the exchange rate with movements in short-term interest rates by the Bank was at times so predictable that financial market participants calibrated their moneymarket activities accordingly. Such an approach presumes a great deal about the central bank’s ability to distinguish different kinds of movements in the exchange rate, and the effect a given development might have on the desired value of the index. Some movements in the exchange rate might be connected to actual or incipient changes in demand for Canadian goods and services – so a depreciation, for instance, might signal a weakening of demand and merit an unchanged or even a lower policy interest rate. Other exogenous movements in the exchange rate might, if allowed to work their way through the economy, affect the net trade balance and therefore merit an offsetting monetary-policy response – so a depreciation would merit a higher policy rate.12 Since all kinds of factors drive actual exchange-rate movements, responding automatically to every wiggle with offsetting adjustments in the policy rate created problems. Downward pressure on the exchange rate during the Asian and Russian crises of 1997–8 prompted policy-rate hikes Canada would have been better off without.13 Those lessons – and the happier experience of Australia, which did not respond that way, and the unhappier experience of New Zealand, which did so even more strongly – likely account for the Bank’s greater willingness to look past movements in the exchange rate in the late 1990s. A corresponding change in its interventions in the foreign-exchange market – from frequent discretionary interventions to smooth fluctuations in the early 1990s, to a relatively automatic stance of symmetric intervention in the mid-1990s, to the period since late 2000 when the Bank has not intervened to influence the Canadian dollar’s value at all – means that the currency now floats quite cleanly (Laidler and Robson 2004, 124–5). The setting of the Bank’s policy rate underwent two useful changes during the decade of two-percent targeting. As noted already, a system that set the Bank Rate weekly 25 basis points above the three-month treasury-bill yield ran from 1980 into the inflation-reduction period. Especially in unsettled financial markets, this approach could produce economically trivial variations in the policy rate from the level intended by the Bank, which observers could interpret as policy signals. In 1994, the Bank began setting a 50-basis-point target range for the overnight rate. In 1996, the top of that range became the Bank Rate. This direct setting eliminated the noise inherent in the old arrangement.
William Robson 21
In the second half of the 1990s, problems arising from the Bank’s discretion over the timing of its policy-rate setting became acute. Ability to move at any time likely tends to focus the attention of the Bank, and therefore of financial-market participants, disproportionately on highfrequency data and market sentiment (Robson 2000a; Laidler 2000). Two high-profile episodes occurred in late 1998 and in late-1999/early 2000: in each instance, the Bank reacted to two successive moves in the US Federal Funds rate by making the same moves one day later, which prompted commentary to the effect that the Bank was joined at the hip to the Fed, and positioning on the part of money- and foreign-exchange market participants based on that assumption (Robson 2000b). The Bank responded to these concerns in late 2000 by moving to eight yearly policy-rate announcement dates, and though it reserves the right to deviate from that schedule, it has not yet done so. This procedure makes a decision not to move the target for the overnight rate as explicit as a decision to move it, creating a roughly six-week cycle for policy formulation and communication by the Bank – helpful for the tactics of monetary policy, and also for accountability.14 Expectations and behaviour Divining inflation expectations is a murky business, and assessing when and to what extent private expectations and behaviour came into line with the inflation targets is necessarily tentative. The spread between nominal and real-return long bonds (Figure 1.8) told an encouraging story early in the period, with the implied 30-year inflation rate falling from the three-percent-plus range in 1995 to two percent and even below in 1998–9, and settling very close to two percent from late 1999 through the end of 2003. Since 2003, however, the inflation rate implied by that spread has risen again. While distortions arising from the thinness of trading in real-return bonds may be distorting the measure, there is enough correlation between monthly moves in the spread and actual year-over-year CPI inflation to raise suspicions that long-term inflation expectations in Canada still give appreciable weight to actual recent experience. Survey responses also give a mixed picture. Short-term expectations and reported pricing intentions also appear to show that recent experience influences answers to such surveys. Surveys of longer-term expectations, on the other hand, appear to show that observers take the target as their best estimate of what inflation will actually be: the Consensus Economics surveys currently show expected inflation over both the 2–3 year and the 6–10 year horizons at two percent.15
22
The Canadian Monetary Order 10 9 8
Percent
7 6 5 4 3 2 1 0 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
Nominal Figure 1.8
Real-Return
Spread
Nominal and real-return 30-year bond yields
Political commitment While no formal change in the dual-responsibility arrangements that give the Bank of Canada instrument independence has occurred since 1995, the passage of time with no adverse incidents has helped entrench the current approach. During his tenure as Governor, Gordon Thiessen located the inflation targets clearly within the dual-responsibility framework, in which a directive – and the likely resignation of the Governor – would be the route to over-ruling the targets or the Bank’s method of pursuing them (Thiessen 1998, 31–2). The period when the Bank tended to undershoot its targets coincided with a period of considerable fiscal stress, which may have inhibited a government worried about market reaction from criticizing monetary policy; later in the decade, a return to growth made friction over the conduct of policy less likely. Early in the regime, the threat of unsustainable increases in public debt was not the only point of dissonance with fiscal policy. Key fiscal parameters, most notably the three-percent threshold for indexation of personal income tax brackets, made inflation above two percent fiscally helpful, which led David and me to wonder if the elected government’s commitment to the inflation target might wane (Laidler and Robson 1994, 10). Subsequent full indexation of the tax system has now
William Robson 23
removed that troubling point. Some important official projections, such as those for the publicly funded pension systems, assume that inflation over the long term will be somewhat higher than two percent (OCA 2004a and 2004b), but to see these as expressions of any kind of official contrary view would be a stretch. Some commentators, including David in the past (Laidler 1991b) and former Bank of Canada Governor Crow (2002) have advocated an explicit change to the Bank of Canada’s mandate that would make low or no inflation the explicit, unique goal for policy. Canada’s experience over the past decade suggests, however, that political commitment to the inflation target and instrument independence for the Bank in striving to hit it is an arrangement with some durability. Accountability Turning to the ability of the population and the elected government to apply corrective action if the Bank fails to achieve its target, the two-percent period certainly deserves a passing grade. Aided by the move to fixed policy-rate announcement dates, more regular conferences with programs that explicitly match insiders and outsiders in each session, the creation of a senior Special Advisor position held by outsiders for one-year terms (including David in 1998–9), and other innovations to promote the exchange of views between itself and outside experts, the Bank has improved its communications to financial-market participants and the public, and fostered a higher level of discussion of monetary policy.16 Notable in this effort is the Bank’s transparency with regard to its assessment of the economy. Each policy-rate announcement is followed by a report with extensive commentary, including estimates of the output gap and details of its forecast for quantity variables – and an inflation forecast that, not surprisingly, converges to two percent. On the debit side, Canada is one of the inflation-targeting countries in which inflation outside the target range triggers no formal review or statement by the central bank.17 From the end of 1995 through the end of 2005, inflation was above three percent or below one percent for 24 months, or 20 percent of the time. I know of no occasion when such a deviation from target prompted any parliamentary or other investigation of the Bank’s conduct of policy.18 Resilience Inasmuch as the resilience of a monetary regime is ultimately an empirical matter, the most that can be said of Canadian monetary policy since
24
The Canadian Monetary Order
1995 is that it has survived more than a decade. Reviewing Canada’s economic experience during the period – much of which can plausibly be related at least in part to the new regime – reveals a number of reasons to expect that the regime may persist for a while to come. Certainly, many results in Canada’s case have been positive. Inflation has been close to target for most of the period, so to the extent that low inflation reduces shoe-leather, menu, search, and other coordination costs,19 Canada has reaped such benefits. The success of targeting has naturally reduced inflation volatility, which presumably has further reduced search and other flexibility costs. The distribution of monthly realizations of year-over-year CPI inflation around the target is ambiguous: skewed to the right in a way that might suggest some implementation problems (Figure 1.9), but centred very close to the target and not indisputably non-normal. Indeed, the high success rate in hitting the annual targets yielded a cumulative difference between the actual CPI and the CPI that would have resulted from perfect targeting over the decade of a trivial 0.6 percent. This greater stability has likely lengthened the time horizon for key contracts. The ratio of long-term debt to total business debt rose from 45 percent in 1981 to over 70 percent in 2005. And the duration of average union contracts has increased from approximately 25 months at the end of the 1970s to nearly 40 months in the 2000s (Longworth 2006).
14
Number of months
12 10 8 6 4 2
4 6 8 0 3. 3. 3. 4.
4. 2 4. 4 4. 6 4. 8
2
8 0 2. 3.
3.
6
4
2.
2.
0. 0 0. 2 0. 4 0. 6 0. 8 1. 0 1. 2 1. 4 1. 6 1. 8 2. 0 2. 2
0
Percent Figure 1.9 Twelve-month CPI inflation rates: December 1995 to December 2005
William Robson 25
Although theory and evidence on possible trade-offs between inflation targeting and variability of output are inconclusive, the stability of real GDP under targeting has been good enough to justify a judgement that the population would not reject the regime on those grounds (Figure 1.10). A further element that to date has worked in the regime’s favour is that nominal and (realized) real interest rates have moved in directions generally considered benign (Figure 1.5). Aided by fiscal consolidation, Canada’s interest rates have dropped considerably, most notably falling below US rates, which many observers long doubted could happen, and the volatility of interest rates has also declined since the early 1990s. One key variable that has not behaved in a way that would obviously reinforce support for the inflation-targeting regime is the exchange rate. The two-percent decade saw first a sizeable depreciation of the Canada–US dollar exchange rate, and then an even larger and more rapid appreciation. The fact that movements in Canada’s terms of trade can largely account for these movements does not make them any less irritating to businesses and others who suffer disruptions from movements in relative prices on either side of the Canada–US border. Since the disruptive effects – actual or apparent – of a volatile exchange rate constitute one of the principal threats to the durability of the current Canadian monetary order, this observation is a suitable segue to some closing speculations about its future. 6
Percentage Points
5
4
3
2
1 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05
Figure 1.10 Standard deviation of quarterly change in real GDP (12-quarter window)
26
The Canadian Monetary Order
The future of Canada’s monetary order On balance, my reading of the evidence to date suggests that Canada’s current monetary order merits the term, and will likely persist in recognizable form for long enough that individuals and businesses can count on it. It exhibits the key characteristics its predecessors partly lacked to a remarkable degree (See Table 1.1). Has Canada tamed the genie of fiat money once and for all? One might answer yes to this question with the caveat that the regime is susceptible of improvement short of overturning it. Or one might answer no, on the grounds that the apparent durability of the regime owes much to benign circumstances which, should they not persist, would cause Canada to abandon it. Fine-tuning the goal Suggestions for minor improvements in the regime tend to focus on four issues: the target inflation rate, the choice of price index, the timeperiod over which to measure inflation, and the possibility of giving output variability explicit independent weight in the central bank’s reaction function. Turning first to the inflation rate itself, two percent appears to be a target arrived at by a process of groping along a historical path rather than any analytical process. On its face, two-percent inflation is far enough from price stability to be seriously deficient – the cumulative loss of purchasing power over the period from year-end 1995 to year-end 2005 was more than 21 percent. For someone subject to 50 percent effective marginal tax rates, two-percent inflation reduces the real aftertax yield on a five-percent investment to less than 0.5 percent in real terms. Current Bank of Governor David Dodge has recently tried to renew interest in a longer-term target less than two percent.20
Table 1.1
Canada’s post-war monetary regimes: Key characteristics
Clear goal Technical power Tactical skill Conforming behaviour Democratic support Accountability Resilience
1950s
1960s
1970–80s
No Yes ? ? No No No
Yes? Yes? Yes No ? Yes No
No Yes No No No? ? No
1991–95 1995–now Yes Yes Yes? ? Yes? Yes? Yes?
Yes Yes Yes Yes Yes Yes? Yes
William Robson 27
One familiar objection to a major step down is fear that a lower target would put Canada at risk of accidental deflation in which the zero nominal interest rate constraint would prevent monetary policy from responding effectively. I do not share this concern. Oddly, those who stress it often see the exchange rate as an important influence on domestic monetary conditions and inflation and worry about appreciation-induced deflation (Thorpe 2006, A1), yet appear not to consider that a depreciation might be an effective remedy. More interesting in my view is the constraint arising from the difference between the marginal calculations of benefits and costs with which economists are comfortable and the importance of round numbers in a target that requires democratic support. Lowering the target from two percent to some number such as 1.92 percent or 1.18 percent is not a serious prospect. One percent would have obvious advantages for communication, and zero would be better yet. But such a quantum drop would require political preparation as well as academic backing, and the currently satisfactory performance of the Canadian economy does not suggest that either will be forthcoming soon. A second potential fine-tune would be targeting a difference price index. Some have made a case for choosing a target index with a view to reducing particular costs arising from inflation – by over-weighting stickier prices, prices of non-traded goods or targeting wages, for example – or with a view to improving the trade-off between price stabilization and output stabilization. Similar arguments might justify the kind of core inflation measures that are now familiar, measures that leave out either the prices of some items that are typically more volatile, or month-bymonth price changes of more than a certain (relative) size. Four types of objections raise doubts about one or more of these options. The extent to which a given choice raises welfare or yields a better point on a stabilization frontier in a model inevitably depends on specific assumptions. Empirical conclusions about stabilization necessarily involve inferences from runs of data that must be short to avoid spanning important changes in the goals or implementation of monetary policy. Stripped-down indices may be satisfactory to those who lean towards a theory of price-level determination that depends on inertia and the output gap, but for those who lean towards monetary theories, an index that covers as much of the economy’s money-based transactions as possible makes the most sense. Finally, and far from trivially, indexes far-removed from ordinary experience present communication and accountability problems. Since at the horizon pertinent for inflation targeting, projections for total and core inflation measures
28
The Canadian Monetary Order
tend to converge (Clinton 2006), it is not clear that the central bank or the public gain anything from attention to measures that are less than comprehensive. A more formidable challenger to the CPI is a price index such as the personal consumption expenditure deflator that uses current-period weights rather than fixed past weights. Current weights are clearly more pertinent for actual experience and monetary control, and presumably remove some sources of upward bias in the CPI. But the practical differences between the two indexes are not that great. Both largely depend on the same data collection. And since revisions are undesirable in a targeted index, the advantages the deflator appears to offer because national income statisticians can revise their estimates with knowledge gleaned at a later date would not be available in a deflator used for inflation targeting. My own preference in this area would be to stick with the CPI, but invest in the more frequent surveys necessary to keep the weights up to date – an improvement in the index that would be desirable for other reasons. Proposals to measure changes in the index over a longer time period also merit attention. Annual inflation measures are obviously arbitrary. Shorter time horizons are unattractive: despite the existence of models that find good stabilization properties from a strategy that seeks to hold the price level at whatever it just was, monetary control is not good enough to achieve month-to-month stabilization without a very wide (in annualized terms) target band, and the amount of base drift such an approach would contemplate might appear to sceptical observers to vitiate the goal altogether. Longer time horizons have the corresponding advantages. They better align the target horizon with the period over which the central bank can influence the price level. And they provide an anchor for expectations at a longer time horizon that may – in a neat specific instance of coherence in a monetary order – create ex ante real interest rates that assist the central bank in hitting its target. Adopting a target for the price level itself is an intriguing alternative. Under such a regime, bygones are never bygones, and the central bank must bring the price level back to its target following any deviation. To my knowledge, the joint questions of the error band and whether there should be a specified time horizon for remedying deviations are not well-explored topics. Having confessed ignorance on this point, I will tentatively assert that one-percentage point either side of the price level in a regime that presumed immediate remedy would foster credibility and accountability.
William Robson 29
Finding a target that is clear enough to permit accountability and promote helpful expectations and behaviour limits the choices in this area. A two-year horizon for an inflation target might be saleable. So might a price-level target. To repeat the point made above with regard to a possible lowering of the target number, such a change would require spending political capital and more convincing academic evidence than now exists. Even more than a possible lower target number, I see nothing in the performance of the Canadian economy to create a sense of urgency about changes of this kind. Should less benign domestic or world circumstances make volatility in output more politically problematic, one can imagine a change in the targeting regime that required the Bank of Canada formally to include output stabilization in its policy setting21 – a formula that looks like a Taylor rule, perhaps with weights on the two objectives that would vary with inflation rate. Specifying the appropriate formula to guide the trade-off seems a daunting task, however, and communicating the rule and holding the central bank accountable for achieving it no less so. Worse, any rule derived empirically – a huge task in itself, given the sensitivity of weights to the model specification – would by definition be inapplicable to the world in which it was applied. Technical or tactical collapse The advent of electronic payments clearing technology and the potential some have seen in it for alternatives to central bank money at the core of the financial system have prompted speculation that central banks might lose control over monetary conditions and inflation. Since Canada was an early mover away from fractional reserves towards systems where demand for central bank money arises from very short-term frictions and uncertainties about credit quality in clearing and settlement, it is a test-case for the central banks’ ability to provide a form of money sufficiently attractive to serve as a fulcrum for policy in this kind of environment. As Freedman (2000) had argued, such systems and alternative sources of liquidity do not undermine central bank influence over the very short-term interest rate. Notwithstanding some early running-in problems and disappointment of expectations that the system might run on a zeroaverage cash setting, no single player or set of players domestically or internationally has emerged to take the Bank’s place as a provider of high-power money to Canada’s financial system, and the overnight rate remains an effective tool for the management of Canadian monetary conditions.
30
The Canadian Monetary Order
What if the central bank was still able to control the price of liquidity, but faced a transmission mechanism so unstable as to render that control ineffectual? In view of the controversy over such basic questions as whether money plays any role in transmission, the possibility cannot be dismissed outright. Thus far, ignorance on this level does not appear to have compromised inflation targeting in Canada or elsewhere,22 but the possibility exists that asset-price bubbles and collapses, or violent exchange-rate swings divorced from fundamentals, might present central banks with challenges so great that the question of tactical versus technical powers might appear abstract. Many of these speculations appear to hinge on the old question, already alluded to with respect to the choice of inflation target, about whether the zero nominal interest constraint is a serious obstacle to expansionary policy. For those who think it is, these possibilities might seem important technical threats to central banks’ control of monetary conditions. For those such as David who think it is not, they are not; the challenges is one of tactics. The relationships between money growth, output, and inflation have varied in Canada over the decades. There is nevertheless enough evidence – including some from the past half-decade of inflation targeting – that narrow money’s old tendency to lead output continues, to cast doubt on the view that interest rates are all that matters. A specific challenge to the Bank of Canada’s control that attracts occasional attention is some combination of economic integration with a larger currency area – presumably the US-dollar area – and exchangerate volatility that could lead Canadians to cease using the Canadian dollar in their transactions: a spontaneous dollarization. Again, it seems rash to rule such a possibility out. Canadians’ nervousness over their national identity and often-disappointing relative economic performance against the US have at times provoked predictions that it was under way. As Laidler and Poschmann noted in 2000, however, the fraction of domestic money supply denominated in US dollars in Canada was no higher in the late 1990s (when weakness in the Canada–US exchange rate prompted vehement predictions along those lines) than it had been in the 1970s, and such rise as was evident during the 1990s was in line with what one would expect from the increase in cross-border trade after the Canada–US free-trade agreement of the late 1980s and the North American Free Trade Agreement of the early 1990s (Laidler and Poschmann 2000, 9). Further research on the Canadian dollar’s continued dominance in domestic uses (Murray and Powell 2002) confirmed that spontaneous marginalization or outright replacement of the
William Robson 31
currency the Bank of Canada controls continues to be more a topic for speculation than a clear and present danger. Adopting the US dollar While spontaneous dollarization may not be a serious prospect, unhappiness with the floating exchange rate likely presents the existing Canadian monetary order with its most important challenge. Exchange rate swings that are large, sudden, and apparently divorced from fundamentals cause intense irritation among Canadians exposed to them, and if such swings overwhelmed the Bank of Canada’s tactical ability to deal with them, failure to hit the inflation target would raise questions about the costs and benefits of the inflation-targeting regime. The logical replacement for it would be a monetary order based on the US dollar.23 The closing years of the last century saw many predictions along these lines,24 inspired partly by the spectacle of monetary integration in Europe, and, in Canada, also by discontent with the low level of the Canada–US exchange rate. Up to now, however, a mix of economic and political considerations made the proposition less than attractive. It is commonplace, for example, that joining a larger monetary system offers more to a country with a history of macro monetary instability than to one without, and after a decade of reasonably successful twopercent inflation targeting, the repercussions of exchange-rate volatility would have to be severe indeed to put Canada into that former camp. Canada also has a well-functioning financial infrastructure – not only the clearing and settlement systems already mentioned, but deposit insurance regimes, competition laws, and solvency oversight – that could not obviously be replaced satisfactorily in the event of a Canada–US monetary union (Robson and Laidler 2002). Among the criteria for a monetary order enumerated in this paper, an obvious gap in a regime based on the US dollar would be accountability, since Canada would have no representation in the US Congress or in US presidential politics, and would therefore have no say over the conduct of US monetary policy. My judgement is that economic integration and political comity between Canada and the US would need to be a quantum jump more intimate before this trade-off would become acceptable to a majority of Canada’s population. The obvious alternative would be a return to a pegged Canada–US dollar exchange rate. Canada has not had such a regime since the 1960s, but there is nothing mysterious about it. The goal is consistent with the
32
The Canadian Monetary Order
powers of a central bank, and the political mechanism for achieving it – with the use of the directive if necessary – is straightforward. The Bank would again become a major player in the foreign-exchange market, and its operations on the domestic front would need to adapt to whatever the maintenance of the peg required. The obvious, brief point to make is that fixing the nominal exchange rate forces real-exchange rate adjustments to occur exclusively through adjustments in relative domestic wages and prices. Since Canada’s real exchange under a peg would still be subject to many of the same forces that have moved it in the past, one might reasonably expect that large swings in Canada’s terms of trade or net external demand would eventually create domestic stresses large enough to call the government’s commitment to the peg into question, presenting foreign-exchange traders with the familiar one-way bet. Pegged exchange rates have tended not to be resilient in the past, and there is no reason to think a future attempt in Canada would be any different.
Wrap At the risk of opening myself to a charge of ‘presentism’, I conclude by repeating that the regime of two-percent inflation targeting that has prevailed in Canada since the end of 1995 appears to constitute a monetary order superior to any of its post-war predecessors. Although I have not attempted to rank Canada’s arrangements against those of other countries, I think the current combination of a logical goal, technical and tactical powers on the part of the central bank, coherent expectations, political support, and resilience make it a subject worthy of study elsewhere. Even if changed circumstances do undermine the order, it has already delivered Canadians a decade of monetary stability unparalleled in generations. And if it does persist for another decade or more, it will constitute an impressive legacy for those who played a part in its evolution.
References Bank of Canada. 1991. ‘Targets for Reducing Inflation’ Press Release reproduced in Bank of Canada Review, March, pp. 5–6. —— ‘Renewal of the Inflation Control Target: Background Information.’ Ottawa: Bank of Canada, May. Boessenkool, Kenneth J., David E. W. Laidler, and William B. P. Robson. 1996. ‘Devils in the Details: Improving the Tactics of Recent Canadian Monetary Policy.’ C. D. Howe Institute Commentary 79. April.
William Robson 33 Boessenkool, Kenneth J., and William B. P. Robson. 1998. ‘Buried by a Falling Dollar: The Bank of Canada’s Misguided Interest Rate Hike Promises Recession.’ C. D. Howe Institute Backgrounder. August. Bothwell, Robert, Ian Drummond, and Edward English. 1981. Canada since 1945: Power, Politics and Provincialism. Toronto: University of Toronto Press. Brunner, Karl. 1984. ‘Monetary Policy and the Monetary Order.’ Reprinted in Karl Brunner and Alan Meltzer. Monetary Policy and Monetary Regimes. Rochester: University of Rochester Graduate School of Management. 1985. Clinton, Kevin. 2006. ‘Core Inflation at the Bank of Canada: A Critique.’ Queen’s Economics Department Working Paper, No. 1077. Kingston. May. Crow, John W. 1988. ‘The Work of Canadian Monetary Policy’ Eric J. Hanson Memorial Lecture delivered at the University of Alberta, Edmonton, January 8, 1988, reprinted in Bank of Canada Review (Spring). Crow, John W. 2002. Making Money: An Insider’s Perspective on finance, Politics, and Canada’s Central Bank. Toronto: Wiley. Freedman, Charles. 2000. ‘Monetary Policy Implementation: Past, Present and Future – Will Electronic Money Lead to the Eventual Demise of Central Banking?’ International Finance 3:2, 211–27. Friedman, Benjamin. 1999. ‘The Future of Monetary Policy: The Central Bank as an Army With Only a Signal Corps.’ International Finance 2:3, 321–38. Heymann, Daniel and Axel Leijonhufvud. 1994. High Inflation. London: Oxford University Press. Howitt, Peter. 1990a. A Skeptic’s Guide to Canadian Monetary Policy. C. D. Howe Institute Commentary 25. Toronto. December. —— 1990b. ‘Zero Inflation as a Long-Term Target for Monetary Policy.’ In Richard G. Lipsey (ed.) Zero Inflation: The Goal of Price Stability. Toronto: C. D. Howe Institute. —— 1997. ‘Low Inflation and the Canadian Economy.’ In David Laidler (ed.) Where We Go from Here: Inflation Targets in Canada’s Monetary Policy Regime. Toronto: C. D. Howe Institute. Jenkins, Paul. 2006. ‘Remarks to the Canadian Institute of Actuaries.’ Ottawa, 29 June. www.bankofcanada.ca/en/speeches/2006/sp06-11.html. Kamhi, Nadja. 2006. ‘LVTS, the Overnight Market, and Monetary Policy.’ Bank of Canada Working Paper 2006-15. Ottawa: May. Kennedy, Sheryl. 2003. ‘Monetary Policy during Economic Shocks: Lessons Learned.’ Remarks to the Canadian Association of Business Economists, Kingston, Ont. 26 August. Laidler, David. 1991a. ‘Qualms about Inflation Targets.’ The 1991 Federal Budget. Kingston: McGill-Queen’s University Press for the John Deutsch Institute. —— 1991b. How Shall We Govern the Governor? A Critique of the Governance of the Bank of Canada. Toronto: C. D. Howe Institute. —— 1999. ‘The Exchange Rate Regime and Canada’s Monetary Order.’ Bank of Canada Working Paper 99-7. Ottawa. March. —— 2005. ‘Monetary Policy and its Theoretical Foundations.’ University of Western Ontario Economic Policy Research Institute Working Paper 2005-08. Laidler, David E. W. 2000. ‘Picking up the Beat: Why the Bank of Canada Should Move to a Fixed Schedule for Policy Announcements,’ C. D. Howe Institute Backgrounder 43, Toronto.
34
The Canadian Monetary Order
Laidler, David, and Finn Poschmann. 2000. Leaving Well Enough Alone: Canada’s Monetary Order in a Changing International Environment. C. D. Howe Institute Commentary 142. May. Laidler, David E. W., and William B.P. Robson. 1994. “The One to Three Percent Solution: Canadian Monetary Policy under the New Regime.” C. D. Howe Institute Commentary 59. March. Laidler, David E. W., and William B. P. Robson. 1998. Walking the Tightrope: Canada’s Financial System between a ‘Yes’ Vote and Quebec Secession. C. D. Howe Institute Commentary 102. March. —— 2004. Two Percent Target: Canadian Monetary Policy since 1991. Toronto: C. D. Howe Institute. Longworth, David. 2006. ‘The Crucial Contribution of the Financial System and Monetary Policy to Economic Development.’ Remarks to the Conference of the Association des économistes québécois. Montréal, Quebec. 5 May. Macklem, Tiff. 2001. ‘A New Measure of Core Inflation.’ Bank of Canada Review. (Autumn) 3–12. Murray, John and James Powell. 2002. ‘Dollarization in Canada: The Buck Stops There.’ Bank of Canada Technical Report No. 90. Ottawa. August. Office of the Chief Actuary (OCA). 2004a. Actuarial Report (21st) on the Canada Pension Plan as at 31 December 2003. Ottawa: Office of the Superintendent of Financial Institutions. —— 2004b. Actuarial Report (7th) on the Old Age Security Program, as at 31December 2003. Ottawa: Office of the Superintendent of Financial Institutions. Parent, N. 2002. ‘Transparency and the Response of Interest Rates to the Publication of Macroeconomic Data.’ Bank of Canada Review (Winter), 22–34. Parent, Nicholas, Phoebe, Munro, and Ron Parker. 2003. “An Evaluation of Fixed Announcement Dates.” Bank of Canada Review (Autumn), 3–11. Powell, James. 1999. A History of the Canadian Dollar. Ottawa: Bank of Canada. Robson, William B. P. 1995. Change for a Buck: The Canadian Dollar after Quebec Secession. C. D. Howe Institute Commentary 68. March. —— 2000a. ‘A Little Rhythm Could Help the Bank of Canada Stay Cool: The Case for Regularly Scheduled Bank Rate Decisions.’ C. D. Howe Institute Backgrounder. March 7. —— 2000b. ‘The Bank of Canada Must Use its Independence – Or Lose It.’ National Post, 19 May. Robson, William B. P. and David E. W. Laidler. 2002. No Small Change: The Awkward Economics and Politics of North American Monetary Integration. C. D. Howe Institute Commentary 167. July. Roger, Scott and Mark R. Stone. 2005. ‘On Target? The International Experience with Achieving Inflation Targets.’ IMF Working Paper WP/05/163. Washington. August. Thiessen, Gordon. 1998. ‘The Canadian Experience with Targets for Inflation Control.’ 1998 J. Douglas Gibson Lecture, Queen’s University, Kingston, Ontario, 15 October. Reproduced in Gordon Thiessen. 2001. The Thiessen Lectures. Ottawa, Bank of Canada. Thorpe, Jacqueline. 2006. ‘Dodge mulls change that could boost rates: Inflation target ripe for adjustment.’ National Post, March 22. A1.
Discussion Peter Howitt
Bill Robson’s collaboration with David Laidler at the C. D. Howe Institute, which started in 1990 and continues to this day, has promoted useful public debate about Canadian monetary policy, and has produced a published record that is of great value to students of the subject. Their 1993 book won the Canadian Economics Association’s Purvis Prize for work on Canadian Economic Policy, its 2005 successor won the Donner Prize for the best book on Canadian public policy, and their frequent C. D. Howe Commentaries have provided Canadians with a continual stream of first-rate monetary policy analysis that I suspect is unrivalled in any other country. In addition to these writings, Bill’s creation of the C. D. Howe shadow Monetary Policy Committee, of which David is a member, has kept the heat on the Bank of Canada and kept public interest in monetary policy alive during a period when the macroeconomy was performing so well that it would have been easy to let things slide. So when Bill speaks on Canadian monetary policy I listen. This paper takes a broader historical and analytical perspective than most of Bill’s writings, one that is based on David’s concept of a ‘monetary order’. It does an excellent job of describing and analyzing the evolution of that order, and certainly belongs on the reading list of any course on Canadian monetary policy. Discussants are supposed to take issue, but there aren’t many issues on which I disagree with this paper. Accordingly my discussion will be brief. Although the paper describes the evolution of policy starting in the 1950s, most of its analysis focuses on the inflation-targeting era that started in 1991. Inflation targeting seems to have been a big success; inflation has come down, and so have the variance of inflation, the variance of output, and the variance of interest rates. Although no one has yet found conclusive evidence that these desirable macroeconomic 35
36
The Canadian Monetary Orde
outcomes can be directly attributable to inflation targeting, it is hard to argue with a policy that has been associated with them. So it is also hard to argue with Bill’s conclusion that ‘Canada’s current order appears robust enough to warrant confidence in its durability.’ Moreover, the concept of a monetary order allows us to consider more than macroeconomic outcomes when evaluating policy regimes, because it is inherently multidimensional. It goes beyond monetary policy narrowly conceived as a mapping from the Central Bank’s information to its instrument setting, and involves the goals of policy, the technical powers and tactical skills of the central bank, the degree to which policy and expectations are aligned, political support for the policy, accountability of the central bank, and the resilience of the policy regime. Bill points out that all the above seem to have improved under inflation targeting. I particularly like the emphasis placed on tactical skills, an aspect of monetary policy that deserves more attention than it has received so far in academic literature. The aspect that Bill focuses on is the manner in which the Bank of Canada goes about setting short-term interest rates, which it now does on a pre-announced schedule of fixed action dates. He points out that the adoption of a fixed schedule has made monetary policy more transparent and more predictable, and has helped to focus private expectations on the fundamentals. I would add that the fixed schedule also seems to have improved decision making within the Bank. There is now a systematic decision-making process leading up to each fixed action date, in which forecasts are made and discussed, risks are assessed, past decisions are re-evaluated, and the entire organization is focused on and feeds into the ultimate decision.25 Not only does this ensure that each decision is carefully considered from a variety of different viewpoints, with no rush to judgement, but the fact that the same process is repeated regularly facilitates learning by doing within the organization. Orchestrating such an elaborate team effort was just not possible when no one knew when the next decision was going to be made. The first item on Bill’s list of conditions for a good monetary order is having a well-defined goal. On this criterion, he gives the current regime of inflation targeting top marks. Having a clear numerical target for inflation, being committed to coming within one-percentage point of that target over the medium term, and having the central bank’s political masters committed to that policy, has yielded great benefits in terms of transparency, and in terms of bringing expectations into line with the objectives of policy. To this I would again add that the clarity of the Bank’s objective probably also helps in the Bank’s internal
Discussion
37
decision making. As Mervyn King has testified,26 committee members are less likely to engage in long-winded digressions when everyone is clearly focused on a common goal. But is the goal really all that clear? Is monetary policy really focused on achieving the target rate of inflation within the stated time horizon, to the exclusion of all other considerations? Clearly not. Like any other central bank, the Bank of Canada takes into account the level of real activity when deciding on the appropriate course of action, and not just because the future behaviour of inflation depends partially on the state of the economy but also because the Bank does not want to precipitate or aggravate a recession. Moreover, it is widely understood that any inflation targeter like the Bank of Canada has more than the single objective of controlling inflation. This is why writers like Svensson (1977) can analyze an inflation-targeting regime with the seemingly contradictory assumption that the Bank is seeking to minimize a loss function involving not only inflation but also the output gap. As Chuck Freedman (2001) has observed, this assumption helps us to understand how an inflation-targeting central bank chooses its response to supply shocks that confront it with a short run inflation–output trade-off. So how could the Bank of Canada have possibly realized the benefits of a well-defined goal and a transparent policy when it is transparently obvious that its real goals are different from the only clear goal that it has publicly defined? I think the way to make sense of this is to recognize that what distinguishes inflation targeting from other regimes is not that it has a single goal but that it has a single medium-term goal, where the medium term is approximately six to eight quarters, and that the goal is an attainable nominal goal which is clearly understood. In terms of the Svensson formulation, what is really important in the central bank’s objective function is not the weight assigned to squared deviations of inflation from its target versus the weight assigned to the squared output gap, but rather the fact that there is indeed a clear target for inflation. The rest of the objective function will affect how it manages short-term fluctuations of the output gap and of inflation around its target, but it is not clear how much the central bank can influence these fluctuations, and in any event reacting one way or another to supply shocks is a second-order issue compared to controlling the mediumterm rate of inflation, which the bank clearly can influence. So yes the Bank of Canada cares about more than the rate of inflation, but only in the sense that a conscientious dentist cares about more than her patients’ dental health – otherwise why would she bother using anaesthetics?
38
The Canadian Monetary Orde
Despite his enthusiasm for inflation targeting, Bill is cautious in his assessment of its future. Asset price or exchange rate volatility might overwhelm the Bank’s dwindling ability to control inflation and/or erode public support for the current regime. I believe there is another important problem with the regime, one that threatens to undermine it even in the absence of a major financial shock, and it has to do with the problem of finding leading indicators for inflation. Inflation being a highly inertial variable, the policy of inflation targeting depends not on reacting to adverse movements in inflation but on anticipating and taking preemptive action to forefend latent future movements. Once you see a rise in inflation it is too late to do much about it for the next year or two. The problem is that the more successful the policy is in anchoring peoples’ expectations of inflation, the harder it becomes to forecast inflation. This is probably why the Bank is finding now that all its leading indicators of inflation, from the output gap to core inflation to private inflation forecasts, are losing their predictive power. As Bill has documented, private inflation forecasts two years and more out are now pretty much stuck at two percent. And no macroeconomic aggregate is going to have much predictive power over a variable that has been two percent plus small change for many years. Thus in effect the Bank of Canada is now trying to control inflation without feedback. If the economy were currently operating above capacity, thus building up pressure that will manifest itself in a serious inflation in a few years if left unchecked, how would anyone know? Not by monitoring the macroeconomic aggregates that have lost their predictive power, and not from surveys of professional forecasters who have seen inflation at two percent for so long that that’s what they will continue to predict until experience proves them wrong. We will only know when we see a prolonged rise in actual inflation, by which time most of the damage will have been done. This, it seems to me, is the most likely source of future difficulty for the current regime, and also the most challenging problem for monetary economists to address.
Notes Essay for the Festschrift in Honour of David Laidler, University of Western Ontario, August 2006, post-conference draft: 10 December 2008. 1. I thank Bob Dimand, Peter Howitt, David Laidler, and Ben Tomlin for their comments and absolve them of blame for any remaining errors or defects. 2. But for the replacement of ‘the Dominion’ by ‘Canada’, this passage in the current version of the Act reads identically today.
Discussion
39
3. Jenkins (2006) provides a clear account of the Bank’s current interpretation of its mandate in explicit contrast to some other views of what monetary policy can and should do. 4. Fiscal deficits large enough to pose a threat of monetization, for example, could render a regime incoherent (Heymann and Leijonhufvud 1994). 5. An interesting special case is a seceding state that commits in advance to maintain a monetary regime, only to face a crisis of confidence at the time of secession that undermines the commitment. On the Quebec–Canada example, see Laidler and Robson (1991) Robson (1995), and Laidler and Robson (1998). 6. A well-remarked specific example being the network externalities that sustain the use of a money in exchange even when high inflation has driven it out of many other uses. 7. One might defend this program – like the later, and even less consequential, ‘six and five’ program in 1983 and 1984 – as an explicitly temporary measure to aid disinflation, unlike attempts elsewhere to reduce inflation permanently by non-monetary means. The ‘cost–push’ views prominent in justifications for the policy, however, can only have exacerbated the confusion about inflation’s monetary roots. 8. The overall effect may have been unhelpful: as I elaborate below, this approach may have added noise to the signal intended by policy-rate changes. 9. Real-time gross-equivalent because it processes individual payments in real time on a gross basis and settles on a multilateral net basis at the end of the day. Kamhi (2006) describes the system and its operation. 10. Initially, the Bank expected the LVTS as a whole to operate on zero net balances. Happily for those who find a money multiplier resting on a zero base conceptually unsatisfactory, central bank money turned out to be so uniquely desirable that the system works better with a small positive net balance. 11. Although the core measures making most sense in such models are unlikely to be the same ones adopted for ease of communicating to the public. 12. Recent Bank of Canada commentary refers to the former movements as Type 1 and the latter as Type 2 – terminology that is aptly obscure. Such analysis often speaks of ‘portfolio shocks’ as driving Type 2 movements. I find portfolio adjustments with implications for aggregate demand and inflation easier to imagine than those without, since one would expect changing views about Canada’s relative attractiveness for investment, whether driven by changes in assessments of Canada particularly or by changes in assessments of alternatives, to influence the terms on which Canadians can obtain funds, and hence actual or incipient domestic demand. 13. As some pointed out at the time (Boessenkool and Robson 1998). See Laidler and Robson (2004, 121–3) for commentary on this episode, and Kennedy (2003) for a view from inside the Bank on its lessons. 14. For the Bank’s assessment of the move to fixed announcement dates, see Parent (2002) and Parent, Munro, and Parker (2003). 15. See www.bankofcanada.ca/en/rates/indinf.html, visited 11 July 2006. 16. One private-sector innovation possible in the new environment that I would deem notable is a ‘shadow’ committee, the C. D. Howe Institute’s Monetary Policy Council, which publishes recommendations of individual members,
40
17.
18.
19. 20.
21.
22.
23.
24. 25. 26.
The Canadian Monetary Orde and the median for the group, four working days before each of the Bank’s policy-rate decisions. Roger and Stone (2005, 12 and Table 3) state that eight inflation targeting countries, including the first six to adopt this regime, require formal central bank explanations for inflation outside the target range – a classification that does not match my understanding of the arrangements affecting the Bank of Canada. I also know of no policy-oriented modeling work at the Bank in which the Bank’s utility function contains any discontinuities when inflation moves outside the target range. See Howitt (1990b and 1997) for discussion of these issues before and after the establishment of targets. In a December 2005 speech, he said that while recent work at the Bank provided somewhat more support for a lower target than had been the case at the last review of the target in 2001, the evidence did not make a compelling case for a lower target at the 2006 renewal (Thorpe 2006, A1). The 2002 renewal of New Zealand’s inflation targets contained an informal prescription of this sort, which may amount to nothing more in practice than its counterpart in the Bank of Canada’s mandate. Commenting on Woodford’s elaboration of models where fluctuations in money have no effect on the output gap, David remarks: ‘the practical problems of implementing Woodford’s recommendations that stem from the difficulty of actually measuring the output gap and estimating the neutral value of the interest rate … [have] so far … not led inflation-targeters into serious policy mistakes, but one wonders just how long their luck will hold out’ (Laidler 2005, 8–9). Abandoning an independent currency would only be a sensible response to another overwhelming technical or tactical challenge if the same challenge was not undermining monetary control in the alternative currency area. Laidler and Poschmann (2000, 5) provide some citations to this literature. See Macklem (2002) for a detailed description of the process. King (2005, pp. 13–14).
References Freedman, Charles. ‘Inflation Targeting and the Economy: Lessons from Canada’s First Decade.’ Contemporary Economic Policy 19 (January 2001): 2–19. King, Mervyn. ‘What Has Inflation Targeting Achieved?’ In The Inflation-Targeting Debate (eds) Ben S. Bernanke and Michael Woodford, 11–16. Chicago: University of Chicago Press, 2005. Laidler, David E. W. and William B. P. Robson. The Great Canadian Disinflation: The Economics and Politics of Monetary Policy in Canada, 1988–93. Toronto, Canada: C. D. Howe Institute, 1993. —— Two Percent Target: Canadian Monetary Policy Since 1991. Toronto, Canada: C.D. Howe Institute, 2004. Macklem, Tiff. ‘Information and Analysis for Monetary Policy: Coming to a Decision.’ Bank of Canada Review (Summer 2002): 11–18. Svensson, Lars E. O. ‘Inflation Forecast Targeting: Implementing and Monitoring Inflation Targets.’ European Economic Review 41 (June 1997): 1111–46.
2 Inflation Targeting in Canada: Optimal Policy or Just Being There? Peter Howitt
2.1 Introduction David Laidler has had the good sense not to have taken too seriously the notion that people are rational maximizers, always acting under rational expectations. One of the central themes of his work is that money is a device for economizing on the costs of processing information. People use it as a buffer stock that automatically absorbs unforeseen changes in income and expenses without the need for deliberation. They also use it as a unit of account, measure of value and standard of deferred payment because it is convenient to use, conventional and easily understood, even if this seems to introduce biases and inefficiencies into their decision making and even if economists can think of better measures and standards.1 In this respect, David stands apart from the mainstream of macroeconomics, which has been characterized over the years by what he has called an irrational passion for dispassionate rationality. But unlike many other critics of unbounded rationality, David does not put his ideas forth as an attack on free market economics. On the contrary, he has insisted that money is part of the institutional mechanism that helps a decentralized market economy to coordinate people’s activities. For example, when demand increases in some sectors and decreases in others, one of the first signals that dealers receive is an unanticipated change in cash holdings. Until it is clear that this is going to be permanent, or at least more than momentary, there is no point in changing production, employment or pricing decisions. Buffer stocks allow firms to keep these activities on schedule without having to make portfolio adjustments every minute of every day, adjustments that would disrupt economic activity in other markets, causing disequilibrium to spread through the system. 41
42
Inflation Targeting in Canada
To assume that people use convenient heuristic devices for arranging their affairs in an uncertain world is not to say that they lack intelligence. On the contrary, it is to say that they are smart enough to realize the folly of acting always in Bayesian fashion by fabricating a model of the world and choosing a plan that would be optimal under the incredible assumption that the model was a true representation of reality. Not only would this be likely to lead them astray, but Bellman’s curse of dimensionality would render the task of computing a solution infeasible by any imaginable twenty-first-century computer unless the model was made tractable by other convenient heuristic devices, like assuming all agents identical, only two possible states of the world, Cobb–Douglas aggregate production functions, and so forth. Coping with economic life in an economy that mystifies even those of us paid the most to understand it requires simple, robust behavioral rules that are easily implemented and well adapted to one’s economic environment, not cumbersome ones whose success depends on far-fetched assumptions. The adjustment mechanisms of a modern economy are largely driven by such rules. Taking money seriously, as David has argued, involves looking at the role that money plays in these mechanisms and rules, a role that has evolved differently across time and space depending on particular historical circumstances. The adaptive, historically bound nature of human behavior that David has portrayed in his writings on money is also at the heart of the philosopher Andy Clark’s book Being There: Putting Brain, Body, and World Together Again, an account of recent developments in cognitive science and artificial intelligence, developments exemplified by the idea of “neural networks.” Clark’s thesis is that human intelligence is not to be thought of as an abstract reasoning capability joined to a memory bank of facts, but rather as a device for controlling the body’s varied set of adaptive behaviors in a way that helps the body cope with the particular environment it finds itself in. Clark calls this view “embodied, environmentally embedded cognition.” It portrays intelligence not as a central computer program solving a well-defined maximization problem but as a decentralized network of autonomous neurons that interact with each other, often sending conflicting messages, and often competing to execute the same task. Intelligence emerges not from the capacity to solve planning problems but from simple chemical reactions that reinforce the neural
Peter Howitt
43
responses that have been associated with improvements in the body’s well-being and weaken the responses that have not. Clark points out that what sorts of neural reactions constitute an improvement is not something that one can evaluate independently of the environment in which the body finds itself, or independently of what other processes are at work. He also points out that the intelligent human is always working to economize on neural resources by manipulating the environment, as when we tie strings around our fingers or make entries on a to-do list in order to remember facts which a computer-like mind would have stored internally; he might have added that the intelligent human relies also on money as a way of coping with an uncertain future and keeping track of material resources. Finally, Clark points out that the process of human adaptation is not guided by an internal model of the world which the brain takes as defining the constraint set for optimization, but rather it consists of simple rules for acting in ways that cope quickly and effectively with environmental hazards such as the presence of predators, the need for food and so on; these rules may sometimes make use of internal representations but typically they need to operate much faster than the construction and use of any such representation would allow, just as people need to make economic decisions faster than would be possible by the use of dynamic programming. The title of Clark’s book derives from a quotation that he attributes2 to Woody Allen: “Ninety percent of life is just being there.” The idea is that intelligence depends not just on a person’s internal capabilities, which are inherently limited, but also on the external physical and cultural environment, which contains external devices which the mind can use as props, to be manipulated with clever tricks and strategies. Thus, “Language and culture … emerge as advanced species of external scaffolding ‘designed’ to squeeze maximum coherence and utility from fundamentally short-sighted, special purpose, internally fragmented minds.” One of David’s contributions to economics is to point out how money plays just such a scaffolding role. In this essay, I argue that the adaptive model of human behavior which Clark presents sheds useful light on much else that David has written about, not only monetary theory but also monetary policy, its interrelationship with monetary theory, the evolution of Canadian Monetary Policy, and most particularly, the policy of inflation targeting which the Bank of Canada and many other central banks around the world are now practicing.
44
Inflation Targeting in Canada
2.2 Monetary policy and adaptation 2.2.1 The two tasks of monetary policy A central bank is the sole issuer of the base money on which a country’s monetary and financial systems rest. Because of this it has two central tasks to perform. First, it needs to ensure that the supply of base money varies enough from day to day so as to avert financial panics and collapses of the sort that used to characterize the seasonal shortages of means of payment in the US in the nineteenth and early-twentieth centuries and the periodic banking crises in England in the nineteenth century. This task requires a central bank to stand ready to act as lender of last resort from time to time. It also requires the central bank to stand ready on a regular basis to buy or sell short-term financial instruments at prices that do not fluctuate wildly from day to day. In effect, a central bank oversees a country’s money market in much the same way, and for much the same reason, that more conventional commercial enterprises manage the markets for the products they buy and sell. That is, to make the market function efficiently it must provide assurance to other transactors that they can trade when they want, on reasonably predictable terms. The other task of a central bank is to ensure the long-run value of the monetary unit. To use a well-worn metaphor, a central bank is the only agent in an economy in a position to provide a “nominal anchor” for the unit in terms of which contracts are written, accounts are kept and prices are quoted. It does this by controlling the growth in the supply of base money.3 We have known since Edgeworth, Wicksell and Keynes that unless the supply of base money is controlled, the overall supply of money and credit cannot be controlled, and we have known for even longer that unless the supply of money is controlled the price level cannot be controlled. Frequently, these two central tasks of a central bank conflict with one other. To maintain a well-functioning money market, the bank must often dampen interest-rate fluctuations by expanding or contracting the monetary base to meet the market’s day-to-day demands. But to avoid long-run inflation, it must not supply whatever is demanded without limit. Thus, there is a constant tension between the two tasks. In carrying out this difficult balancing act, a central bank must deal with three major sources of difficulty: interference from partisan political interests, an ever-diminishing ability to control nominal spending and lack of understanding of the process by which its own actions affect the course of the price level.
Peter Howitt
45
2.2.2 Politics and central banking As for political influence, the central banks that have been most successful in controlling inflation have been those with the greatest degrees of independence from the executive and legislative branches of government, like the late-twentieth-century central banks of Germany and Switzerland.4 The simplest explanation for this phenomenon is that governments find it hard to resist demands for deficit spending when they can rely on a central bank to finance the deficits by printing money, whereas the costs of inflation that might result from increased monetary expansion may lie too far in the future to influence their decisions. Central bankers that do not enjoy such independence often spend much of their time and effort finding ways to gain some degree of de facto autonomy that would help insulate them from political pressure. The infamous Coyne affair presents a case in point. When it was resolved by the resignation of the governor, his successor, Louis Rasminsky, took office only after securing agreement from the government concerning the nature of the joint responsibility of the Bank and the government for the conduct of monetary policy. A crucial part of this agreement was the “directive clause”, according to which if ever the governor and the minister of finance were unable to resolve their differences in private, the minister could exert ultimate authority, but only by issuing a directive that was to be tabled in the House of Commons. Moreover, the directive would have to specify exactly what instructions the Bank was required to follow, and it would have to specify a limited time period during which the instructions would be binding on the Bank. Although this clause affirmed that the Bank was de jure subordinate to the minister of finance, all the more so because of the implicit understanding that a governor faced with such a directive would have no choice but to resign, nevertheless it can be argued that it enhanced the Bank’s independence in the sense of making it more difficult for a government to exert pressure on the Bank to pursue inflationary policies for partisan political reasons, the sort of pressure from which the Bank is most in need of protection. That is, the threat of dismissal would not be a very credible one with which to exert such pressure because the government would be required to make its reasoning explicit and very public.5 2.2.3 Monetary control The ever diminishing control over nominal spending that I identified above as the second major difficulty that central banks must deal with arises because of the increasing depth and sophistication of financial
46
Inflation Targeting in Canada
markets in advanced countries. Because of these developments, the base money that a central bank controls is a vanishing part of a country’s entire stock of liquidity. This poses the question of how a central bank can continue to provide a nominal anchor to a system which increasingly resembles Wicksell’s imaginary pure credit economy. When the Bank of Canada wants to tighten monetary conditions, it raises the “operating band” whose upper limit is the bank rate and whose lower limit is the rate the Bank will pay on positive balances of participants in Canada’s Large Value Transfer System (LVTS).6 This effectively raises the overnight interest rate. If tightness is expected to persist, a chain of substitution will transmit the rise to other interest rates. In turn, this rise in domestic interest rates will produce at least an incipient rise in the demand for Canadian dollars on the foreign exchange market as investors shift from foreign to Canadian securities, and the result is likely to be a stronger Canadian dollar. By affecting interest rates and the exchange rate, Bank of Canada actions can thereby exert an influence on the flow of aggregate expenditures. Businesses and households facing higher interest rates and/or tougher credit terms will tend to spend less. Businesses, including financial intermediaries, will also be less willing by themselves to grant credit to their customers, which will reinforce the effects on expenditures by reducing the overall availability of credit7 in the economy. How this works will depend very much on how expectations react to the tightening of policy. If investors take the tightening as a signal that the Bank is resolved to stem potential inflationary pressures that will require it to maintain its stance for some time, the anticipation of persistent tightening will hasten and amplify the rise in other interest rates and in the exchange rate. Without supportive expectations, however, a central bank has only limited scope for controlling the level of aggregate expenditures. If no one thought that interest rates beyond the overnight rate would be affected by the Bank’s tightening, there would be little effect on the longer-term interest rates that influence spending decisions in the economy. Nor would there be any point in the Bank trying to intervene directly in longer-term asset markets with the hope of affecting those rates through open market operations without supportive expectations, because financial markets have grown too large in relation to the Bank’s balance sheet. Thus to produce a major effect the Bank must make people believe that the tightening it induces by raising the operating band will continue until it does have an effect. In other words, the “announcement effect” of monetary policy is becoming
Peter Howitt
47
more and more the primary channel through which a central bank can control aggregate spending. 2.2.4 Understanding the economy This announcement effect is a psychological phenomenon that rests on prevailing opinion. Exactly how it works is not something that we clearly understand, which leads me to the third major difficulty of central banking, namely the difficulty of knowing how central bank actions (and pronouncements) affect the ultimate objectives of policy. For we do not have a good idea of how variations in interest rates, exchange rates, the supply of liquid assets, the availability of credit and the other variables that are more or less directly affected by monetary policy in turn affect the level of aggregate demand in the economy. Nor do we understand in much detail how variations in aggregate demand affect the overall level of economic activity and the course of the price level. On the basis of accumulated evidence, we can assert confidently that monetary policy affects these ultimate targets with a long and variable lag, but that isn’t much help in negotiating the narrow path between meeting the needs of trade and controlling inflation. Consider for example what we know about the price level. Almost any undergraduate textbook you can find nowadays will tell you that it is determined by the intersection of an aggregate demand curve and an aggregate supply curve. But when you probe the underpinnings of these aggregate constructions, you find little reason to believe in their existence. Why, for example, should the aggregate amount of planned expenditures be a decreasing function of the price level? Certainly not because of the quantitatively insignificant Pigou effect. Instead the textbooks typically rely on the Keynes effect; as the price level falls, then the real supply of money in the economy will be larger in relation to its demand,8 and this in turn is supposed to cause interest rates to fall, thus inducing an increase in investment expenditures. A cursory examination of the literature on investment demand shows, however, that of all the purported determinants of investment demand, the rate of interest is the one whose effect can be identified with the least confidence. Moreover, what reason do we have for thinking that the rate of interest responds to the excess demand for money rather than the excess demand for loanable funds? In any event, since the dominant element of “money” in a modern economy is trade credit rather than the bank money recorded in official statistics, why should we think that a rise in the price level would not be offset by an endogenous increase in money? The one thing we can be sure of is
48
Inflation Targeting in Canada
that a rise in the price level will result in a real transfer of wealth from creditors to debtors, because of the ubiquity of nominally denominated debt contracts. But this effect seems far more likely to result in a rise in expenditures than a fall. Direct evidence on the price-formation process9 suggests that the most common strategy of price-setters is that of full-cost pricing. Individual prices are typically set as a fixed mark-up over a long-run average of perunit cost. The major exceptions to this rule arise in the case of import prices, which are often “priced to market”, and in the case of basic commodities whose prices vary daily on organized exchanges. Thus it seems that the first place in which a change in aggregate demand would affect the overall level of prices would be through its effects on basic commodity prices and on labor costs that respond to fluctuations in the derived demand for labor. However, volatile commodity prices typically have a small weight in a comprehensive measure of the price level, and the subsequent feedbacks on the price level that work through costs of manufactured goods to which these commodities are an intermediate input are slow to develop because most price-setters do not want their prices to vary with every short-run fluctuation in costs. And the effects on wage costs are typically slow to develop because of the prevalence of long-term nominal contracts in labor markets. Even after basic commodity prices and wages have started to respond to the rise in aggregate demand, the price level may still not be affected, because of the fixed overhead costs prevalent in manufacturing. That is, since the typical firm operates under conditions of decreasing average cost, the rise in demand will at first cause unit costs to fall, not rise. An increase in basic commodity prices and labor costs will offset this, but it will have to persist for a long time before a lot of firms will perceive that, all things considered, they have experienced a significant and persistent enough rise in unit cost to warrant raising their own prices. Moreover, once a significant number of prices start to rise, because price-setting is not synchronized, and because the main element of cost to most firms is the price of the intermediate goods they buy from other firms, the response of the overall level of prices will be drawn out over a long period of time. Each firm will wait until enough of its own suppliers have raised their prices by enough, and for long enough, before reacting. What appears like a small delay at the individual level can easily amount to a very attenuated process when aggregated.10 Because of this, the response of the overall price level, far from being the instantaneous equilibration of demand and supply, will be a delayed and sluggish response to changes in demand and supply. This sluggishness
Peter Howitt
49
will be amplified by the fact that workers will respond after the fact to rises in the cost of living, and to each others’ wages, by insisting on ex post compensation in the form of higher wages for themselves. Every step in this multidimensional wage-price spiral takes time. Thus, as countless econometric studies have confirmed, inflation will be a highly inertial process – hard to get started and hard to slow down once started. The story I have sketched above sounds reminiscent of the augmented Phillips curve of textbooks, according to which inflation will accelerate whenever unemployment is below its natural rate. However, the bulk of econometric evidence presented in the recent symposium on the Phillips curve published in the Winter, 1997, issue of the Journal of Economic Perspectives provides almost no support for the simple relationship described in the textbooks. Nor does it offer any good empirically supported alternative analytical framework for understanding in detail what determines the course of inflation. 2.2.5 Central banks as adaptive agents In summary, central banks are faced with the problem of trying to regulate a variable (the rate of inflation) whose variations they cannot control except with a long and poorly understood lag and with an uncertain degree of control, and whose movements, once begun, are exceedingly difficult to reverse. They cannot just react ex post to unwanted changes in the variable, because once a significant inflation has arisen it is too late to avoid many of the consequences. They must instead try somehow to anticipate those movements. And the only way this can be done is to search for “leading indicators”, that is for statistical clues that will allow them better to foresee incipient trends in inflation. Over the years, this has led central banks to adopt different strategies using the level of market interest rates, measures of liquidity, exchange rates, the term structure of interest rates, the money supply, commodity prices, measures of unit labor costs, monetary-conditions indices …, in their search for a solution to their dilemma. They have followed fads in monetary theory, have deployed various econometric techniques to estimate the temporal structure of the wage price spiral, and so forth. In doing so, they have been adapting to new experience and to developments in economic theory, which in turn have adapted to these different policy experiments. Moreover, the economy about which central banks are trying to learn keeps changing. Indeed, not only has there been extensive technological change taking place in finance and commerce, but the very attempt to use leading indicators in the control of inflation has degraded the quality of these indicators, in accordance
50
Inflation Targeting in Canada
with the “Lucas critique.” As central banks adjust to evidence about the economy, the economy adjusts to these attempts. Thus a central bank is constantly engaged in a social process of adaptation and learning, a process in which it is just one of the actors, and in which none of the participants have a clear understanding of how the process works or how it is evolving. Some economists would argue that as a profession we have little to say about such a process. And while there is considerable truth in this view, it does not imply that macroeconomists should focus their attention on other aspects of the economy. For the same can be said about almost any other aspect of the working of an economic system taken as a whole. Somehow or the other, free enterprise economies have developed mechanisms by which economic problems get solved through the decentralized actions of countless individuals, each acting on the basis of private information and motivation, and none of them with a clear understanding of the overall system in which their actions play a part. We might not yet have much understanding of how this works. But at least we can try to learn from experience. The same can also be said of central banks, except that they can go one step further. They can try to learn not just from experience but also from experimentation. And that is what has happened over time, as central banks around the world have experimented at various times with monetary targeting, interest-rate pegging, exchange rate control, inflation targeting, and so on. Moreover, it appears as if they have learned something along the way, at least to judge by the way that inflation has come down in the past two decades in almost all OECD (Organization for Economic Cooperation and Development) countries, to the point where inflation ceases to be a major concern, outside of the financial markets in which participants eagerly await inflation statistics not because inflation is a major problem but because it might allow them better to anticipate movements in monetary policy and hence movements in the interest rates and exchange rates affected by central bank actions. In this respect, central banks are aptly characterized by Clark’s adaptive model of behavior, since they must find rules of operation that react quickly to ongoing developments, and adapt to changing circumstances, without the aid of a reliable internal representation of reality. The job of central banking is not just to find the right model of the economy and use it to compute optimal instrument settings but rather to design mechanisms for controlling how its instrument settings are going to be regulated, mechanisms that will not just be appropriate for a particular
Peter Howitt
51
environment but which will also adapt to an environment that changes unexpectedly, often in response to its own changes. This is why even though all central banks in the world make extensive use of macroeconomic models in the process of instrument setting, the model is typically just one input to the process, often supplemented or even overridden by judgment and informal analysis.11 It is also why the models being used keep evolving over time – in Canada from RDX2 to QPM and now TOTEM, a “dynamic stochastic general equilibrium” model much like those in one stage or another of development in just about every central bank in the world. Moreover, just as the neurons being controlled by human intelligence often send conflicting messages that need to be resolved quickly into an action, so the various departments and officials of a central bank often send conflicting messages, and some method must evolve for resolving these into real-time action. All this in a hazardous external political and economic environment that must somehow be manipulated to if the limited and fragmented intelligence embedded within the institution is to result in coherent outcomes. 2.2.6 The case for discretion rather than rules The main implication to draw from the above view of the nature of central banking is that a central bank could not function effectively if it were bound by rigid rules that determined its instrument settings in every conceivable circumstance. It must be free to use discretion, not so that it can react to changing indicators of the economy, which a complicated enough rule would allow, but more importantly so that it is free to change its strategies when the evidence shows that these strategies aren’t working or that the economy has changed in unforeseen ways, and to adopt promising new ideas for achieving the conflicting goals of monetary policy. To illustrate this case for discretion rather than rules, consider the Canadian experiment with “Monetary Gradualism”, which was started in the fall of 1975 with the goal of bringing inflation down by use of targeted reductions in the growth rate of M1. This strategy was in conformity with the then academically dominant theory of monetarism, so much so that Milton Friedman (1975) described the speech in which Governor Bouey announced the program as “the best speech I have ever heard from a central banker.” However, the policy foundered when it became apparent that the relationship between M1, the level of interest rates and the level of nominal income, whose stability was crucial to the effectiveness of the policy, had begun to shift unpredictably.
52
Inflation Targeting in Canada
To some extent bad luck may have played a role in the failure of monetary gradualism. For it was shortly after the initiation of this program that technological innovation in banking made it economical for chartered banks to offer daily interest checking accounts, which fell outside the definition of M1. Similar technical developments made it profitable for banks to begin offering sweep accounts that reduced the need for medium- and large-size business firms to hold their transaction balances in current accounts that were included in M1.12 However, in retrospect, it seems that there would probably have been a problem even without bad luck. For, as Charles Goodhart has argued, in every country in which a similar monetary targeting regime was adopted, the empirical “demand function” for the targeted aggregate soon began to fall apart. The failed attempts at monetary targeting in Canada and elsewhere illustrate well the need for a central bank to retain the flexibility to learn from mistakes. If the Bank of Canada had been forced to adopt a binding rule in 1975, it is hard to see how anyone at the time could have picked a better rule than the one it voluntarily chose. But then what would have happened when the Bank learned that the rule was doing more hard than good? Presumably, the Bank would have tried to persuade whatever authority was charged with enforcing the rule to revise it, which would hardly have enhanced the credibility that a binding rule is supposed to promote. As it happened, the Bank simply abandoned M1 targeting and tried something else. The case for discretion over rules is simply that it provides a central bank with the flexibility to learn from mistakes this way instead of condemning the bank to repeat them.
2.3 Inflation targeting In February 1991, the Bank of Canada and the Department of Finance announced the new inflation targeting policy that the Bank has followed, with periodic revisions, ever since. It appears that this policy has been a great success, judging from the fact that inflation came down ahead of the targeted reductions and has remained within a narrow band of around two percent per year, with a few temporary deviations, ever since. I say that the policy has worked despite the fact, pointed out by Ball and Sheridan (2005), that inflation targeting countries have not reduced inflation by any more than non-targeting countries since the early 1990s, once one controls for the country’s average rate of inflation prior to the inflation-targeting era. My reason goes as follows.
Peter Howitt
53
My own interpretation of the evidence presented by Ball and Sheridan is that when central banks around the world were learning to rein in inflation, those that had the most difficult time doing so were those that had the least independent central banks, for the straightforward reasons discussed in section 2.2 above. These countries ended up adopting inflation targets as a way of reducing the inflationary political pressure on the central bank. According to this view, what was critical was not so much that the central bank signed on to inflation targets but that its political masters signed on. In effect, this made the central banks more independent de facto, by increasing their freedom to pursue their objective of low inflation. Thus countries that adopted inflation targets started with higher inflation than non-targeters, because of their low central-bank independence, and ended with inflation rates that looked much like the non-targeters, because of their enhanced de facto independence. This means that inflation fell by more on average among inflation targeters than non-targeters, not because they started with higher inflation rates and were reverting to the mean but because they were starting with less independence and would not have been able to revert to the mean without the increased independence offered to them by inflation targets. According to this view, those central banks that adopted inflation targets should have been the ones that had the least degree of independence de jure in the pre-inflation-targeting era. This seems reasonable given the countries that adopted inflation targets, such as New Zealand, Canada, the UK, Sweden and Australia. To confirm this impression more formally, I took the 20 countries in the Ball–Sheridan sample, for each one calculated the median of the four measures of central bank independence surveyed by Eijffinger and de Haan (1996), each measure having been transformed linearly so as to vary from 0 to 1. The coefficient of correlation between this median measure of central bank independence and a dichotomous indicator of whether the country adopted inflation targets turns out to be –0.46. My interpretation of the Ball–Sheridan evidence is also corroborated by the fact that the coefficient of correlation between inflation and this measure of independence fell (in absolute value) from 0.75 in the pre-targeting era to 0.37 in the post-targeting era. 2.3.1 Inflation targeting as a successful adaptation The adoption of inflation targets by the Bank of Canada can be seen as a successful adaptation to a particular economic and political environment. Of course the general idea of inflation or price-level
54
Inflation Targeting in Canada
targeting did not originate with the Bank, having been espoused by Simons (1936) among others, and having been put into practice by New Zealand shortly before the Bank’s adoption and by Sweden long before the Bank’s adoption. Even the specific idea of implementing the policy in Canada in 1991 originated not in the Bank but in the Department of Finance (Crow, 2002). But the Bank was clearly an active partner in adopting the new policy. As is common with historically embedded cognition, the Bank was economizing on its own internal resources by opportunistically learning from others and following others’ suggestions. The adaptation was successful largely because it dealt with all three of the main difficulties facing the central bank. First, the Bank managed to gain extra insulation from political pressure as a result of the move, as argued above, because its political master, the Department of Finance, also signed on. Once these targets had been sanctioned by the minister, there was little scope for him to pressure the Bank into more inflationary policies. Thus even when the Canadian government decided in 1993 that it was politically expedient not to renew the contract of the governor that had become the public face of inflation targeting, it chose to replace him with his former senior deputy, who remained committed to those targets, and under whose regime the measured rate of CPI inflation in Canada remained for several years at or below 1.5% per year.13 As for dealing with the control problem, as many have observed, inflation targeting sharpens the focus of central bank announcements and creates increased transparency. It thereby makes it easier for the Bank to communicate its reasoning to the public, thus allowing people to better anticipate future movements in the overnight rate. The increased transparency also allows people to see that the Bank is really doing as announced, and thus enhances the credibility needed for the crucial announcement effect to work. The announcement effect itself is of course a good example of how an adaptive agency opportunistically manipulates aspects of its external environment, in this case expectations of external private agents, in such a way as to leverage its own limited and fragmented analytical capabilities. Inflation targeting has helped the Bank to deal with its third major problem by allowing it to continue to learn from experience and experimentation without sacrificing its own credibility. Recall what happened when the Bank attempted to follow a policy of targeting the growth rate of the money supply in the second half of the 1970s. After having committed itself to targeted reductions in M1 growth,
Peter Howitt
55
the Bank soon learned that the demand for M1 was undergoing negative shocks which were nullifying the effects of these reductions on inflation. This put the Bank in the awkward situation of having to choose between allowing inflation to persist, thereby defeating the ultimate objective of the policy, or violating its commitment to the targets, thereby undermining its own credibility and reducing its ability to talk inflation down. Nor was the Bank of Canada the only central bank in the world that found itself in such a dilemma. As Goodhart (1984) pointed out, every country in the world that undertook a monetary targeting policy found that the demand for whatever M that they were targeting somehow suddenly started to decrease. If these central banks had been committed to inflation targets rather than money-growth targets, the lessons that they learned in the 1970s when they tried reducing monetary growth could easily have been put to use without jeopardizing the long-run goal of inflation reduction. They could have started right away aiming for much lower monetary growth, or switched to controlling some other monetary aggregate, or they could even have abandoned the discredited policy in favor of some other approach to inflation control that placed less reliance on monetary aggregates. In the end that’s what many of them did, but only after a lengthy and costly delay caused by the understandable wish to maintain their reputation for constancy. Moreover, the open and transparent framework of inflation targeting helps the Bank of Canada not just with the control problem, as explained above, but also with the problem of understanding the economy.14 This is because transparency allows the Bank to explain more clearly than ever what is going on when it changes tactics. It can explain openly that what it is doing is just a tactical policy that in no way diminishes its commitment to the publicly announced policy of inflation control, as for example when it abandoned the use of its monetary conditions index or when it moved to a schedule of fixed action dates for setting interest rates. In other words, the Bank is now free to benefit from new information that teaches it something about how the economy works and indicates that a change of tactics is in order, without having to sacrifice its credibility, rather than having to hope that no such lessons will be forthcoming. 2.3.2 The game-theoretic view of inflation targeting In contrast to the account I have just presented, many contemporary accounts of inflation targeting are presented as extensions of the following
56
Inflation Targeting in Canada
game-theoretic model, according to which inflation targeting is a rule, which, being imposed on the central bank, delivers it from the temptation to deliver surprise inflation. According to this account, the central bank that is not constrained by such a rule would use its discretion to yield to this temptation unless expected inflation were so high as to make the cost of the surprise even higher than the gain in terms of reduced unemployment. A simplified formal model of this game-theoretic view goes as follows.15 At each date t, we have pt as the rate of price inflation, pte as the expected rate of inflation, ut as the rate of unemployment, v as the natural rate of unemployment, a as the relative importance of inflation and b as the inverse slope of Phillips curve. The parameters v, a and b are all positive. The central bank sets the rate of inflation in each period knowing that the rate of unemployment will be determined by the expectationsaugmented Phillips curve: ⎛ 1⎞ pt ⫽ ⫺ ⎜ ⎟ ( ut ⫺ v ) ⫹ pte . ⎝ b⎠
(1)
Its objective is to minimize the social loss from unemployment and inflation: Lt ⫽ ( ut )2 ⫹ a(pt )2 .
(2)
According to this consensus model, the key choice of monetary policy consists in choosing the combination of inflation and unemployment that minimizes the social loss (2) subject to the trade-off (1). It is standard to assume that rational expectations prevail, which in this non-stochastic world implies that the expected rate of inflation entering the Phillips curve trade-off will be determined by perfect foresight. pte ⫽ pt .
(3)
According to this theory, the difference between having an inflation target imposed on the central bank and leaving it free to use its discretion can be modeled as the difference between whether the central bank chooses the actual rate of inflation before or after people form their expectations. Under discretion, people have to form their expectations
Peter Howitt
57
first, and the bank retains the option of varying inflation afterwards. In that case, the central bank will minimize (2) subject to (1), taking the value of pte as predetermined. Thus it will set pt ⫽
bv ⫹ b 2pte . a ⫹ b2
(4)
From (3) and (4), the equilibrium values of actual and expected inflation will be pt ⫽ pte ⫽
bv ⬎ 0, a
under discretion.
(5)
Suppose, however, that the central bank is constrained by an inflation target. This target determines the rate of inflation pt before people form expectations. Whoever sets the target can anticipate that expected inflation will be governed by (3). So according to the Phillips trade-off (1) the rate of unemployment will always equal the natural rate v no matter what target is chosen. In that case, the only variable in the loss function (2) affected by the target will be the rate of inflation, which should therefore be set to its optimal value of zero. Hence pt ⫽ pte ⫽ 0,
under the optimal inflation target.
(6)
It follows immediately that having the best inflation target is better than discretion, in the sense that it results in a lower value of the loss function (2). The assumption of rational expectations together with the expectations-augmented Phillips curve (1) imply that in both cases the rate of unemployment will be the same, namely the natural rate v. So allowing the central bank to use discretion just results in a higher rate of inflation, with no pay-off in terms of lower unemployment. Many economists have been convinced by this game-theoretic argument, because it does not rely either on mistaken perceptions by the public16 or on faulty objectives or incompetence on the part of the central bank. The presumption is that the objective being pursued by the central bank is a worthy one, and that everyone is acting in a way that is individually optimal. In effect, the argument portrays discretion as a misplaced softness, like the weakness of a judge that always feels sympathy for a criminal and treats bygones as bygones. An inflation target is needed to stiffen the bank’s backbone. Failing that, we would need to appoint a central banker who had no concern for unemployment, who would act as if the parameter a in the objective function were
58
Inflation Targeting in Canada
indefinitely large, and who would therefore approximate the first-best inflation target even though not constrained to do so. Thus inflation-targeting has been embraced by many monetary economists as a way, although perhaps not the only way, to avoid the useless inflationary bias of benevolent but uncommitted central banks who cannot be counted on to resist the temptation to help society by giving it a bit of surprise inflation. In effect, this is exactly the opposite of the adaptive view explained above, for while the game-theoretic analysis would rationalize inflation targeting as a way to use rules rather than discretion, the adaptive view sees them as an example of discretion rather than rules. 2.3.2.1 The New Keynesian view I digress briefly to discuss (a simplified version of) the variant of this argument presented by Svensson and Woodford (2005), who replace the “neoclassical” Phillips curve (1) by a “New Keynesian” version: ⎛ 1⎞ pt ⫽ ⫺ ⎜ ⎟ (ut ⫺ v ) ⫹ bpte⫹1 ⫹ et , ⎝ b⎠
0 ⬍ b ⬍ 1,
(7)
in which what matters on the right-hand side is not last period’s expectation of this period’s inflation but this period’s expectation of next period’s inflation, with a coefficient less than unity, and where explicit account is taken of the random price-shocks, et. Svensson and Woodford also consider an explicitly dynamic loss function equal to the expected discounted value of (2) for all periods. Using this variant, they argue that the bank needs to be committed to inflation targets not just to avoid the inflation bias of discretion but also to avoid what they call the “stabilization bias” of discretion. That is, following a positive price shock that disturbed the Phillips curve, an optimal monetary policy under commitment in the New Keynesian model would require the central bank to accommodate inflation somewhat during the period of the shock but then to bring inflation below target in future periods, even if the price shock were purely transitory. This future tightening of monetary policy would increase future expected losses, but the expectation of lower inflation would allow a more favorable inflation–unemployment trade-off during the period of the shock, and the overall effect of the tightening would be to reduce the discounted sum of expected losses. The problem, however, is that if the central bank is not committed to this future tightening it will
Peter Howitt
59
not undertake it, since the benefits in the form of improved inflation expectations will be bygone. More generally, Svensson and Woodford argue that the optimal inflation target should be history-dependent in a way that would never be implemented by a central bank minimizing the “true” social loss function under discretion. Hence, the central bank needs to be constrained by a rule that implements the optimal inflation target. As an explanation for why inflation targeting works, this New Keynesian argument adds little to the preceding game-theoretic argument. The optimal rules derived by Svensson and Woodford all converge in the long run to a fixed target of zero inflation, for exactly the same reasons as in the Kydland–Prescott analysis. Svensson and Woodford supplement this fixed long-run target with a state-contingent shortterm inflation target. But inflation-targeting central banks as a rule do not publicly commit themselves to state-contingent targets. Nor is it realistic to suppose that they deliberately aim at tighter monetary policy long after a positive price shock has finished having a direct effect on the economy. So although the management of expectations in the face of price shocks is certainly an important part of inflation targeting,17 the kind of policy that this New Keynesian argument rationalizes is not the kind that inflation targeters have been following.18 2.3.3.2 What is lacking: commitment or understanding? The game-theoretic view that I have just spelled out bears little resemblance to the adaptive view that I outlined in the preceding section. Rather than having to learn about how the economy works, everyone is portrayed as having learned all there is to learn. The struggle to retain control over something that might provide a nominal anchor to the system is assumed away by the taking of inflation itself as an instrument of the central bank. Instead of learning and adaptation, it presents commitment and the resistance to inflationary temptation as the central aspects of monetary policy. The rest of this section presents the case for seeing monetary policy as more of a matter of learning than as a matter of commitment.19 Consider the theoretical basis of the argument from the previous section. First, it is predicated on the basic idea of the expectationsaugmented Phillips curve, and the related idea of the ineffectiveness of monetary policy. The reason why discretion yields no pay-off is that it can be anticipated, whereas the only sort of policy that can affect the economy is that which produces unexpected changes in inflation. As a general proposition, policy ineffectiveness has long
60
Inflation Targeting in Canada
since been discredited by numerous empirical studies, as has been the augmented Phillips curve and its prediction of a unique natural rate of unemployment.20 Moreover, there is little reason to think that private expectations enter into the inflation process in the way portrayed by this theory. For although there is some empirical support for the idea that inflation depends upon some measure of economy-wide excess demand for goods and services and on lagged inflation, the lagged inflation terms do not appear to serve so much as proxies for forward-looking expectations as for timing delays in the setting of the millions of individual prices that comprise the price index.21 One seller’s price will change with a brief lag when there is enough change in the prices of other factor inputs and/or competing goods have changed, but as we pointed out above, a succession of small lags can amount to a very sluggish process in the aggregate. The place where expectations of monetary policy seem to enter most critically is in financial markets rather than in the markets for goods and labor services. And when they respond to inflation, they do so without delay. If investors begin to suspect that the Bank of Canada is not committed to a disciplined monetary policy, the first thing that happens is that interest rates begin to rise and the exchange value of the Canadian dollar begins to fall. The fact that expectations of monetary policy can have their most significant effects on the economy instantaneously implies that the game-theoretic model of discretionary policy outlined above has little basis in reality. According to that model, expectations of monetary policy matter mainly in the markets for goods and services, in which wages and prices are often fixed in advance for months or even years. If this were true, then it would make sense to suppose as is done in the game-theoretic model of discretion that the central bank can rationally take expectations as predetermined when choosing its policy action. The “temptation” to raise the level of economic activity with some surprise inflation might exist if society were indeed locked into expectations. In reality, however, the temptation just doesn’t arise, as practitioners of central banking have long maintained.22 Central bankers are keenly aware that although there are long and variable lags between monetary stimulus and any resulting rise in the level of economic activity, there are no lags at all between such a stimulus and the currency depreciation and capital flight that will occur if the stimulus is taken by investors as a signal of future weakness in the currency. Because of this, there is no reason
Peter Howitt
61
for believing that discretionary central banks have the inflationary bias that the game-theoretic view attributes to them. Another weakness of the game-theoretic model is its assumption that central bank actions are motivated by the desire to minimize a cost function with inflation and unemployment as objectives.23 Aside from some serious doubts as to the form of that cost function and whether or not it could have anything to do with conventionally defined economic welfare, it is doubtful that this is a reasonable description of the motivation of any governor of the Bank of Canada or of any other central banker. Instead, I believe the evidence favors the account that has been given over the years by such astute observers as Thornton (1802), Bagehot (1873) and Milton Friedman,24 namely that responsible people entrusted with such important and delicate jobs as the management of a country’s central bank are typically motivated by the desire to be seen as having done a good job, to have acquitted themselves well. They pursue this objective by doing everything possible to avoid major inflations, financial panics and runs on the currency, while carrying out the day-to-day job of making available the base money needed for the financial system to function.
2.4 The evolution of monetary policy If we look beyond inflation targeting, I believe that the adaptive view of monetary policy in general provides a more coherent account than the game-theoretic view of the way monetary policy has evolved in North America and elsewhere. One of the strongest arguments in favor of taking the trial-and-error point of view is that one does not have to be a committed monetarist to agree with Friedman and Schwarz that the history of major depressions and inflations in the US and other countries is to a large extent a history of errors made by those in charge of monetary policy. Not a history of insufficient resolve to resist the temptation to do good, as would be portrayed by now conventional rational expectations theory, or a history of too low a weight placed on inflation in the central bank’s loss function, but a history of insufficient understanding of how best to carry out the task of central banking. In the 1950s and the 1960s, the Federal Reserve Board operated largely on the principle of what its Chairman William McChesney Martin called “leaning against the wind”; that is, of trying to counteract short-term movements in nominal interest rates. As Friedman argued, this policy was almost guaranteed to amplify the rise in unemployment when aggregate demand fell, and to add momentum to inflation in the
62
Inflation Targeting in Canada
face of either supply or demand shocks that started an incipient inflation spiral. In making this criticism, Friedman was reopening a debate that goes back to the “real bills doctrine” espoused by Adam Smith and others, a debate that is still unresolved, and which involves the fundamental conflict referred to above between the two main tasks of central banks. Friedman’s insistence that central banks pay strict attention to conventional monetary aggregates was treated by conventional Keynesians in the early 1970s as crankish nonsense. Monetary policy was largely influenced by those Keynesian ideas and by the related Radcliffe view according to which monetary policy was largely ineffective because conventional monetary aggregates were only one component of the multidimensional notion of “liquidity” that really influenced the level of aggregate expenditures in the economy, to the extent that these were influenced at all by financial considerations. It would be hard to find a monetary economist today that did not agree that Friedman was right, and did not agree that high inflation is certain to occur in a situation in which all known monetary aggregates are growing persistently at rates between 15 and 25 per cent per year. Yet in the early 1970s in Canada such was the situation, and the then governor of the Bank of Canada, Louis Rasminsky apparently saw no reason to change course.25 Rasminsky was not a lover of inflation. Nor was he, by all accounts, a man of weak resolve easily given to short-run temptations. He just made an honest mistake. He saw the need to preserve “orderly markets”, which led him, like Martin and other central bankers at the time, to pursue a strategy of dampening interest rate movement. He didn’t see the conflict between this strategy and the goal of avoiding a disastrous inflation. The next governor of the Bank of Canada, Gerald Bouey, apparently saw things differently. He set out on the course of “Monetary Gradualism”, that was described in section 2.2 above, which foundered when the targeted M1 aggregate began to exhibit an unstable relationship with interest rates and nominal income. What ruined monetary gradualism was not lack of commitment but lack of understanding. And so it has been in case after case. The disappointments that we have seen in monetary policy have largely been the result of error, not deception. Moreover, the history of central banking has been one of trial as well as error. When mistakes have been made, central banks have learned from them, as have economists. We have learned that interest
Peter Howitt
63
rates must often be allowed a wide latitude from time to time, and that when all monetary aggregates start growing at high rates, inflation is sure to follow if nothing is done to reverse course. We have learned that a policy of publicly focusing on a single aggregate is not enough, because once people get used to the new regime, the chosen aggregate will cease to be indicative of the overall level of aggregate demand. Another aspect of monetary policy that can be explained by seeing it as a question of trial and error rather than strategic manipulation is the apparent herd instinct of central bankers. As the above sketch of Canadian monetary history has indicated, the history of successive attempts to formulate a monetary strategy in Canada parallels in broad outline that of most other advanced countries.26 And the resulting movements in inflation have been remarkably similar in different countries despite many differences in institutional and political arrangements that give different central banks quite different incentives and different opportunities to make credible commitments. This herd behavior on the part of central bankers has a straightforward interpretation if we view monetary policy as a history of trial and error. For as Keynes once argued so persuasively, people faced with unquantifiable uncertainty tend to rely on custom and convention. When there is no rational basis for assessing the probable consequences of alternative courses of action, the prudent course is to do what is commonly accepted as the sensible thing. And surely there is nothing about which we have more uncertainty than the precise way in which central bank actions today will affect such variables as the level of economic activity and the price level, two years from now. Of course, many people in uncertain situations will strike out on uncharted courses that defy conventional wisdom. Those that succeed are recognized with hindsight as creative geniuses. But the financial world does not generally accord credibility to people with such proclivities. On the contrary, it tends to favor those who don’t deviate much from conventional opinion. Those whose expectations are just a little ahead of the crowd are handsomely rewarded in their speculations without having ever to form an independent thought about the fundamental determinants of profitability. And those who make mistakes managing others’ fortunes are least likely to be punished when they have been following “best practice.” No one is more charged with such social responsibility than a central banker. From this point of view, what is surprising is that they are not even more alike, that some central banks, such as that of Canada in 1975
64
Inflation Targeting in Canada
and of New Zealand in 1990 have taken the lead and initiated new untried policies. There is another feature of this herd behavior that is easy to understand from the point of view of this essay, and that is its rough coincidence with developments in monetary theory. In the 1960s, central bankers were guided largely by Keynesian ideas. In the 1970s and 1980s, they seemed to take their inspiration from the monetarism that had then become so popular in academic writings on macroeconomics. In the 1990s, the proliferation of inflation targeting can be seen as a direct extension of the Kydland–Prescott game-theoretic analysis, for one solution to the conundrum of time inconsistency, as indicated above, is to delegate monetary policy to a central bank with a “contract” mandating the pursuit of a specific range of inflation. In all these cases, the globalization of monetary theory has implied that what can best be rationalized in terms of conventional wisdom in one country is the same as in another, at least in so far as conventional wisdom accords with academic fashion. As another example of something more easily explained by the adaptive view of monetary policy than by the game-theoretic view, consider the fact that economists have been proffering policy advice to central bankers over the years. The rational expectations assumed in the game-theoretic approach would make such advice redundant. For as several observers have noted, the assumption of rational expectations implies that monetary theories cannot at the same time be new and correct, since correct theories are already assumed to be in use. The game-theoretic view gives no role for advocacy other than perhaps providing more precise estimates of the slope of the expectations-augmented Phillips curve. If everyone is behaving optimally then what are the economists maximizing?27 Under the adaptive view, however, policy makers should continually be seeking and even paying for advice as to how the economy works, if not for improving their performance, then at least for acquiring a stamp of academic approval.
2.5 Who is ahead of whom? When the Bank of Canada and the Department of Finance announced the Bank’s new policy of inflation targeting in 1991, Canada was introducing a new goods and services tax. The new tax was clearly going to create a problem for the Bank by causing an upward blip in the price level. Even if the Bank could prevent this blip from turning
Peter Howitt
65
into an inertial inflationary spiral, the immediate rise in inflation that would accompany the blip threatened to undermine the Bank’s credibility. Inflation targets were seen at first as a means for dealing with this problem. More specifically, the Bank estimated the firstround effect of the new tax on the price level, under the assumption that the path of wages would not be affected, and the policy was designed to limit the price blip to that estimated amount. Inflation targets were announced after the blip had had its first-round effect, rising to about six per cent, and the announcement promised to halt this rise and to bring inflation gradually down to within a band of one to three percent over the following three years. When I first heard of this announcement, and for sometimes afterwards, I was very skeptical. Along with many other academic economists, I thought it was foolish for the Bank to announce that it was going to control something like inflation, which it can only affect through a long and variable lag, with such a high degree of precision. To me the idea reeked of fine-tuning, and I thought the Bank was setting itself up for a fall. I should add that although David Laidler also had qualms, he was much more supportive than I was (see Laidler, 1991). As usual, he showed better judgment than I did, because in the end, the Bank managed to contain the price blip just as planned. The price level stopped rising and then inflation quickly came down to within the target range, where it has been almost continuously ever since. This is just one example of how central banks seem to have learned a lot during the past half century without the support of academic economists. In the US, for example, economists often poked fun at the Greenspan method of sifting through all sorts of data with no apparently coherent guidance from economic theory. But there is no disputing his success in maintaining a low course of inflation, even in circumstances (with unemployment dipping below 4%) where established theory was predicting that inflation was going to accelerate. When the most successful central bankers appear to pay little attention to economic theory, and the economic theory which seems to rationalize the success of inflation targeting is riddled with empirical contradictions, it is reasonable to infer that something is wrong with theory, not practice. It suggests that in the social process of adaptation in which both practitioners and students of monetary policy play a role, the students now have more to learn from the practitioners than the other way round. Now such a situation would not appear
66
Inflation Targeting in Canada
anomalous to anyone familiar with the historical relationship between science and technology, where fundamental scientific breakthroughs have been the result rather than the cause of successful innovations made by practical people solving mundane problems, often with little or no understanding of why the innovation works.28 Inflation targeting seems to work, and we are still waiting for a full-blown monetary theory that can provide an empirically successful explanation of this practical experience. Monetary theory has fallen behind practical experience partly because instead of taking money seriously as a critical part of an economy’s coordination mechanism, it has been mesmerized by a priori ideas like rational expectations. This hypothesis does more to hide than to illuminate the process of learning and adaptation that lies at the heart of monetary policy and of much else in economic life. For it asserts that there is only one correct theory of how the economy works, and central banks know it as well as anyone else.29 The fact that this assumption has appealed equally to new classicals and new Keynesians, whose views of how the economy works are in fundamental contradiction with one another, has not apparently reduced its popularity in either camp. But it has put modern monetary economics out of touch with the reality of monetary economies. As central bankers continue to improve their methods, they will find less and less need to continue to pay lip service to academically fashionable economic theory that cannot account for that success. But there are plenty of problems remaining with inflation targeting. One is the vexing question of how to deal with asset-price bubbles, about which much has been written. Another is the problem of the loss of early warning signs of inflation in an environment where inflation expectations have become entrenched. Specifically, with inflation having become 2% plus white noise, all leading indicators of inflation have lost their predictive power, including measures of inflation expectations, monetary aggregates, core inflation and the output gap. The best predictor is just 2%. Moreover, as inflation becomes more entrenched, private forecasters have less and less incentive to spend resources trying to predict something so predictable, and as a result, asset prices and expectational surveys no longer reflect much information coming from peoples’ forecasting activities. But as we have emphasized above, given the inertial nature of inflation, the policy of inflation targeting depends very much on early warning signs. Latent inflationary pressure can build up for a long time before its symptoms become apparent, and once they do become apparent they are very costly to
Peter Howitt
67
reverse. So the more our leading indicators continue to deteriorate, the more danger we face of waking up one morning and finding that we have been unknowingly fostering a vigorous inflation that can only be eliminated by a prolonged economic slowdown. In that sense, conducting monetary policy is becoming more and more like trying to fly an airplane blindfolded without crashing, and central bankers are facing the very difficult problem of finding a way to operate without feedback, a problem that inflation targeting has done more to create than to solve. So, adaptive central banks will still be looking to academia for answers for some time, even as they continue to look for answers on their own. Moreover, there is a growing literature on learning in macroeconomics30 that takes the adaptive view of economic life, some of it devoted explicitly to questions of monetary policy. So economic theory still has a chance to catch up with the practical lessons being learned by central banks. If it does catch up, this will be because of people that follow the path that David Laidler has always been on, constructing and judging theories according to their explanatory power rather than according to conventional a priori beliefs, and taking seriously the role of money as a central institution of economic life.
Notes Parts of the paper are drawn from my unpublished essay entitled ‘Learning About Monetary Theory and Policy’, which benefited from many conversations on the subject with David Laidler and also with Joel Fried. John Crow, Chuck Freedman, Nicholas Rowe, T. K. Rymes and seminar participants at Carleton University and the Laidler Festschrift provided valuable comments. 1. For example, Laidler (1974, 1984, 1988, 1993). 2. Clark’s attribution to Woody Allen seems to be incorrect; according to all the references found in a Google search, Allen is reported as having said something like ‘80 percent of success in life is just showing up’. But perhaps the cinematic reference is nonetheless apt in the present case, because ‘Being There’ is the title of a movie in which a simpleton played by Peter Sellers was taken by everyone as being a genius. 3. This is not to say that the central bank must use the base as its instrument. On the contrary, operating directly on the base would conflict with the obligation to maintain a well-functioning money market, especially when, as is now the case in Canada, the base consists entirely in notes. The point is just that whatever instrument the central bank uses must have the effect of limiting the growth rate of the base.
68
Inflation Targeting in Canada
4. See Bade and Parkin (1987). 5. There are other levers by which a government might still exert such pressure, as John Crow (2002, Chapter 2) observes, because of its power to veto a governor’s appointment or reappointment and to disapprove salary increases for the governor and deputy governors, but presumably the elimination of one possible lever does help to insulate the Bank to some extent from the pressure. Crow also rejects Pierre Fortin’s (1996) claim that ‘No sane minister of finance would ever dare start such a process’, on the grounds that there might indeed by circumstances under which a sane minister could produce a well thought-out directive. But it seems unlikely that these circumstances would include cases in which the minister was trying to engineer an inflation without having to pay a price in the short run. 6. For a simple description of how the Bank of Canada operates under the LVTS, see Howard (1998). 7. By “availability of credit” I mean the non-interest terms on which financial intermediaries are prepared to lend, such as credit limits, collateral requirements, allowable debt–service ratios, minimum monthly payments, and so on, many of which become more restrictive when monetary policy is tightened. 8. This excess supply is purely momentary in the textbook IS-LM analysis, according to which the rate of interest adjusts rapidly to equate the supply and demand for money. By contrast, the “active-money” approach that David Laidler has been proposing (see Laidler, 1999b, for a recent exposition) implies that the excess supply would persist for much longer, since the rate of interest adjusts only far enough to equate the supply and demand for loanable funds. The adjustment process implied by Laidler’s analysis is considerably more complex than that of the textbooks, and as Laidler has acknowledged, there is much about it that we do not yet understand. 9. See Blinder et al. (1998) for example. 10. On the mechanics of such processes see Blanchard (1986). 11. See Macklem (2002) or Sims (2002). 12. See Freedman (1983). 13. See Howitt (1997). Crow (2002) argues that inflation targeting actually resulted in a decrease of the Bank of Canada’s independence because it brought the Finance Department into the process of formulating monetary policy more than ever in the past. Although this does mean a shift of policymaking power from the Bank to the Ministry, I would argue nonetheless that it made it even more difficult than before for the minister of finance to pressure the bank into financing deficits with highly inflationary policies. 14. Transparency also helps, of course, to make the Bank more accountable to the general public. Although this involves a sacrifice of the Bank’s independence from political factors, it clearly does so in a way that does not imply increased inflationary pressure. As David Laidler (2005) points out, this kind of dependency is needed to make the Bank of Canada fit into a modern democratic system in a way that Henry Simons might approve.
Peter Howitt
69
15. The seminal article from which this game-theoretic approach originates is Kydland and Prescott (1977). Other central contributions were made by Barro and Gordon (1983), Canzoneri (1985), Rogoff (1985), Cukierman and Meltzer (1986) and Walsh (1995). 16. Except perhaps the perception that the monetary authority that they had created would serve them well. 17. See, especially, section 2.5. 18. The empirical validity of the New Keynesian argument can also be challenged on the grounds that the Phillips curve (7) on which it is based, implies, counterfactually, that inflation should on average rise during recessions. See Fuhrer and Moore (1995). 19. A similar conclusion has been reached by Romer and Romer (1997). For another interpretation of US monetary policy that emphasizes learning, see Sargent (1999). Simultaneous learning by a central bank and private agents is analyzed by Honkapohja and Mitra (2002). 20. On policy effectiveness, see Boschen and Grossman (1982). On the Phillips curve, see the symposium in the Journal of Economic Perspectives in the winter issue of the year 1997. 21. For evidence of this using Canadian wage contract data, see Riddell and Smith (1982). For US evidence, see Fuhrer (1997). 22. See Blinder (1998). 23. As Chuck Freedman has emphasized to me, the way a central bank responds to price shocks, trying to allow some of the effects to come out in inflation and some in real output, makes it look as if the bank was trying to minimize a loss function involving these two variables. But these short-run responses are part of the detailed implementation of policy having little to do, in my opinion, with the first-order questions of keeping inflation anchored and avoiding financial collapse. 24. Friedman has been cited indirectly on this subject by King (1997). 25. See Courchene (1976). 26. I have argued this in greater detail in Howitt (1993). 27. On this point, see Rymes (1979). 28. See Rosenberg and Birdzell (1986). Laidler (1999a, esp. pp. 323 ff.) makes a case that textbook IS-LM analysis is another example of theory following practice. What others have called the Keynesian ‘revolution’ is in Laidler’s analysis a theoretical synthesis of various ideas and proposals that had sprung up in the decades preceding Keynes’s General Theory, many of them in response to urgent policy concerns. 29. On the same point, see Laidler (1999a, xi, xiii) 30. See for example, Evans and Honkapohja (2001).
References Bade, Robin, and Michael Parkin. “Central Bank Laws and Monetary Policy.” Unpublished, University of Western Ontario, June 1987. Ball, Laurence, and Niamh Sheridan. “Does Inflation Targeting Matter?” In The Inflation Targeting Debate, edited by Ben S. Bernanke and Michael Woodford, 249–76. Chicago: University of Chicago Press, 2005.
70
Inflation Targeting in Canada
Bagehot, Walter. Lombard Street: A Description of the Money Market. London: Henry S. King & Co., 1873. Barro, Robert J., and David Gordon. “A Positive Theory of Monetary Policy in a Natural Rate Model.” Journal of Political Economy 91 (August 1983): 589–610. Blanchard, Olivier J. “The Wage-Price Spiral.” Quarterly Journal of Economics 101 (1986): 543–65. Blinder, Alan S. Central Banking in Theory and Practice. Cambridge, MA: MIT Press, 1998. Blinder, Alan S., Elie R. D. Canetti, David E. Lebow and Jeremy B. Rudd. Asking About Prices: A New Approach to Understanding Price Stickiness. New York: Russell Sage Foundation, 1998. Boschen, John E., and Herschel I. Grossman. “Tests of Equilibrium Macroeconomics Using Contemporaneous Monetary Data.” Journal of Monetary Economics 10 (November 1982): 309–33. Canzoneri, Matthew B. “Monetary Policy Games and the Role of Private Information.” American Economic Review 75 (December 1985): 1056–70. Clark, Andy. Being There: Putting Brain, Body and World Together Again. Cambridge, MA: MIT Press, 1998. Courchene, Thomas. Money, Inflation, and the Bank of Canada. Montreal: C. D. Howe Institute, 1976. Crow, John W. Making Money: An Insider’s Perspective on Finance, Politics and Canada’s Central Bank. Etobicoke, ON: Wiley, 2002. Cukierman, Alex, and Allan H. Meltzer. “A Theory of Ambiguity, Credibility, and Inflation Under Discretion and Asymmetric Information.” Econometrica 54 (September 1986): 1099–1128. Eijffinger, Sylvester C. W., and Jakob de Haan. “The Political Economy of CentralBank Independence.” Princeton Economics International Finance Section, Special Papers in International Economics, no. 19, May 1996. Evans, George, and Seppo Honkapohja. Learning and Expectations in Macroeconomics. Princeton, NJ: Princeton University Press, 2001. Fortin, Pierre. “The Great Canadian Slump.” Canadian Journal of Economics 29 (November 1996): 761–87. Freedman, Charles. “Financial Innovation in Canada: Causes and Consequences.” American Economic Review Proceedings 73 (May 1983): 101–6. Friedman, Milton. “Remarks at the Milton Friedman Seminar.” Sponsored and published by C. J. Hodgson, Richardson Inc., Montreal, October 2, 1975. Fuhrer, Jeffrey C. “The (Un)Importance of Forward-Looking Behavior in Price Specifications.” Journal of Money, Credit and Banking 29 (August 1997): 338–50. —— and George Moore. “Inflation Persistence.” Quarterly Journal of Economics 110 (February 1995): 127–93. Goodhart, Charles A. E. Monetary Theory and Practice: The U.K. Experience. London: Macmillan, 1984. Honkapohja, Seppo, and Kaushik Mitra. “Performance of Monetary Policy with Internal Central Bank Forecasting.” European Central Bank Working Paper No. 127, 2002. Howard, Donna. “A Primer on the Implementation of Monetary Policy in the LVTS Environment.” Bank of Canada Review (August 1998): 57–66.
Peter Howitt
71
Howitt, Peter. “Canada.” In Monetary Policy in Developed Economies (eds) M. Fratianni and D. Salvatore, 459–508. Handbook of Comparative Economic Policies, Vol. 3. Westport, Connecticut: Greenwood Press, 1993. —— “Low Inflation and the Canadian Economy.” In Where We Go from Here: Inflation Targets in Canada’s Monetary Policy Regime (eds) David Laidler, 27–67. Toronto: C. D. Howe Institute, 1997. King, Mervyn. “Changes in UK Monetary Policy: Rules and Discretion in Practice.” Journal of Monetary Economics 39 (June 1997): 81–97. Kydland, Finn E., and Edward C. Prescott. “Rules Rather than Discretion: The Inconsistency of Optimal Plans.” Journal of Political Economy 85 (June 1977): 473–91. Laidler, David. “Information, Money and the Macroeconomics of Inflation.” Swedish Journal of Economics 76 (March 1974): 26–41. —— “The ‘Buffer Stock’ Notion in Monetary Economics.” Economic Journal 94 (Supplement 1984): 17–34. —— “Taking Money Seriously.” Canadian Journal of Economics 21 (November 1988): 687–713. —— “Qualms about Inflation Targets.” In The February 1991 Federal Budget (eds) Martin F. J. Prachowny and Douglas D. Purvis, 29–37. Kingston, ON: John Deutsch Institute for the Study of Economic Policy, Queen’s University, 1991. —— “Price Stability and the Monetary Order.” In Price Stabilization in the 1990s (eds) K. Shigehara, 331–56. London: Macmillan, 1993. —— Fabricating the Keynesian Revolution: Studies of the Inter-War Literature on Money, the Cycle, and Unemployment. New York: Cambridge University Press, 1999a. —— “The Quantity of Money and Monetary Policy.” Bank of Canada Working Paper 99-5, April 1999b. —— “Review of ‘Making Money: An Insider’s Perspective of Finance, Politics, and Canada’s Central Bank’ by John Crow.” Canadian Journal of Economics 36 (August 2005): 758–64. Macklem, Tiff. “Information and Analysis for Monetary Policy: Coming to a Decision.” Bank of Canada Review (Summer 2002): 11–18. Riddell, Craig, and Philip Smith. “Expected Inflation and Wage Changes in Canada.” Canadian Journal of Economics 15 (August 1982): 377–94. Rogoff, Kenneth. “The Optimal Degree of Commitment to an Intermediate Monetary Target.” Quarterly Journal of Economics 100 (November 1985): 1169–89. Romer, Christina D., and David H. Romer. “Institutions for Monetary Stability.” In Reducing Inflation: Motivation and Strategy (eds) C. Romer and D. Romer. Chicago: University of Chicago Press, 1997. Rosenberg, Nathan, and L. E. Jr Birdzell. How the West Grew Rich: The Economic Transformation of the Industrial World. Basic Books, 1986. Rymes, Thomas K. “Money, Efficiency, and Knowledge.” Canadian Journal of Economics 12 (November 1979): 575–89. Sargent, Thomas J. The Conquest of American Inflation. Princeton: Princeton University Press, 1999. Simons, Henry. “Rules Versus Authorities in Monetary Policy.” Journal of Political Economy 44 (February 1936): 1–30.
72
Inflation Targeting in Canada
Sims, Christopher. “The Role of Models and Probabilities in the Monetary Policy Process.” Brooking Papers on Economic Activity 2 (2002): 1–40. Svensson, Lars E. O., and Michael Woodford. “Implementing Optimal Policy through Inflation Forecast Targeting.” In The Inflation-Targeting Debate (eds) Ben S. Bernanke and Michael Woodford, 19–83. Chicago: University of Chicago Press, 2005. Thornton, Henry. An Enquiry into the Nature and Effects of the Paper Credit of Great Britain. Fairfield, NJ: Augustus M. Kelley, 1962 (1802). Walsh, Carl E. “Optimal Contracts for Central Bankers.” American Economic Review 85 (March 1995): 150–67.
Discussion John Crow
I’m enormously pleased to be here to celebrate David’s contributions. With my background, I tend of course to evaluate, and value, what he’s done (will be doing?) from the policy side. And so it was gratifying when I was involved to see someone with his knowledge, experience, sanity (together with, from time to time, a dose of passion) apply himself to regular commentary on what monetary policy was up to and why. He’s kept the Bank on its toes, especially of course in figuring out whether it begged to differ. But also, to pick up a theme of Peter’s paper, I can suppose that in observing the Bank of Canada do what it did (and it did quite a lot) through lively, and sometimes distinctly challenging, monetary policy times, he will have learned a thing or two. One thing that it seems to me both David and Peter must have experienced as events unfolded, is a deepening appreciation of the complicated and varying role played by the government in monetary policy affairs. This is something that the government itself (notoriously perhaps from a transparency viewpoint, but true to form from a political loss-function angle) will generally be diligent to hide. It is also something that the Bank of Canada, given its situation (at the very least advisory) within the governmental set-up, is not in a position to explain or discuss if the government doesn’t want to – and those folk hardly ever do want to.
Which masters need educating? I applaud the paper’s critique of the game-theoretic view of monetary policy. In particular, the practical policy answer that has in recent years been claimed to stem from this (see, as a notable exponent, 73
74
Inflation Targeting in Canada
Stanley Fischer, interestingly enough now Governor of the Bank of Israel) – namely, monetary policy goals should be set by the government, with the central bank limited to carrying out the necessary operations – represents a misreading of what really was at issue for better monetary policy. That is to say, if anyone needed to be tied down to a consistently anti-inflationary monetary policy, it was the political authority, and not a central bank continually tempted to seek a short-term inflation– unemployment advantage. What the central bank itself needed to run better monetary policies than before was better intellectual support (in the way Peter has recounted) and, unfortunately, an interregnum that showed all and sundry what it actually meant to live under policies tolerant of inflation. So, the big change that I’ve witnessed has been the willingness (so far so good) by so many governments around the world, including that of Canada, to be tied down through public commitments, up to a point and after their fashion, to reasonably specific, monetary policy goals. (And with this, we can for now say farewell to the many-faceted thing that is the preamble to the Bank of Canada’s enabling legislation.) Now let me look more closely at the Canadian experience, in particular the government–central bank dynamic, in the above regard. And like Peter, I’ll look specifically at the introduction of inflation targets. 1 Admittedly, and as David has contended, the Bank was less than totally clear in the late 1980s and early 1990s as to where it was going to end up in reducing inflation. In short, we never did clearly define ‘price stability’. In my view, we went as far as we could in the political economy climate that prevailed, and no one at least could mistake our sense of direction. And any controversy surrounding Canadian monetary policy was not so much about clarity as about the actual direction in which it was evidently going. In any event, where it was very clear was in its intra-Ottawa communications regarding the Goods and Services tax. These were that it would not be financing any second-round consequences (i.e., wage-price catchup processes) from the 1.5 per cent net direct increase in consumer prices that was preordained by that tax, starting at the beginning of 1991. So for a government with a huge debt and a large deficit, any general inflation reaction presented an immediate issue. 2 I am not suggesting that advocacy by the government of inflation targets in late 1990 was necessarily totally event driven, to the exclusion of principle. If one surveys the range of important economic policy
Discussion
3
4
5
6
75
changes anywhere, at any time, you’ll find that the role of events in the decision, even if only in helping the decider to ‘seize the day’, was a vital one. Opportunism, if not everything, is of crucial importance. (Another, particularly striking example is the way in which a financial confidence crisis helped Mr. Chrétien and Mr. Martin resolve actually to fight the deficit in 1995.) At the same time, to the extent that there was a principle involved in the government’s approach to inflation targets, it was one of avoiding having the Bank go ‘too far’.1 Consistent with this interpretation is the fact that in the bilateral negotiations over the form of inflation reduction targets, the government was pressing for higher rather than lower inflation targets. It also hedged its advocacy by arguing for a shorter rather than longer time span for inflation targets to be in effect. But if the Bank was going to surrender autonomy by committing itself to a formal agreement with the government (inevitably, as I saw it, as the junior partner), it wanted targets that were as ambitious as economically possible and as long as politically possible. The Bank wanted, in brief, medium- to long-term commitment, and the more ‘price stability’ (eventually) the better. In the end, a deal was struck. The fact that the arrangement apparently has delivered beyond the initial expectations of most commentators (I recall David being if not as sceptical as Peter, at least a fence-sitter – fair enough in the circumstances) has perhaps compensated for the fact that the longer-run governmental commitment to them that the Bank originally extracted was withdrawn in late 1993. So to summarize this important part of the Canadian monetary policy experience in terms of the game-theoretic model as critiqued by Peter: • It is questionable how much ‘principles’ regarding good monetary policy were actually involved. It can well be contended that government allowed itself to be nailed down (against general political instincts – Finance Minister Martin’s, for example) by the targets, not because it particularly wanted targets and government commitment, but because it was facing an interest-rate crisis that needed desperate measures. • To the extent that the Bank was in fact reined in by the targets, it was for principles entirely opposite to those prescribed by that model. This was confirmed by subsequent government actions.
76
Inflation Targeting in Canada
What can we really rely on in doing policy? I’m now going to turn to another leg of Peter’s commentary – namely, his suggestions that 1 we should allow, if we’re honest, that none of us knows that much about the transmission dynamics of monetary policy (at least much we can count on to work reliably on schedule); and 2 in such circumstances a more goal-linked objective (the ultimate goal of course has to be a healthy economy) gives the central bank salutary room to adapt and manoeuvre in working out what works best in terms of generating its worthy goal by monetary policy means. In short, don’t get trapped into using as a target, a particular component of the supposed, but ultimately rather dodgy, transmission mechanism, for example, M1, and its ilk. The caution embedded in Goodhart’s Law comes readily to Peter’s and my mind in this regard. But if the transmission mechanism is itself not that reliable, what does this tell us about the chances of hitting the inflation numbers? Peter’s answer is that a central bank can try, try, again. 1 I’ll not track through all the ways in which the Bank of Canada’s view of the transmission mechanism per se or the relative size and time-distribution of relevant parameter coefficients may have evolved over time. I remember it all rather imperfectly and in any case that’s not important. 2 What I will emphasize is that when the Bank did its initial homework in late 1990 on how to implement inflation reduction targets, it was extremely cautious as to its recipe for success in hitting the numbers. 3 While inflation may well be, in the end, always and everywhere a monetary phenomenon, our real-time decision framework involved charting a path from our overnight rate instrument through financial markets, through product and labour markets, and from there into consumer prices at a specific point in the future. Furthermore, given the great importance, especially in regard to a policy initiative that had attracted so much scepticism, of coming close to that initial target, we were determined to give ourselves as much time to get there as we thought we could decently get away with. I remain surprised that we got away then with as much lead time as we did.
Discussion
77
Perhaps indifferent outsiders thought we were going to fail anyhow, in any number of ways. 4 To be more specific, in early 1991, the likely model forecast error in terms of hitting a particular consumer price number, say two years or so out, seemed huge relative to the target. Accordingly, an executive decision was made that the announced performance band around the target number was going to be arbitrarily fixed at one percentage point on each side. This range was chosen on the grounds that anything wider would look ridiculous. Of such stuff is policy, even monetary policy, sometimes made. 5 In light of these technically uncertain beginnings, it is perhaps remarkable that the track record in inflation targetry has turned out as well as it has. Indeed, it seems to have turned out so well that it might be seen as a case of Goodhart’s Law in reverse. 6 Good luck, in particular an especially favourable non-monetary policy environment, may have been involved. Also, inflation target devotees have to wrestle with, or rationalize, the fact that the US has also greatly improved its inflation record without any support (at least explicit support) from inflation targets. Still, the targeting record both here (and, I guess, in the UK) is sufficiently gratifying that we can’t rule out the arrival, finally, of the kind of announcement/credibility effect that Peter refers to, and the kind of holy grail that central bankers (or any reasonable person) have been looking for from the time they started doing policy – the ability to get what you want without having to try so very hard. Now to my final set of comments.
The best of all possible targets? Peter is cautious. He allows that ‘inflation targeting seems to work’, but goes no further than that. Fair enough perhaps, though far less effusive than how the Bank itself promotes the exercise. Fifteen years is not that long – not a generation even. (David is also a bit of an economic historian.) And there are some questions about targeting procedures that remain to be satisfactorily answered – as there will no doubt always be. Peter poses a couple himself. The issue of asset price bubbles resonates strongly with academics, policymakers and, now, the media. He also highlights the more subtle fact that with less variability in inflation, central banks’ capacity to foresee, and forestall, emerging inflation pressures could well be eroded. But he doesn’t indicate where we might
78
Inflation Targeting in Canada
go next if fashion, or luck, changes. So let me add a thought in that regard. What about nominal income targeting as something to consider if and when inflation targets ‘don’t seem to work’? Without attempting to exhaust all the pros and cons, I’ll supply three sets of considerations in their favour. 1 First, in terms of the monetary policy transmission mechanism, nominal income targets have it over inflation targets in being impacted sooner by monetary policy actions, given that these actions have their effect fundamentally through aggregate demand. And, as Peter has strongly emphasized, given how little is regularly reliable in that mechanism, the shorter the transmission required to get a result the better. It is true that there is a reliability offset in the targets themselves, since income numbers are subject to revision in a way the CPI is not (for better or for worse). But that surely cannot be held to be a fatal flaw – just a complication to be dealt with. 2 Second, nominal income targets have useful automatic stabilization properties, especially in the face of supply-driven price shocks. Such shocks are matters that have to be handled on an ad hoc basis with inflation targets, and under Canadian arrangements through the awkwardness of special agreements between the Bank and the minister of finance. 3 Third, if the Bank of Canada is in tight with the government to the degree it already is under the arrangements through which inflation targets are formulated, a move to nominal income targets is surely a rather small further step to take. I would agree with anyone who argues that inflation targets ‘look’ better than income targets from a central bank’s supposed ‘independence’ standpoint, since they reinforce, at least in terms of appearance, the point that inflation is, in the end, a monetary phenomenon. But also from that standpoint, monetary aggregate targets might ‘look’ better than inflation targets, since ‘money’ is something a central bank actually produces, and we’ve gone past that hurdle. In any event, if the central bank is already tied up by the government through having to get government agreement on particular inflation targets, but in a process where the government is not called upon to explain its views (and does not), it does not seem that central bank independence, practically speaking, matters so very much any more.
Discussion
79
Notes 1. Let me add here that I don’t completely share the Ball and Sheridan view, expanded upon by Peter, to the effect that it was the ‘least independent’ central banks that embraced inflation targets. I don’t share it because I don’t regard (or at least didn’t regard) the Bank of Canada as being among the ‘least independent.’ Furthermore, measuring central bank independence is tricky. In fact, I had occasion once to examine closely the Eijffinger–Schaling yardstick that is included in his four-measure gauge of policy independence. And what was clear to me was that there were anomalies when one judged it against reality. Specifically, it seemed to me that both the Netherlands central bank (their ‘home country’ central bank, by the way) and the Bank of Canada had virtually the same legal set-up. However, in the results of Eijffinger and others, the Dutch were virtually at the top (home country bias?) while the Bank of Canada was at the bottom, alongside Australia. However, in point of fact, the Netherlands central bank had not been doing serious monetary policy for years (having a currency tightly pegged to the deutschemark), whereas the Bank of Canada in fact did it, and did it seriously.
3 The Lender of Last Resort: Lessons from Canadian History Angela Redish
Introduction Since at least the early nineteenth century, the merits of a lender of last resort for the financial system have been a matter of both theoretical and empirical debate. The debate has naturally included definitional issues – what is a lender of last resort, and competing interpretations of historical events and, for some, the larger question about the necessity for government institutions in a market economy. The goal of this essay is to review some of the recent theoretical literature on the lender of last resort and, since that review concludes that the necessity for, and desirable characteristics of, a lender of last resort depend on the empirical values of a variety of parameters, to examine historical evidence in the light of the theoretical frameworks. A one-sentence summary of David’s career would surely involve a phrase like ‘insightfully combining the history of monetary thought with a hands-on approach to monetary policy’; equally surely nothing combines these two as does the subject of lender of last resort, so it is no surprise that David has written on both the history of thoughts about lender of last resort, as well as on its current place in monetary policy, concluding in a recent (2004) paper that ‘the idea of the central bank as a lender of last resort is as old as the idea of central banking itself and … retains considerable relevance for modern central banking even in regimes from which positive stocks of high powered money have apparently disappeared, or are about to disappear for clearing and settlement systems.’ Indeed, questions about lenders of last resort provide the nexus between micro and macro approaches to monetary economics, and lie at the heart of our understanding of monetary economies. 80
Angela Redish
81
I proceed by briefly reviewing some of the theoretical debates over the necessity for a lender of last resort. I then describe the structure of the Canadian monetary and banking systems in the period 1870–1934, and characterize the outcome of that structure showing that the system was both stable (few bank runs, no panics) and unstable (many costly bank failures). I then argue that the system that emerged included a market component that absorbed solvent but illiquid banks, and a government component that supplied market liquidity.
Why have a lender of last resort? Depending upon the author, potential components of a lender of last resort policy include open market operations, a discount window, deposit insurance, and ‘bail out’ provisions. These complement other government roles in enhancing financial stability such as regulation and supervision of financial markets and their participants. Discussions of the need for a lender of last resort invariably begin with the propositions of Walter Bagehot who argued in the 1870s that the Bank of England should act as a lender of last resort by having an explicit policy of lending freely at a penalty rate, to solvent but illiquid banks, in times of financial crisis. Historians of economic thought have elaborated this story, emphasizing that it was not original to Bagehot and that Bagehot saw his prescription as particularly context-specific.1 Pride of place for first discussing the need for a lender of last resort is given to Francis Baring (1797) and then to Henry Thornton who articulated the role that the Bank of England played in the financial sector and argued that it should hold higher reserves than dictated by shortrun market considerations so that it could issue more notes in a time of crisis (Laidler 2002, 18). Laidler emphasizes the differences in the views of Thornton and Bagehot.2 He argues that Bagehot believed that the Bank of England’s position as an issuer of high powered money and central reserve for the banking system arose from past government intervention and it was only this specific historical context that necessitated the lender of last resort. In contrast, Thornton, and many others since, argued that natural economies of scale and network externalities lead market based financial systems to a central reserve. Bagehot’s views held such sway that there was little further analysis of the need for and characteristics of a lender of last resort for a century. Theoretical analysis was reignited by the classic paper on the causes of bank runs by Diamond and Dybvig (1983). They presented a model
82
The Lender of Last Resort
in which a bank run was a potential Pareto-inferior equilibrium – if everyone else withdrew their deposits it would be rational for you to withdraw your deposits. The model assumed that the essential characteristics of a bank included liquidity mismatch and sequential service. Suspension of convertibility, deposit insurance, or a discount window could eliminate the run equilibrium. Calomiris and Gorton (1991) argued that historically bank runs were not random sunspot events, but occurred in the face of negative fundamental shocks. They presented a model of bank liquidity crises wherein the value of the assets of a bank cannot be known with certainty to a depositor, and showed that if depositors get information that they think implies that the bank may be insolvent, then they will attempt to withdraw money before the possible failure. Since the downside risk (the cost of withdrawing if the bank is actually solvent) is minimal, depositors may respond to relatively small pieces of information. In this model, a lender of last resort could resolve the information problem by making loans only on good collateral. The proposition that a lender of last resort should lend to illiquid but not insolvent institutions could be interpreted as promoting lending to specific institutions or as lending to any institution that has good collateral.3 For example, Goodfriend and King argue that Bagehot recommended ‘monetary’ rather than ‘banking’ policy, because his prescription would have expanded the quantity of high powered money. They further argue (1988; 239) that ‘monetary policy [changing the quantity of high powered money] would have been both necessary and sufficient to prevent banking crises before the creation of the Federal Reserve; banking policy [all other policies of the central bank], in contrast would have been neither necessary nor sufficient.’ An alternative interpretation of Bagehot's views is that he supported emergency loans to threatened institutions, or organized intervention to support an institution. For example the Bank of England’s organization of a loan to Barings in 1890, and the similar assistance proposed by the New York Fed for LTCM in 1998, have been used as examples of Bagehot type policy. Others have argued against Bagehot’s prescription on the grounds that it is infeasible or too narrow. Charles Goodhart and Lixin Huang (1999), for example, made the case that the distinction between ‘illiquid’ and ‘insolvent’ is virtually impossible to make in the time frame that the authorities are operating in, and that it may not be relevant in that it may be socially optimal to provide funds to ‘insolvent’ financial institutions since contagion imposes negative external costs.
Angela Redish
83
The key ingredients to the need for a lender of last resort are incomplete information and illiquidity (imperfections in the market for liquidity) and the most recent papers have focused on spelling out the ways that these problems can generate a role for government intervention. For example, Gorton and Huang (2002) argue that a banking system comprising ‘unit banks’, will naturally develop a coalition of banks that will issue notes and act as a lender of last resort in the face of banking panics. Clearing houses in the US performed these functions under the national banking system. Why then did the Federal Reserve system emerge? They argue that for a coalition to function, banking panics are necessary; however, panics are welfare-reducing as they impede the operation of markets. Government intervention (through deposit insurance or lender of last resort facilities) can yield the benefits of the coalition without the necessity of banking panics.4 Gorton and Huang essentially conflate the roles of deposit insurance and lender of last resort, but Rochet and Vives (2004) argue that in today’s financial markets, behaviour in a crisis is driven by uninsured depositors. They model a financial system in which the banks hold uninsured deposits (e.g. Certificates of Deposit, CDs) and face liquidity shocks. The impact of the shocks depends on the underlying soundness of the banks, but there is a ‘positive probability that a solvent bank cannot find liquidity assistance in the market’ motivating emergency discount-window loans. A driving force in the model is the potential discounting of assets in a ‘fire sale’, which could reflect a liquidity premium or a sale of ‘bad loans’; it is assumed that the central bank monitors the banking system and so, unlike other participants in financial markets, can discriminate between these two motives.
The structure of the Canadian banking sector before 1935 In 1868 there were 33 chartered banks in Canada with approximately 123 branches. The system expanded in the early 1870s to 51 chartered banks with 230 branches; expansion of the branch network continued and by 1904 there were 38 banks with 1145 branches. By 1925, consolidation in the industry (discussed below) left 11 chartered banks, although the branch network had expanded to 3840 branches.5 Relative to the banking systems of the UK and the US, the Canadian chartered banks were lightly regulated. Charters were created by an Act of Parliament and the Treasury Board permitted a bank to open only after a minimum of $250,000 (half the minimum authorized capital) had
84
The Lender of Last Resort
been subscribed. Charters were temporally limited and the government renewed all charters at roughly 10-year intervals by passing a Bank Act which renewed the charters and typically introduced some changes to the regulatory environment. All charters imposed double liability on shareholders and in the event of a bank failure there was frequently a call on the shareholders. Banks were required to give monthly reports (basically balance sheets) to the Minister of Finance, which were then published in the Canada Gazette. The banks could lend (essentially) only on real bills. They were permitted to issue bank notes in denominations of $5 and above, with the requirement that such notes be convertible on demand into legal tender (gold or Dominion notes) and that a bank’s total note issue not exceed its paid-in capital. In 1870, note issues were 40% of demand liabilities (notes plus demand deposits) and in 1925, they still represented 25%. In 1935, the Bank of Canada began operations and eased the chartered banks out of the note issue. In addition to chartered bank notes, Dominion notes were issued by the Federal government, in denominations less than $5 (i.e. $4, $2, and $1) and in large denominations6 which the chartered banks used to clear inter-bank balances. Dominion notes were convertible on demand into gold at Receiver-General offices across the country. In 1870, the issue was constrained by the requirement that up to $9 million be backed by 25% in either gold or Canadian debentures payable in sterling and guaranteed by the British government, with a minimum of 15% in gold. Over time, the fiduciary issue rose, and by 1914, the limit was $50 million with a minimum of 25% gold backing. There was no required reserve ratio against either notes or deposits issued by the chartered banks, although 40% of cash reserves had to be held in Dominion notes. In practice, this was rarely a binding constraint. In the absence of the legal requirement of holding reserves, the banks typically held about 15%.7 In the nineteenth century, the banks did not perceive a limit on their note issues (relative to deposits.) By 1906, the note issue, particularly in the Fall, was close to the constraint that it not exceed the paid-in capital. In the Fall of 1907, the Minister of Finance allowed a 15% increase in note issues during the Fall on which they were obliged to pay 5% p.a. interest (see below for further discussion of this episode). In 1908, this facility was made permanent. In 1913, the creation of the Central Gold Reserve created an alternative method for the banks to expand their note issue; they could deposit legal tender in the Central reserve in Montreal and then issue (essentially 100% backed) notes.8
Angela Redish
85
In 1890, following losses to bank note-holders in some notorious bank failures, discussed below, the Bank Act revisions included the establishment of a Bank Circulation Redemption Fund. Each bank was required to pay into the Federal government 5% of its circulation on which they would be paid 3% interest. In the event of a bank suspending convertibility of its notes, this Fund would be used to pay 6% interest on its notes until they were redeemed. Notes were also made a first charge on the assets of the Bank, and after 1890, no losses were experienced by note holders. In 1891, the Canadian Bankers Association (CBA) was established as an industry association, and in the 1900 Bank Act revision it was given jurisdiction over the supervision of the clearing houses. Clearing houses were established relatively late in Canada, possibly in part because the small number of banks and the branching system enabled the majority of transactions to be cleared internally. In 1887, a clearing house was established in Halifax, in 1889 in Montreal, and two years later in Toronto and Hamilton. In 1927, a central settlement plan was adopted, whereby each bank telegraphed its balances to Montreal. The extent of the influence of the CBA is the subject of considerable debate. The Act incorporating the Association gave them authority to appoint the curator for a suspended bank; jurisdiction over the note issue, that is, to keep track of notes issued and destroyed; as well as management of the clearing houses. They openly operated to control the interest rate paid on deposits at 3%.9 Although the banks were required to publish monthly balance sheets, there was no requirement that they be inspected or audited. In 1923, following the failure of the Home Bank, the office of the Inspector General of Banks was created and banks were required to submit to government inspection. Until 1914, all Dominion notes were convertible into gold, and at the margin the notes were 100% backed by gold as described above. At the onset of the war, two changes were introduced, one temporary and one permanent. The convertibility of notes into gold was suspended, and not reintroduced until July 1926. In addition, the government permitted the issue of Dominion notes in loans to chartered banks. Essentially, this opened a discount window for the banks where they could borrow Dominion notes against good collateral. To recap, the Canadian banking system prior to the creation of the Bank of Canada had a number of special, if not unique, features. It was a concentrated system of largely unregulated nationwide branching banks, which could issue notes against general assets and whose shareholders had double liability.
86
The Lender of Last Resort
Experience of the system Banking systems are measured in multiple dimensions, efficiency, contribution to growth, and stability. Here I am interested simply in the question of the extent and sources of the stability of the system in the absence of a lender of last resort. The Canadian banking system is touted for its stability, particularly in contrast to US experience. However, this conclusion needs to be carefully nuanced. Undoubtedly, the US banking system experienced far more failures than did the Canadian system, but the size of each Canadian bank meant that a failure was likely to entail far greater losses than a US bank failure. Table 3.1 shows that Canadian depositors lost more than US bank depositors in the period 1881–1900, and arguably more than US national bank depositors in the 15 years earlier. Table 3.2 lists the Canadian banks that failed, and contextualizes their scale by the assets of the total banking system. The key difference between the US and the Canadian banking systems was not the lack of bank failures, but the absence of regional suspensions of convertibility and of bank runs. In the US there were banking panics in 1873, 1884, 1890, 1893, 1907, and 1914.10 In all six cases, clearing house loan certificates were issued, and during the ‘crises’, which would include 1914, there were widespread suspensions of convertibility. In Canada there was no issue of clearing house loan certificates. A related difference between the experiences of the US and Canadian banking systems lies in the elasticity of the bank notes system. Figure 3.1 shows the (negative of) the percentage change in the outstanding bank notes (i.e. excluding Dominion notes) between October of one year and January of the next. The average decrease was 15%, and in the first decade it was about 20%.11 Data for the US from Friedman and Schwartz (1963) show that the quantity of currency did not fluctuate from October to January.12 In contrast, the available evidence on short-
Table 3.1
Losses on deposits Canadian banks
1865–80 1881–1900 1901–20
0.01* 0.16 0.01
US national banks 0.06 0.08 0.01
US (all banks) 0.21 0.15 0.05
* This figure is for 1867 (Confederation) to 1880. In 1866, the Bank of Upper Canada failed, so the figure for 1866–80 is 0.07. Source: Bordo et al. (1996, 65).
Angela Redish Table 3.2
87
Canadian bank failures, 1868–1934
Year
Assets ($m)
Bank
1868 1873 1876 1879* 1879 1879 1881 1883 1887 1887 1887* 1888* 1893 1895 1899 1905 1906 1908* 1908 1908 1910 1910 1914 1923
1.2 0.2 0.8 1.3 3.0 0.7 1.0 3.8 3.2 1.8 1.3 4.9 2.0 9.5 2.3 0.7 15.9 1.5 0.3 19.2 0.8 2.6 1.5 27.0
Commercial Bank NB Bank of Acadia, NS Metropolitan B., Q Stradacona, Q Consolidated Mechanics, Q Bank of PEI Exchange Bank Central Bank, Ont Maritime Bank London, Ont Federal Bank, Commercial Bank, Man Bque du Peuple, Bque Ville Marie Bank of Yarmouth, NS Ontario, Bank Bque de St. Hyacinthe Bque de St. Jean Sovereign St Stephen’s, NB Farmers Bank B of Vancouver Home Bank
Paid to noteholders (%) 100 0 100 100 100 57.5 59.5 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
Paid to depositors (%) 100 0 100 100 100 57.5 59.5 66.5 99.6 10.6 100 100 100 75.2 17.5 100 100 100 30.3 100 100 0 0 25
* indicates voluntary liquidation. Source: "Statement showing the Number of Chartered Banks that have gone into liquidation since 1867" Department of Finance evidence. [Macmillan] Royal commission on Banking and Currency in Canada, 1933. Appendix 15.
term interest rates shows dramatic seasonal fluctuations in the US, and virtually none in Canada. An additional feature of the Canadian banking system was the practice of holding secondary reserves in the New York money market. When asked about this practice in 1923, Sir Edmund Walker (General Manager of the Canadian Bank of Commerce) defended the practice (of not keeping the money in Canada) with the comment that ‘The practice is older than Confederation. … Canada has been saved almost every time when there has been a serious panic the United States by our ability to withdraw from New York our balances there to serve our people at home. Saved in ’93, saved in ’97, saved many, many times.
88
The Lender of Last Resort 300
25.00%
250
20.00%
200 15.00% 150 10.00% 100 5.00%
50 0
18 1880 1882 1884 1886 1888 1890 1892 1894 1896 1998 1900 1902 1904 1906 1908 1910 1912 1914 1916 1918 1920 1922 1924 1926 28
0.00%
Seasonal (left axis) Figure 3.1
Notes (annual)
Seasonal contraction of notes: October to January
The amount in New York must be added to the cash reserves in Canada.’ This reader is somewhat puzzled by the idea that reserves were safer in New York where there were occasional suspensions of convertibility than in Canada where there weren’t. It also seems unlikely that the use of the New York money market can explain differences in the monetary experiences of the US and Canada since US regional banks would also have had access to this market.
Implicit deposit insurance Was the Canadian banking system inherently stable, or did its stability arise primarily from implicit deposit insurance provided by either the Federal government or the CBA? In the latter view, dubbed the ‘imminent failure hypothesis’ in a recent paper, Kelly Bushe and John Chant,(n.d) the consolidation of the Canadian banking sector between 1900 and 1920 occurred because large banks were forced to merge with insolvent banks (see Table 3.3). The argument that the absence of bank failures reflected regulatory forbearance has been made most clearly by Kryzanowski and Roberts, who argue that the Minister of Finance had made it clear that he would not allow depositor losses in the event of a bank failure.
Angela Redish Table 3.3
89
Canadian bank amalgamations, 1868–1920
Year
Bank
Absorbing bank
1875 1875 1875 1875 1883 1901 1901 1902 1903 1903 1905 1906 1907 1908 1909 1910 1911 1912 1912 1913 1913 1914 1917 1918 1918 1919
St Lawrence Niagara District City Bank Royal Cdn Union Bank, PEI Summerside, PEI Bank of BC Commercial B., NS Halifax Bdg Co. Exchange Bk, NS People’s Bank, NS Merchants Bk, PEI People’s Bk, NB Crown Western Union, NS United Empire B Eastern Twps Traders Bank of NB Bque Internationale Metropolitan Quebec Bk of BNA Northern Crown Ottawa
N/A Imperial Consolidated Consolidated Nova Scotia Bank of NB Commerce Union Bk, Hx Commerce Montreal Montreal Commerce Montreal Northern Standard Royal Union Bk Commerce Royal Nova Scotia Home Nova Scotia Royal Montreal+losses Royal+losses Nova Scotia
Assets ($m) N/A 2.8 4 1 0.3 16 2 6 1 6 2 1 1 6 25 1 28 51 12 3 12 21 78 28 66
Source: Beckhart (1929, 338)
The ‘imminent failure hypothesis’ is not quite a myth, but it does have some resonance, for example, Williamson (1989) and Kryzanowski and Roberts (1993). In fact, the historical accident that the failures of the Ontario Bank and the Sovereign Bank, in 1906 and 1908, respectively, immediately preceded the National Monetary Commission may explain part of the durability of the ‘imminent failure’ hypothesis. Joseph Johnson wrote the chapter on the operation of the Canadian banking system and focused attention on the response of the Canadian banks to those failures. Both were ‘open door’ liquidations. The Bank of Montreal took over the liabilities (and assets) of the Ontario bank, and a consortium of 12 banks assumed the liabilities of the Sovereign. Williamson cites Johnson’s account of the takeover of the Ontario Bank and concludes that ‘The collective behaviour of Canadian banks
90
The Lender of Last Resort
not only served to minimize the costs of liquidating insolvent institutions; it also appears to have prevented widespread banking panics’. But, Johnson (1910, 125) concluded his discussion of the takeover with the comment that ‘it is generally understood that the venture was a profitable one for the Bank of Montreal’, suggesting that this was not a bail out of an insolvent institution. The coalition that supported the Sovereign Bank was not so lucky, and Carr, Mathewson, and Quigley (1994; 6) suggest that the banks lost about 3–4% of the funds advanced. The National Monetary Commission interviewed several Canadian bankers in 1909 and focused on precisely this issue. When Edward Vreeland stated: ‘I understood it was not the exception but the practice for the other banks or the association to take over a failed bank?’, the Secretary of the CBA stated: ‘No. That has only happened in two cases. … The practice is to let them close.’ Sir Edward Clouston was asked the same question: ‘I notice that in the case of the Sovereign and Ontario they were taken over and liquidated by the chartered banks or a combination of them. Would you say that has become the policy of the chartered banks in case of a weak bank or one that is about to suspend?’ His answer, again, ‘No’.13 The words of the Minister of Finance in 1923, in the aftermath of a major bank failure, provide a clear description of policy. He spoke of a merger that had occurred several years earlier: ‘it [the bank being absorbed] was not insolvent; nor indeed was it on the edge of insolvency. The Directors saw however that the bank was on the decline and that the day was coming when it could not be successful.’14 Evidence from share prices supports the argument that banks that were absorbed between 1900 and 1920 were not insolvent. Carr, Mathewson, and Quigley (1994) examine share prices of the absorbing banks before and after the merger and find that in no case did absorption lower the share price. Busche and Chant conduct an event study in which they examine the performance of 19 banks, some mergers, some mergees, and some neither, for which data on dividends and capital gains are available for the period 1900–1931. They find that there were no significant differences in the returns to an investor in the three different types of bank, and they conclude that there is no evidence that mergers were ‘of strong stable banks being forced to absorb weak imminently failing banks’. Finally, the fact that Canadian banks did fail and depositors did suffer losses seems prima facie evidence that the depositors were not insured. In 1910, the Farmer’s Bank failed with all deposits written off and the
Angela Redish
91
double liability of the shareholders used to redeem bank notes.15 In 1914 the Bank of Vancouver failed, and again depositors suffered losses despite the call on the double liability. Finally, in 1923, the Home Bank failed – a bank with approximately 1% of Canadian demand deposits – and again double liability was called. Depositors in general received 25 cents on the dollar from the Bank’s assets and capital. Again, depositors and shareholders argued that the government was responsible for the Bank continuing business after it was insolvent, and the government appointed a Royal Commission to investigate. The commission concluded that the Minister of Finance had ‘acted inappropriately’ in not auditing the bank in 1916 when directors of the Bank had complained about misrepresentations in the monthly reports. The government agreed to pay 35% of deposits under $500, to address cases where the losses had created hardship.
The evolution of standing lending facilities There is one episode prior to 1914, in which some authors argue the Department of Finance acted as a lender of last resort. This occurred in the Fall of 1907. At that time, the US was in a financial crisis and short-term interest rates rose to 21%.16 The Canadian banks tightened lending and wheat farmers, anxious that loans to finance the movement of the crop not be restricted, persuaded the government to lend Dominion notes to the banks for that purpose. Five million dollars were borrowed and repaid between November 1907 and May 1908 (M1 was approximately $240 million in 1907, and base about $90 million). While this has been seen as a lender of last resort action, the banks had not asked for the programme (it was grain farmers who had lobbied for it), and it was not intended to provide liquidity to any specific bank. It seems more appropriate to see it either as a subsidy to farmers (Rich, 1989) or as an expansionary monetary policy. In August 1914, when the threat of war became imminent, the Canadian banks approached the Minister of Finance, because they expected a run. The minister acceded to their request and, by Order in Council (ratified by Parliament later in the month), he declared bank notes legal tender, made Dominion notes inconvertible, and established a discount window where the Canadian banks could borrow Dominion notes (on good security). Dominion notes issued under this authority (when ratified, known as the Finance Act) did not have any provision for gold cover, but this did not pose a threat as long as convertibility was suspended.17
92
The Lender of Last Resort
In 1923, the government passed a legislation that continued the discount window provisions of the Finance Act, but required that bank notes and Dominion notes be convertible into gold after July 1, 1926. At that point, the Finance Act became potentially inconsistent with the Dominion Notes Act, and in 1928, when interest rates in New York rose above the (rarely altered) discount rate there was a clear arbitrage opportunity. The banks could borrow Dominion notes, convert them into gold, ship the gold to New York to invest, and earn considerably more than the cost of the borrowed Dominion notes.18 The government used moral suasion to prevent the banks from doing this either directly or indirectly, and Canada effectively left the gold standard. While the banks had the right to borrow under the Finance Act, and on occasion did so, borrowings were slight, and there is no suggestion in literature that they were necessary – the Department of Finance simply represented a cheap source of funds. In July 1933, the government established a Royal Commission on Banking and Currency in Canada to examine the Canadian banking system and the ‘advisability of establishing a central bank’. In September 1933, after holding hearings across the country, the Commission reported in favour of establishing a central bank. The objectives of the bank (Article 236) should be to provide ‘a more rational and unified control over the credit structure’, an instrument for exchange rate policy and an institution to advise governments and facilitate international monetary cooperation. In the report there is no mention of the need for a lender of last resort, or of the role of central banks as lenders of last resort, other than the undefined (and repeated) ‘regulation of the volume of credit and currency’. Even in the discussion of bank failures and mergers in the description of the Canadian banking system, there is no mention of a potential role for a central bank. In the Act establishing the Bank of Canada (23–25 Geo V c43) there is equally no implicit or explicit mention of the lender of last resort.19 It was not until 1997 that the Bank of Canada Act explicitly introduced language referring to financial stress, viz., ‘18. The Bank may … (g.1) if the Governor is of the opinion that there is a severe and unusual stress on a financial market or financial system buy and sell any other securities … to the extent determined necessary by the Governor for the purpose of promoting the stability of the Canadian financial system’. In summary, the Canadian banking system never suffered the banking panics that characterized the US and the British banking systems in the
Angela Redish
93
nineteenth century before they developed central banking institutions. The explanation would seem to lie in the combination of the ‘elasticity’ of the currency, the branching system of the Canadian banks, and the acceptance of a high degree of concentration in the banking system.
Conclusion Economic theorists can design environments where a lender of last resort improves welfare. Yet the environments are specific – certain markets exist and others don’t; some kinds of moral hazard are more significant than others. These models can alert policy makers of the factors that should be in play, but it seems unlikely that they will ever be definitive. Equally, we can hope that empirical evidence on the parameters at issue will be thin. Central banks will be flying by the seat of their pants for a while. That said, history can provide another source of information about the realm of the possible. The widespread acceptance that the Bank of England became a ‘central bank’ after the 1870s when it accepted its role as a lender of last resort, and the explicit founding of the Federal Reserve system in the aftermath of the Panic of 1907, leave an impression that banking systems were unstable in the absence of a central provider of liquidity. The Canadian historical experience provides a counter to these experiences and highlights their dependence on particular institutional environments, in turn very much the product of a particular chronology and political process.
Notes This paper began while I served as Special Advisor to the Bank of Canada, and benefited greatly from discussions with Bank staff. The opinions and errors are all mine. 1. O’Brien (2003) states that Lombard Street ‘rationalized and analyzed an idea 80 years in the making’. 2. One difference that Laidler notes is that the earlier writers were framing their policies at a time when Bank of England notes were inconvertible, while Bagehot wrote when the Bank was constrained by the exchange rate. See also Martin (2002) who rationalizes the Fed lending at a discount while Bagehot recommended lending at a penalty rate by the difference between policy in a fiat money and a gold standard environment. 3. The word ‘good’ here is significantly ambiguous – good in ‘normal’ times or good in the ‘crisis’? 4. Gorton and Huang motivate their description by the contrast between the US and Canadian experiences but do not incorporate the higher rate
94
5.
6. 7.
8.
9.
10.
11. 12. 13. 14. 15.
16.
17.
The Lender of Last Resort of losses to Canadian depositors in the late nineteenth century into their analysis. In 1895, the five largest banks owned 41% of the assets of the chartered banks; by 1927 they owned 82% (Bordo, Rockoff, and Redish 1996; 65). These ‘large legals’ were only legal tender among banks, and ranged in denomination from $1000 up to $50,000. For example, in an interview with the National Monetary Commission, Mr Coulson with the Canadian Bankers Association (CBA) reported that ‘There is an understanding with the banks of the association that we are to keep 15% in cash.’ He is not specific about what it is 15% of, but ratios of gold plus Dominion notes to notes plus demand deposits were typically over 15%. It is not clear why the banks chose not to raise their capital stock, although fear that this would lower the dividend rate may be part of the explanation. In his report to the National Monetary Commission, Joseph Johnson cited an anecdote that sheds light on both the significance of the clearing house and the operations of the CBA (which he considered controlled by the ‘older large banks’. The general manager of a younger bank wished to pay 3.5 % on deposits and was denied access to the clearing house. He declared: ‘I find that I can get along very well without the use of the clearing house and my not belonging to it puts them to a great deal more trouble than it does me’ (Johnson 1910, 135). The dissident bank was likely the Northern Crown Bank which was created in 1908, and in 1918 was absorbed by the Royal Bank (Evidence of Mr Knight before the National Monetary Commission.) Sprague (1910) refers to the events of 1873, 1893, and 1907 as crises, 1884 as a panic and 1890 as a period of ‘financial stringency’. Calomiris and Gorton (1991) add 1914 to the list. Dominion notes in the hands of the public were less than 20% of the notes in the hands of the public, and had little seasonal variation. See also the chart in Beckhart (1929, 377), showing a 2% seasonal fluctuation of National Bank notes. Both interviews published in the National Monetary Commission, Vol. 9. Knight p. 65, and Clouston p. 180. Cited in Beckhart (1929, 341). Beckhart, p. 336. Depositors and shareholders petitioned the government for compensation and a Royal Commission was established to investigate the responsibility of the government for allowing the Bank to begin operation (i.e. the issue of a Treasury Board certificate permitting the Bank to open). The Commission concluded that there had been misconduct in connection with the application for the certificate but that the Department of Finance was not liable. See Carr, Mathewson, and Quigley (1994). Homer and Sylla (1996; 358). Twenty-one per cent is the monthly average ‘high’ reported for call loans. The ‘extreme quotation’ is 125%, but it is not clear what type of loan was being priced. The Act was in fact an ‘Act to Conserve the Commercial and Financial interests of Canada’, perhaps a more appropriate title!
Angela Redish
95
18. See Shearer and Clark. The exchange rate moved slightly outside the gold points after February 1929, and remained so until explicit exchange controls were introduced in 1931. Advances under the Finance Act were $1.4 million, $14.l million, and $57.6 million in January 1927, 1928, and 1929, respectively. 19. The preamble to the Act is the only place where the Bank’s objective function is even implied and it reads: ‘Whereas it is desirable to establish a central bank in Canada to regulate credit and currency in the best interest of the economic life of the nation, to control and protect the external value of the national monetary unit and to mitigate by its influence fluctuations in the ganaral level of production, trade, price, and employment, so far as may be possible within the scope of monetary action and generally to promote the economic and financial welfare of the Dominion: Therefore …’. This language is taken almost verbatim from c.206 of the Macmillan Commission report.
Bibliography Bagehot, W. (1873) Lombard Street: A Description of the Money Market. London: H. S. King. Beckhart, B. H. (1929) The Banking System of Canada. New York: Henry Holt and Co. Beckhart, Benjamin H. (1929) ‘Canada’, in Foreign Banking Systems, H. Parker Willis and B. H. Beckhart (eds). London: Pitman and Sons, 289–488. Bordo, Michael, Angela Redish and Hugh Rockoff (1996) ‘Two Long Roads to Stability: Canadian and American Banking 1870–1925’, Financial History Review, Vol. 1 (April) 49–68. Busche, K. and J. Chant (n.d.) ‘The Imminent failure hypothesis in Canadian Banking’ ms. Simon Fraser University. Calomiris, C. and G. Gorton (1991) ‘The Origins of Banking Panics: Models, Facts and Bank Regulation’ in G. Hubbard, Financial Markets and Financial Crises. Chicago: University of Chicago Press, 107–73. Carr, J., G. Mathewson, and N. Quigley (1994) ‘Stability in the Absence of Deposit Insurance: The Canadian Banking system, 1890–1966’ ms. Institute for Policy Analysis, Toronto. Charles Goodhart and Haizhou Huang (1999) ‘A Model of the Lender of Last Resort’, IMF Working Paper, 99/39. Douglas W. Diamond and Phillip H. Dybvig (1983) ‘Bank Runs, Deposit Insurance and Liquidity’, Journal of Political Economy, Vol. 91, 401–19. Friedman, Milton and Anna Schwartz (1963) A Monetary History of the United States, 1867–1960. Chicago: University of Chicago Press. Goodfriend, M. and R. King (1988) ‘Financial Deregulation, Monetary Policy, and Central Banking’ Federal Reserve Bank of Richmond Economic Review, (May/June): 3–22. Gorton, G. and L. Huang (2002) ‘Banking panics and the origin of central banking’ NBER Working Paper 9137. Homer, Sydney and Richard Sylla (1996). A History of Interest Rates (third ed). New Brunswick, NJ: Rutgers University Press.
96
The Lender of Last Resort
Johnson, Joseph F. (1910) The Canadian Banking System. Report prepared by the National Monetary Commission for the US Senate, 61st Congress, 2nd Session Senate Doc. 853. Washington, DC: Government Printing Office. Kryzanowski, L. and G. Roberts (1993) ‘Canadian Banking Solvency, 1922–40’ Journal of Money Credit and Banking, May: 361–76. Laidler, D. (2002) ‘Two Views of the Lender of Last Resort: Thornton and Bagehot’ University of Western Ontario ms. —— (2004) ‘Central Banks as Lenders of Last Resort – Trendy or Passe?’ University of Western Ontario Working Paper 2004–8. Martin, Antoine (2002) ‘Reconcilling Bagehot with the Fed’s Response to Sept. 11’, Federal Reserve Bank of Kansas City. Research Working Paper: RWEP 02-10. O’Brien, Denis (2003) ‘The Lender of Last Resort Concept in Britain’, History of Political Economy 35(1): 1–19. Rich, G. (1989) ‘Canadian Banks, Gold and the Crisis of 1907’, Explorations in Economic History 26(2): 135–60. Rochet, J.-C., and X. Vives (2004) ‘Coordination Failures and the Lender of Last Resort: Was Bagehot Right after all?’, Journal of the European Economic Association 2(6):1116–47. Sprague, O. M. W. (1910) History of Crises under the National Banking System. National Monetary Commission, 61st Congress, 2nd Session, Senate, Doc. No. 538. Williamson, Stephen D. (1989) ‘Bank Failures, Financial Restrictions, and Aggregate Fluctuations: Canada and the United States, 1870–1913’, Federal Reserve Bank of Minneapolis Quarterly Review, Vol. 13, No. 3 (Summer 1989) 20–40.
Discussion Neil T. Skaggs
Angela Redish provides us with a topic that can be addressed from a variety of directions. It is at once • a topic near and dear to David Laidler’s heart, • an attempt at writing the history of a (contested) topic, • a description of the structure and historical evolution of the Canadian banking system, • an examination of the stability of the Canadian banking system relative to the US banking system, • a discussion of whether an implicit deposit insurance scheme in fact existed before the official system was created, • a brief history of the evolution of Canada’s standing facilities and of the Finance Act of 1914, and • a very brief look at the establishment of the Canadian central bank in 1933. I want to focus on only a subset of the issues raised here. First, defining exactly what it means to be a ‘lender of last resort’ isn’t easy. At what point did the Bank of England become a lender of last resort? That depends partially on whether one considers expansive monetary policy during periods of financial stress to be acting as a lender of last resort. I am thinking particularly about the financial stress of February 1793 in Great Britain. War between Britain and France had just broken out, and a number of reputable banks and commercial houses found themselves illiquid. The Bank of England stood firm during the crisis, increasing its loans and note circulation and, for the time being, saving the convertibility of the pound. However, in The Paper Credit of Great 97
98
The Lender of Last Resort
Britain, Henry Thornton noted that the Bank’s refusal to lend to country banks which lacked ‘approved London securities’ led to multiple failures and made the Bank’s overall situation worse than it need have been. Thornton remarked that the ‘event … showed that the relief of the country was necessary to the solvency of the metropolis. A sense of the unfairness of the burthen cast on the bank by the large and sudden demands of the banking establishments in the country, probably contributed to produce an unwillingness to grant them relief’ (p. 181, n.). This episode might be classified as one in which the Bank of England engaged in an expansionary monetary policy while refusing to operate as a lender of last resort. Yet had the Bank been willing to set aside its rules and lend to the country bankers, as it loaned to other bankers, the distinction between the two sets of banks would not have been evident, and the entire affair would have appeared as only an expansionary policy. The events of 1797, when Great Britain was forced off the gold standard, also make clear the fact that successfully acting as a lender of last resort is much easier when the currency is inconvertible than when it is convertible. Again referring to Thornton’s Paper Credit (pp. 112–13), the London commercial houses were quite willing to continue their operations using paper pounds issued by the Bank of England. However, country banks, acting through their London agents, withdrew gold from the Bank at such a rate that suspension was necessary. Yet Thornton notes in the course of his discussion, the Scottish banks, which depended on the Bank of England for their gold, just as did the English country banks, did not face a demand for gold by their depositors and hence did not demand gold from London. So, another factor in the lender-of-last-resort puzzle emerges: the attitude of the populace towards the banking system goes a long way toward determining whether a lender of last resort is even needed. By extension, if it is needed, the prevailing attitude of depositors determines the extent to which a central bank must go to stanch a generalized bank run. On to a second issue, the relative stability of Canadian banks compared to US banks: the period before 1914 raises some tricky issues. I was surprised that the record of US banks in comparison to Canadian banks wasn’t much worse than it was, given the injurious restrictions placed upon US banks by the National Banking System. But then, the US problem wasn’t so much one of inadequate capital, and hence bank failures, as inadequate liquidity. On this score, the blame can be laid squarely on the legislation that prevented the US system from
Discussion
99
expanding the quantity of bank notes to meet demand. Almost by definition, a national lender of last resort would have addressed the problem. But so would have a removal of the restrictive cash and bond reserve requirements that were largely to blame for the liquidity crises in the first place. Institutional frameworks are a prime determinant of the need for, and success of, a lender of last resort. As always, the devil is in the details.
References Thornton, Henry. 1802 [1978]. An Enquiry into the Nature and Effects of the Paper Credit of Great Britain. Fairfield, NJ: A. M. Kelley.
4 The Asian Economic Revolution and Canadian Trade Policy John Whalley
Abstract This chapter discusses the broad orientation of Canada’s trade policy relative to two major historical phases of development based on a secure national market behind the National Policy from 1879 until the 1930s, and progressive integration with the US through Bilateral Agreements (1930s), the Auto Pact (1965), the Canada–US Free Trade Agreement (1987), and the North American Free Trade Agreement (NAFTA) (1994). Currently, Canada exports approximately 85% to the US, but imports from China account for 8% and are growing at over 20% a year. Sharply unbalanced (surplus) trade with the US is counterbalanced by unbalanced deficit trade with China. A scenario of elevated growth in Asia (principally China, India, and the Association of Southeast Asian Nations or ASEAN) poses challenges of relative disintegration from North America and growing global integration centred on Asia. Seemingly, a series of implications follow, including positioning Canada within the emerging network of regional agreements in Asia, more resource-based and Western Canada focused trade and infrastructure development, and responding to capital market integration with Asia. Broader Issues are raised including the potential adjustments facing Centre Canada as Asian imports of manufactures displace both imported manufactures for the US and domestic production.
4.1 Introduction and summary Canadian trade and external economic policy has undergone two major phases since Confederation in 1867. The first was associated with the events surrounding Confederation itself and the search for 100
John Whalley 101
secure markets abroad reflected in the Elgin-Marcy Treaty with the US (1854) and its termination in 1866, and subsequently, the search post-Confederation for trade arrangements with the US and the UK. The failure to develop secure arrangements lead to the National Policy of Prime Minister, Sir John A. MacDonald, in 1879 and development behind a national tariff to be based on transportation subsidies and Westward settlement. In the 1930s, this led to a three-part tariff with the highest tariff against the US and Commonwealth preferences agreed in the 1932 Ottawa conference. The second phase can be seen as progressive North American integration beginning with bilateral negotiation with the US in the 1930s and is characterized by ever-deepening trade engagement resulting in the Auto Pact (1965); the Canada–US Free Trade Agreement (1987) and subsequently tri-lateralized under the NAFTA (1994). This is despite multilateral engagement and negotiation first in the General Agreement on Tariffs and Trade (GATT) and the World Trade Organization (WTO). During this phase, Canada’s export share with the US grew from over 30% in the late 1990s to 75% at the time of the Canada–US Free Trade Agreement to approximately 85% today. This growing trade is heavily concentrated in automobiles and parts following the 1965 Auto Pact, with Canada’s production in manufacturing concentrated in Ontario. This trade growth has also been central to four decades of consistent growth in Central Canada. It is argued in this chapter that high growth rates in developing Asia, and most notably in China and India, if they persist for several decades (as has been the subject of conjecture in discussion of the BRICs (Brazil, Russia, India, and China))1 will inevitably affect trade patterns and subsequently, Canada’s interests in access to foreign markets. The continuation of these growth rates for the coming decades in developing Asia is what I term the Asian Economic Revolution. Currently, Canada has 8% of its imports from China which are growing at 22% per year.2 With only 1.6% of exports going to China, this large trade deficit partially accommodates a large trade surplus with the US (the US only accounts for 56% of Canada’s imports). The prospect is for China and other Asian imports to progressively displace imported manufactures from the US, and for Canada’s export of resources to Asia to grow. This in turn will influence both Canada’s negotiating approach to economic integration abroad, and in turn have implications for domestic policy. To narrow the focus of the chapter, I discuss only the larger countries in developing Asia (China, India, the ASEAN member-countries) and do not comment on developed Asia (Japan, Korea, Taiwan, Hong Kong).
102
The Asian Economic Revolution and Canadian Trade Policy
Japan and Korea, for now, have larger trade shares than all countries in developing Asia except China, however, the presumption is that this will change in the next few decades. For Canadian policymakers, the growth prospects for Asia over the next few decades have significant implications for policy more broadly. Will US imports of manufactures be progressively displaced by Asian suppliers with ever-growing surplus trade with the US and deficit trade with Asia? Will trade-based manufacturing in Ontario and Quebec decline and growing resource exports occur from Western provinces? If trade shares with Asia have the potential to grow and access to these markets becomes more important, how should Canada modify its tradenegotiating stance? The Asian economies are now seemingly engaged in a wave of new regional negotiation that goes beyond conventional goods and services negotiations to cover a range of less conventional issues (such as competition policy, mutual recognition, movement of persons, and others) and these agreements are as much process oriented as agreements on limitations on the use of trade instruments. Should Canada participate more actively in such negotiations even if seemingly different in focus from WTO or bilaterals with the US? How will capital market activity surrounding this integration change? Should long-term infrastructure projects (pipeline approval processes with environmental considerations) reflect an increasing Western focus for potential Asian trade more so than a Southern US focus? Should non-trade policies, such as immigration policy factor in the increased trade that follows where culturally attuned communities develop? A central issue concerns whether Asian growth will persist, and there are conflicting views and opinions. Some such as Lin3 see Chinese growth, in particular, as now firmly rooted in a 20-year experience that will persist for many decades. But others see Chinese growth as more fragile, with growing inequality, deepening environmental problems, demographic constraints of an aging population, and problems of financial structure. Growth prospects elsewhere in Asia, in India and in the ASEAN region, encounter similar divisions of opinion. These are broad-ranging issues for the policy framework in Canada with implications beyond trade and economic integration. I focus more heavily on the trade policy dimensions as reflects my own area of relative expertise, but the broader questions facing Canadian policy remain. First, I briefly review the historical pattern of Canadian development and the extent of Canadian integration with the US that has evolved in recent years. Next, I discuss Asian growth, in China and India specifically, and discuss alternative views on its longer-term sustainability.
John Whalley 103
Finally, I turn to the seeming implications for trade policy negotiations and capital market integration. I will conclude with a discussion concerning possible broader policy implications.
4.2 Phases of Canadian development and the Asian challenge The broad characteristics of Canada’s development since Confederation are well known4 and so I will only briefly summarize here and provide limited documentation, focusing more heavily on the potential departures from these processes which continued high growth in Asia may portend. As summarized in Whalley,5 trade concerns regarding access to natural resource exports led upper and lower Canada into negotiation with the US and the subsequent forging of the Elgin-Marcy Treaty (1854), providing the two provinces duty-free access to a range of resource products. In 1866, under the terms of the treaty, the US gave notice of termination and the treaty lapsed. This was one catalyst underpinning Confederation (1867) and the search for more secure markets spurred by nation-building. This was in part to be met through a secure internal market pending negotiation of trade arrangements abroad with enhanced bargaining power of the newly joined provinces. Following Confederation in 1867, there were repeated attempts by the new Dominion to negotiate secure trade arrangements with both the US and the UK. These failed, and in 1879 Prime Minister Sir John A. MacDonald, committed the then government to the National Policy. This was to be a strategy for Canada’s development behind a national tariff which would emphasize East–West linkage, Western development, transportation subsidies, and the development of a national transportation system. It was to lead to the development of East–West transportation routes, Western settlement, and the development of East–West trade within Canada rather than the seemingly more natural growth of North–South trade with the US which would have occurred in a barrierfree world. Following the adoption of the National Policy (1879), debate in Canada continued concerning the merits of a freer trade stance towards the US, with overtures in 1896 and still later. This culminated in the negotiation of a free-trade arrangement with the US in 1911 by then Prime Minister Sir Wilfrid Laurier following pressure from Western farmers who sought reciprocity with the US. The agreement was ratified by the US Congress, but upon return to Canada a vote was
104
The Asian Economic Revolution and Canadian Trade Policy
forced in the Parliament and the government was to fall. The resulting free-trade election won by the Conservative party based on opposition to US free trade then cemented the policy stance in Canada towards internal development behind a national tariff. This was to intensify in the 1930s with the adoption of Commonwealth preferences, and effectively a three-level tariff with the highest tariff directed against the US. The change from and effective reversal of this policy stance began in the mid-1930s through bilateral negotiation with the US which effectively lowered trade barriers on North–South trade. This was born of the experience of the global depression (1929–31) and the attempt by the then US President Franklin D. Roosevelt to revive the global economy, in part, by reciprocal trade negotiations under the Reciprocal Trade Agreements Act (1934). This authorized the president to lower US trade barriers bilaterally if such reductions were reciprocated by foreign trade partners. Between 1934 and 1938, three negotiations occurred with Canada which bilaterally lowered the US Smoot–Hawley tariff of 1929 and provided significant reductions in the trade barriers faced in the US markets. Canada granted the US with the most favoured nation (MFN) status with some limited preferences. These events preceded reductions in bilateral barriers which were to follow in the post-war years both multilaterally under the GATT and bilaterally. After World War II, Canada had a trade structure which reflected a heavy reliance on natural resource exports, including wheat. Roughly 30% of exports were to the US, however, 70% of imports, largely in the form of manufactures came from the US. Canada’s exports were largely to Europe, and following the period of development under the National Policy, the ratio of trade to GDP was much smaller than it is currently (around 20% as against 45% today). Further trade reductions began multilaterally under the GATT concluded in 1947. The early four GATT rounds before the entry of Japan in 1955 and the formation of the European Economic Community under the Treaty of Rome in 1957 (now the European Union (EU)) limited lowering barriers on both sides as the primary focus was on bindings of tariffs. But trade barriers under the GATT later declined as a result of the later Dillon (1957–9), Kennedy (1963–4), and Tokyo Rounds (1973–9) largely focused on manufacturers, and Canada’s imports from the US were stimulated as a result. These negotiations reflected US–EU (and later Japan) reciprocity in which Canada participated under MFN.
John Whalley 105
It was during these years, however, that both the size and composition of Canada’s trade with the US began to change fuelled by the Canada–US Auto Pact (1965). This arrangement effectively removed duties on both finished automobiles and parts on a bilateral basis and was to lead to a sharp growth in two-way trade centred on Ontario, and to a lesser degree Quebec. In the years that followed, Canada was to move to a position that accounted for nearly 30% of production activity in automobiles and parts in North America (more than Canada’s population share in North America), and bilateral Canada–US trade grew significantly. The reversal in policy stance reflected in the National Policy (1891) was striking. This process was then fuelled by the signing of the Canada–US Free Trade Agreement (1987) containing many more provisions, including special dispute settlement over the use of antidumping and countervailing duties, provisions in such sectors as energy, wine, financial services, and a range of instrument agreements covering tariffs, safeguards, and other matters. The key 1965 arrangements in automobiles and parts were effectively grandfathered into the 1987 agreement. The later trilateralization of this agreement into the NAFTA (1994) continued this process of external engagement for Canada’s economy through added preferential access to an expanding Mexican market. A large segment of this trade, however, remained in automobiles and parts (roughly 35% of trade). Canada’s international engagement was then furthered in the GATT/WTO process with the completion of the Uruguay Round in 1994. By the mid-1990s, in contrast to the National Policy years, Canada’s trade amounted to 45% of GDP. Approximately, 85% of exports were to the US. Of these, around 35% were in automobiles and parts, and of Canada’s manufactured exports, 90% originated in Ontario. Relative to the years of the National Policy, Canada’s economy had become deeply integrated with the US with substantial two-way trade encompassed in a cross-border integrated major manufacturing industry, with North– South trade growing rapidly relative to East–West trade within Canada. This picture of continuing ever-deepening integration by the Canadian economy into a larger North American economy has been the presumption behind a range of policy-related activities beginning in the mid-1980s and the MacDonald Commission. Ever-rising trade shares with the US, trade negotiations on access dominated by the seemingly overwhelming interest in the market of Canada’s near exclusive trading partner and perceived further growth of integrated North–South trade.
106
The Asian Economic Revolution and Canadian Trade Policy
It is the prospect of sustained and rapid growth in developing Asia that begins to challenge this presumption. While trade with Asia remains relatively small, its rapid growth poses several challenges for Canadian policy makers. One concerns a trade composition change which reverses the current direction and suggests a move towards resources as in the pre-Auto Pact years as Asia’s demand for resource exports grows, suggesting major implications for Central Canada. Another concerns growing deficits in Asian trade and surpluses in North American trade as imports of manufactures from Asia displace those from the US. Yet another is renewed emphasis on East–West infrastructure involving pipelines, ports, and other elements as lumber and mineral deposits in the West are shipped to West Coast ports. Further, is Canada’s trade policy negotiating stance given the recent rise of regional agreements in Asia assuming a different form than arrangements under the GATT/WTO and the Canada–US Free Trade Agreement? Issues begin to rise as to whether other areas of policy will be influenced by these developments. One area specifically concerns Canada’s immigration policy if, as is often argued, trade growth is facilitated by domestic communities with cultural and other ties to particular trading partners, a re-examination of current policy may be warranted.
4.3 Growth in Asia and its sustainability The growth which has occurred in Asia since the second World War, first in Japan, subsequently in Korea and Taiwan, and now in China, India and also the ASEAN is one of the striking features of global economic performance in recent decades. The combined size of China and India (40% of the global population) and their cumulative growth rate of approximately 8–9% suggests that if these current rates are sustained in the coming decades, major global change will ensue. Trade patterns and investment flows will be affected and with it the global economy. For Canadian policy makers, the challenge will be whether the objective of trade policy is to benefit from integration into this changing global environment and how it is to be achieved. With continued growth in Asia, there is the prospect of reduced North American integration. Central to this discussion is the speed and sustainability of growth in Asia. Table 4.1 sets out growth rates of the larger key developing Asian countries experiencing rapid growth in recent years as contrasted with Organization for Economic Co-operation and Development (OECD) growth for the period 1985–2004.
John Whalley 107 Table 4.1 Annual growth rates in China, India, and ASEAN(5) compared to OECD*
1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Average 1985–2004
China
India
13.50 8.80 11.60 11.30 4.10 3.80 9.20 14.20 13.50 12.60 10.50 9.60 8.80 7.80 7.10 8.00 7.50 8.30 9.30 9.50 9.45
5.63 4.84 4.27 9.86 6.44 5.81 0.91 5.27 4.87 7.46 7.65 7.39 4.48 5.99 7.13 3.94 5.15 4.09 8.61 6.91 5.83
ASEAN(5)† 0.23 4.60 6.15 8.58 9.14 8.16 6.88 6.26 7.05 7.70 8.10 7.17 3.77 −8.79 3.13 5.85 2.51 4.27 5.28 5.98 5.10
OECD 3.83 3.12 3.35 4.74 3.88 3.00 1.36 2.04 1.21 3.12 2.59 3.02 3.34 2.47 3.11 3.55 1.18 1.38 1.99 3.23 2.77
* Source: A. Antkiewicz and J. Whalley, ‘The Sustainability of Growth in the BRICSAM countries’ (mimeo), 2006, p. 9. Based on World Development Indicators database and authors’ own calculations. † Calculated as a weighted average using PPP–GDP share as used in the IMF–WEO database.
China has experienced the most rapid growth, driven first by agricultural reforms and the introduction of the responsibility system in 1976, and subsequently by growth in output and exports of manufactures. Currently, China has nearly 60% of GDP originating in manufacturing and export growth which in the last few years has been in the 35% annual range. There has been active debate on the reliability of the GDP data that underlie these growth rates6 noting the inconsistency between high measured growth and reduced energy consumption in the late 1990s and early 2000s. A central element of this growth experience has seemingly been inward flows of foreign direct investment (FDI) beginning in early 1992 and driven by policy reforms and tax incentives. China currently accounts for a significant portion of FDI flows from OECD to non-OECD countries, running annually at approximately $60 billion.
108
The Asian Economic Revolution and Canadian Trade Policy
Foreign invested enterprises, who utilize this FDI, account for over 50% of manufacturing activity and nearly 60% of exports.7 India’s growth shows acceleration following policy reforms8 in the late 1980s and thereafter (with the exception of 1991) shows high growth, but at rates below those of China. This growth profile has accelerated in the last two years, prompting some to link India and China as comparable large Asian economies, whose continued growth will transform the global economy. In fact, the Indian experience differs from that of China. The initial base in terms of GDP per capital is lower (perhaps half), and there have been much smaller inflows of FDI (around $5 billion in 2003/2004). Trade growth has also been lower (around 10% per year), and the share of manufacturing in GDP has changed relatively little. There has been growth of output from the service sector, and a significant productivity surge.9 The ASEAN growth experiences differ among the individual ASEAN countries, but the mean growth rate of ASEAN(5) (Thailand, Malaysia, the Philippines, Indonesia, Singapore) is also high as compared to OECD growth rates. The impact of the Asian financial crisis of 1998 are most pronounced among this group of countries. Here again the sources of growth seem to differ from India and China. A claim often made is that Japanese outsourcing to countries such as Indonesia and Malaysia in the late 1970s was a significant trigger. The combined growth rate of these countries, led by China, substantially exceeds that of the OECD, as can be seen from Table 4.1. This has led to speculation10 that by the year 2050, India and China could exceed both the US and the EU in GDP terms. This scenario suggests a changed global economic structure for Canada, signalling increasing trade with the Asian economies and declining relative trade with North America, trade that is more heavily conducted through West Coast ports, and trade that is more heavily concentrated on resource products. A central element in evaluating the realism of such scenarios is the sustainability of these growth rates over extended periods of time in the Asian economies, and there are substantial differences of view on this front. China’s high growth has now been sustained for over 20 years, despite claims from Western economists of its potential fragility. The central concerns have focused on China’s financial structure, where high ratios of non-performing to performing loans in the banking system have prompted suggestions of an interruption of growth. Longer term environmental constraints, and the demographic implications of the one-child-per-family policy have also been raised.
John Whalley 109
But there are other issues. One concerns growing inequality in China, with the ratio of urban to rural average incomes increasing from around 1.8 in the early 1980s to 3.4 today. The concern is that political pressures for more redistributive policies may lower future growth. Another is the potential plateauing of foreign direct investment. The sharp growth in FDI since the early 1990’s is often taken to be significant contributor to growth in both GDP and exports, and this could also constrain growth. Yet another is the absorptive capacity of the OECD economies for increased Chinese exports. China’s share of world trade is doubling approximately every three years, and so eventual constraints may prevent the absorptive capacity of the global economy. Those who view China as embarking on a historical transformation, related to but distinct from the description of the industrial revolution in Western Europe,11 stress the longer-term transformation in social, as well as economic structure involved and portray long-term growth at high rates for several decades.12 In the case of India, the potential constraints on growth differ. In the short term, a central factor concerns infrastructure and power. Other constraints relative to the case of China differ. FDI inflows have been relatively small, and trade growth has been more modest and so external factors in future growth performance seem less significant. Inequality has not increased until very recently, and non-performing loans in the banking system are not a major concern. Increases in manufacturing output are only modest, and urban unemployment has fallen little. The increase in output from the services sectors seems to reflect productivity growth in these sectors as much as employment growth. The drivers of growth, however, appear different and this has caused some to be more cautious of future growth. Opinions vary and there are those who portray India as lying on a golden turnpike with growth building on 50 years of democratic process,13 to those who question future growth.14 The growth rate of ASEAN countries is lower than that of India and China, as a higher base of GDP/capita results from a smaller population size. ASEAN growth was subject to major interruption during the financial crisis of 1998 and the consequences of that episode still persist in Indonesia, even though recovery in the remainder of ASEAN countries has been more swift. Financial turbulence and contagion as shorterterm interrupters of ASEAN growth are therefore a factor. Capital inflows have been more modest and growth rates of exports have been considerably below those of China.
110
The Asian Economic Revolution and Canadian Trade Policy
The growth performance in Asia, however, despite concerns over medium- to longer-term sustainability raises the prospect of major global economic transformation. These economies’ growing markets, accelerated intra-regional trade,15 are the sources of significant potential capital outflows, all of which present opportunities and pose issues as well for Canadian trade and other OECD policy makers as they contemplate the coming decades.
4.4 Canada’s trade and global integration policy – challenges of Asian growth Canada’s trade between 2001 and 2005 with developing Asia is shown in Table 4.2. The data indicates only current small export penetration by Canada of developing Asian Markets, but much more substantial and rapidly growing penetration of Canadian markets by Asian suppliers. Asian exports are predominantly of manufactures (and mainly machinery). Canadian exports are heavily wood pulp, ores, fertilizer, and organic chemicals. The highest growth rates occur on the import side, and are for imports from China. China’s export growth rate globally in recent years is in the 35% range, and so the prospect may be for even more rapid growth in the coming years, but the implication would seem to be one of growing trade deficits with Asia, and growing surpluses with the US. The imports entering from Asia will likely displace both US imports of manufactures and domestic production in Canada generating potentially significant adjustment consequences for Canadian manufacturing. As previously mentioned, there are a wide range of trade policy issues raised by these developments beyond the central developmental issues associated with a North American or global integration focus. Examples include policy in key product areas such as textiles and apparel, based on the termination of the Multi-Fiber Agreement (MFA) and the bargaining of remaining tariff-based restrictions for access improvements; and the question of how to proceed with a growing use of dumping actions in Asia (and worldwide). With slower export growth, one central issue will likely be how Canada can manoeuvre to improve access to these Asian markets. In part, some reliance can be placed on WTO disciplines through lowered and bound tariffs in these markets for MFN suppliers, and the package of disciplines agreed to by China in 2001 under their WTO accession terms. But the pace of regional negotiation in Asia has been accelerating in recent years and the resulting agreements are quite different in both form and
Table 4.2
Exports US China ASEAN India World Imports US China ASEAN India World
Canada’s imports and exports with selected countries from 2001 to 2005 2001
2002
2003
2004
2005
Share in 2001 (%)
Share in 2005 (%)
Percentage change 2001–5
3,51,751 4264 2058 676 4,04,085
3,45,366 4132 2475 675 3,96,381
3,26,700 4798 2309 762 3,81,000
3,48,142 6655 2784 860 4,11,840
3,65,741 7060 2793 1082 4,35,834
87.0 1.1 0.5 0.2 100.0
83.9 1.6 0.6 0.2 100
4.0 65.6 35.7 60.1 7.9
2,18,290 12,724 7000 1155 3,43,111
2,18,497 16,004 7222 1327 3,48,957
2,03,803 18,581 7563 1423 3,36,104
2,08,971 24,100 8353 1577 3,56,056
2,15,109 29,498 8154 1787 3,80,691
63.6 3.7 2.0 0.3 100.0
56.5 7.7 2.1 0.5 100
−1.5 131.8 16.5 54.7 11.0 John Whalley 111
Unit: millions Canadian dollars. Source: Industry Canada, Trade Data Online, http://www.strategis.gc.ca.
112
The Asian Economic Revolution and Canadian Trade Policy
focus from more conventional WTO agreements and the Regional Trade Agreements which were largely mutual instrument limitation based. Evaluating both the impact and the potential future significance of recent regional agreements requires synthesizing these arrangements, as many go well beyond the tariff barrel agreements studied by economic theorists. By way of example, the 2002 US–Singapore bilateral agreement is 210 pages of text and 1375 pages of annexes with a 299-page annex on Rules of Origin.16 One way to approach these agreements is as tariff plus agreements, a recognition that in many cases the plus component dominates the tariff part in length of text and likely in significance given that in many cases the MFN tariff rates are sufficiently low enough so that the margins of preference involved have limited impacts on trade. But a further key element in assessing these agreements is to recognize their significance as process rather than simple instrumentbased agreements that limit the use of trade-based interventions (tariffs).17 Several of the recent regional agreements, and especially those of China, are centrally labelled as cooperation and partnership agreements. Commitments are set out to cooperate in tourism, rural development, standards setting, and in other areas including poverty alleviation and the promotion of mutual understanding. Tariff arrangements are but one part of these agreements. Some agreements contain broad commitments to cut tariffs with details to be substantially negotiated. In some cases, agreements on Rules of Origin, and even bilateral dispute settlement are to follow the concluded broad agreements.18 And the move into newer areas not covered by current WTO disciplines (or current or prospective negotiations) such as standards setting, competition issues, mutual recognition and other areas is reflected in the language used to label agreements. The recent China– ASEAN agreement is a Framework Agreement on Comprehensive Economic Cooperation. Having noted this tendency for regional agreements to evolve as formalized management arrangements for bilateral economic interaction, more so than as treaty-based agreements to mutually limit the use of border measures, a further striking feature is the diversity among the agreements themselves. In part, this is reflected in the sharply differing focus of agreements across the partners involved. Many are agreements between countries of sharply asymmetric size and follow a seemingly standardized pattern common to other agreements involving a large trading partner. China’s recent agreements19 appear customized to
John Whalley 113
individual partners and reflect perceived Chinese opportunities in each bilateral negotiation. I now discuss only a subset of recent agreements involving both ASEAN as a single entity, and individual ASEAN countries negotiating separately (Malaysia, Thailand, and Singapore) due to the sheer number and complexity of agreements.20 These go substantially beyond conventional free trade agreements in dealing with items not yet subject to WTO disciplines. They have instead become platforms for packaging a range of new and ongoing issues previously dealt with separately (such as visas/work permits) into a combined negotiation. These additional issues added to trade negotiation differ both between multilateral and regional negotiations, and across individual regional negotiations. To date, ASEAN has concluded three formal framework agreements (with China, Japan, and India) and aims to activate more substantial arrangements and has a further two (with Korea, and Australia and New Zealand jointly) under negotiation. Singapore has concluded six substantive agreements (with the US, the European Free Trade Association (EFTA), Australia, New Zealand, Japan, and Jordan) and is negotiating a further 12. Thailand has concluded two full agreements (with Australia, Bahrain), two framework agreements (with the US, India), one ancillary agreement to a wider ASEAN agreement only covering vegetables and fruits (with China), and is negotiating a further five agreements. Malaysia has a single bilateral investment treaty with the US, but is negotiating a further five agreements. These agreements vary greatly in length, specificity, and coverage; some are detailed with substantial specificity (specifically the Singapore–US agreement). What is striking about these agreements is their breadth of coverage. Several areas beyond current WTO disciplines are dealt with including competition policy, mutual recognition (both of professional qualifications and product standards and testing), movement of persons and visa/work permit arrangements, investment, and cooperation in specific areas. Of all the ASEAN regional agreements, six contain provisions relating to competition policy: Singapore–US, Singapore–EFTA, Singapore–New Zealand, Singapore–Australia, Singapore–Japan, and Thailand–Australia. The two ASEAN-wide agreements (with China and India) have no coverage of competition policy. Singapore at the time of negotiation with the US had no formal competition law, but is now in the process of enacting such laws.
114
The Asian Economic Revolution and Canadian Trade Policy
Investment is dealt with in more of the country agreements than is competition policy and is also covered in ASEAN agreements. There is also more commonality in approach, with central commitments being (either or both) National Treatment and MFN treatment for foreign investors (typically) alongside provisions relating to expropriation compensation, and (in some cases) repatriation of earnings. Issues related to mutual recognition in these agreements arise under a number of chapter headings in the various texts. Issues of product testing and standards are in chapters on technical barriers to trade and sanitary and phytosanitary restrictions, and particular sectoral chapters, such as telecommunications, touch on product standards while chapters on services deal with recognition of professional certification. The ASEAN bloc-wide and country agreements also deal with issues related to movement of persons and usually in separate chapters. This is reflective both of the growing significance of visa and work permit issues in the global economy and the absence of multilateral venues for addressing them.21 In this case, country or regional agreements provide the platform for raising an issue under consideration, but provide no clear multilateral forum for discussion. Bargaining of diverse issues and with it, the greater probability of achieving results presumably provides the rationale for this. The ASEAN agreements also contain a range of commitments relating to cooperation in a number of areas. Cooperation agreements are typically vague and appear tempting to dismiss as lacking in substance, however, they do represent a commitment to a deepening of bilateral relationships in specific and designated areas and hence are of substance from a process point of view. The stress on cooperation in the China–ASEAN agreement is manifest in the title of the agreement as a ‘Framework Agreement on Comprehensive Economic Cooperation’ indicating the significance attached to cooperation in bilateral relationship building. There are commitments to strengthen cooperation in five key sectors (agriculture, information and communication technology, human resource development, investment, and Mekong River basin development). Though inevitably vague and difficult to interpret as legal text, these cooperation elements play a role in the deepening of countryto-country relationships in these agreements. For countries where sequential relationship building and deepening is seen as a critical way to proceed in international negotiations, the cooperation provisions of these agreements can play a major role as far as international economic management is concerned. These new partnership agreements unlike
John Whalley 115
traditional tariff-based agreements are not typically negotiated as one-off treaty arrangements by countries with an exclusive focus on legal provisions and detailed text. Canada thus far has not pursued active negotiations of this form with the developing Asian economies. In part, the presumption of continued reliance on North American trade and commitment to WTO process may be one reason, although Canada has been actively involved in regional negotiation elsewhere. Since these negotiations tend to be broader than trade negotiations in coverage, they are also of symbolic importance in establishing linkages to policy structures and processes in Asian countries. Another recent issue which has received much attention relates to the policy response elicited by a recent wave of both actual and proposed buyouts by Chinese companies of entities outside China.22 While the majority of these cases have not resulted in completed transactions, they pose the issue of whether these activities might be a catalyst for a wider discussion of the implications for global arrangements covering cross-border acquisitions and how Canada might respond. For some years, Chinese outward FDI has been relatively small (in the US$3–4 billion range in 2004) and heavily concentrated on both greenfield and joint venture activity, much of it occurring in Hong Kong. In the last two years, a change which has occurred is both a focus on direct acquisition and the emergence of potentially large transactions, some in the US$15–20 billion range, and with it a focus which goes well beyond Hong Kong. Examples include the Lenovo buyout of IBM’s PC business, CNOOC’s (China National Offshore Oil Company) bid for Unocal, prospective bids by MinMetals for Noranda, the Haier Group bid for Maytag, and others. No direct WTO issues are raised by these, but questions of subsidization, lack of transparency, and national security have all been raised. National security issues regarding foreign acquisitions are not new and extend to Exon-Florio provisions of the American Defence Production Act of 1998 following concerns in the US in the late 1980s over Japanese buyouts. However, issues of subsidization of foreign acquisitions through low interest loans from central banks and the transparency of organizational form of acquiring entities (State-Owned Enterprises (SOEs)) are new. The overarching feature is the apparent absence of globally agreed disciplines covering not only cross-border acquisitions, but more broadly all cross-border factor flows. This situation stands in contrast to the goods and services flows covered by the WTO. Both the failed multilateral agreement on investment (MAI) and WTO competition
116
The Asian Economic Revolution and Canadian Trade Policy
policy negotiations did not touch directly on the newer issues in recent debate on Chinese buyouts, and neither do earlier bilateral trade and investment treaties. A number of factors underlie this recent upsurge in Chinese cross-border acquisition activity. One is large accumulated Central Bank reserves in China (now close to US$800 billion), and a seeming change in policy stance by the Central Bank of advancing low interest loans to SOEs for foreign acquisitions rather than continued accumulation of US treasury notes. Chinese concerns over security for supply of resource inputs (especially oil) for Chinese manufacturing enterprises also motivate the change in activity. And for private manufacturing groups in China, the use of foreign acquisitions as a way of obtaining distribution networks in the OECD seems to be a factor. The picture is one of macro imbalances combined with pragmatic niche-driven foreign acquisition activity involving both Chinese SOEs and private groups, with differing resource and manufacturing acquisition structures. Thus far, central banks around the world have not engaged themselves in extending low interest loans for foreign acquisitions, but the policy structure in China with large communally owned production units (by national, provincial, and municipal governments) makes this logical from China’s standpoint. An issue for Canada is whether to link approval of individual transactions to negotiations (or renegotiations) of bilateral investment treaties. A broader approach is to seek a global regulatory framework covering purchases by prospective foreign parent entities, a matter rarely touched on by previous WTO and OECD investment discussions which have largely focused on translating existing system of trade rules (National Treatment and MFN) into investment rules. These are but two of several issues which could be discussed in light of the changing environment and illustrate the desirability of an adaptive response on the trade and economic policy integration front to developments in Asia.
4.5 Concluding remarks and broader policy implications In a changing world of sustained high growth in developing Asia for several decades, there are a wide range of policy challenges posed both for Canada and other OECD countries. The vista is broad, and I have maintained a focus in this chapter on trade and economic integration policies. But in many ways the challenges have broader implications. China’s integration into the global economy has the potential to raise major global adjustment pressures. Analytical economists argue that
John Whalley 117
these adjustments are simply the mirror image of the gains from trade, and there is little doubt of the benefits to consumers of low-priced high quality Chinese imports. But the size of the low-wage labour pool in China and its current limited deployment in trade-related activities suggests that we may only be in the early stages of China’s global integration process. Currently, foreign-invested enterprises in China account for over 50% of China’s exports but employ only 3% of the workforce.23 Cumulative FDI into China may be in the $500 billion range, while the OECD capital stock using a rough capital output ratio of 3:1 may be $75 trillion. The amount of capital that could still be deployed in the medium term with immigration from constrained low-wage labour in China appears large and hence the potential adjustments would seem to be a key concern to policy makers throughout the OECD. And this is prior to any further adjustment pressures from growth in India and ASEAN are considered. The adjustments involved relate not only to the country composition of trade, with growing imports of manufactures from China displacing higher priced US imports, but also to domestic production. Since the majority of manufacturing activity in Canada is located in Ontario, the change implied in regional balance could be substantial. Central Canadian production of automobiles and parts will be further impacted by Asian auto production displacing North American production sold in US markets. The adjustments continue with increased exports of resource-based products from Western suppliers, refocused infrastructure towards shipments from West Coast ports, and additional policy areas. One facet is immigration policy as trade research24 tends to indicate more proclivity for trade to occur between countries where similar cultural communities reside. Canada’s trade strategy for increasing exports to growing Asian markets could conceivably affect changes in immigration policy. Seemingly, these prospective changes have sufficient plausibility to be factored into trade and other policy making in Canada over the next few years. How best to approach and respond to these changes remains the challenge.
Notes I am grateful to Josh Svatek for his assistance. 1. D. Wilson and R. Purushothaman, Dreaming with BRICs. The Path to 2050. Goldman Sachs, Global Economic Paper no. 99, October 2003. Online: http: www.gs.com.
118
The Asian Economic Revolution and Canadian Trade Policy
2. Minister of Public Works and Government Services Canada. ‘Seventh Annual Report on Canada’s State of Trade. Trade Update’. Foreign Affairs and International Trade Canada, June 2006. June 2006. Available online: p. 24. 3. J. Y. Lin, ‘Is China’s Growth Real and Sustainable?’, in Asian Perspective, vol. 28, no. 3, 2004. 4. C. Hamilton, R. Hill, and J. Whalley, Canadian Trade Policies and the World Economy. Research Studies, vol. 9, Royal Commission on the Economic Union and Development Prospects for Canada, Toronto, ON: University of Toronto Press, 1985, xv, p. 147. 5. R. Hill and J. Whalley, ‘Canada–US Free Trade: An Introduction’ in J. Whalley with R. Hill, Canada–United States Free Trade, vol. 11 of the Research Studies, Royal Commission on the Economic Union and Development Prospects for Canada, Toronto, ON: University of Toronto Press, 1985, pp. 1–42. 6. See T. Rawski, ‘Measuring China’s Recent GDP Growth: Where Do We Stand’, in China Economic Quarterly, vol. 2, no. 1, 2002. 7. See J. Whalley and X. Xin, ‘China’s FDI and Non-FDI Economies and the Sustainability of Future High Chinese Growth’, NBER Working Paper No. 12249, National Bureau of Economic Research, 2006. 8. A. Panagariya, ‘India in the 1980s and 1990s: A Triumph of Reform’, IMF Working Paper 04/43, International Monetary Fund, 2004 and also D. Rodrik, A. Subramanian, and B. DeLong, ‘India Since Independence: An Analytical Growth Narrative’ in D. Rodrick (ed.), In Search of Prosperity: Analytic Narratives on Economic Growth. Princeton, NJ: Princeton University Press, 2003. 9. See D. Rodrik and A. Subramanian, ‘From “Hindu Growth” to Productivity Surge: The Mystery of the Indian Growth Transition’, NBER Working Paper No. 10376, National Bureau of Economic Research, March 2004. 10. Wilson and Purushothaman (2003). 11. See Karl Polanyi, The Great Transformation: The Political and Economic Origins of Our Time. Boston: Beacon Press, 1944. 12. See J. Y. Lin, Fang Cai, and Zhou Li, China’s Miracle: Development Strategy and Economic Reform. Hong Kong: Chinese University Press, 2002. 13. V. L. Kelkar, ‘India: On the Growth Turnpike’, paper presented at the Narayanan Lecture at Australian National University, Canberra, 27 April 2004. 14. K. S. Parikh, ‘Indian economy: A Shining Star or a Passing Comet?’ in Nagesh Kumar, Rahul Sen and Mukul Asher (eds), India–ASEAN Economic Relations: Meeting the Challenges of Globalization. Singapore: Research and Information Systems for Developing Countries (RIS) and Institute of Southeast Asian Studies (ISEAS), 2006. 15. China, for instance, is now India’s second largest export market while five years ago that share was small. 16. Office of the United States Trade Representative (USTR) online: http://www. ustr.gov/Trade_Agreements/Bilateral/Singapore_FTA/... Final_Texts/Section_ Index.html. 17. J. Whalley, ‘Recent Regional Trade Agreements: Why so many, why so much variance in form, why so fast, and where are they headed?’, Paper prepared for CESifo Summer Institute, Venice, 2006.
John Whalley 119 18. ASEAN website: http://www.aseansec.org/4979.htm. 19. See J. Whalley and A. Antkiewicz, ‘A BRICSAM Strategy for Canada?’ in A. F. Cooper and D. Rowlands, Split Images: Canada Among Nations 2005, Montreal: McGill-Queens University Press, 2005. 20. This discussion draws in part on O. G. Dayaratna Banda and J. Whalley, ‘Beyond Goods and Services: Competition Policy, Investment, Mutual Recognition, Movement of Persons, and Broader Cooperation Provisions of Recent FTAs involving ASEAN Countries’. NBER Working Papers: 11232, National Bureau of Economic Research, 2005. 21. See E. Ng and J. Whalley, ‘Visas and Work Permits: Can GATS/WTO Help or is a New Global Entity Needed?’ 2004 (mimeo) and also J. Nielson, ‘Current Regimes for Temporary Movement of Service Providers: Labour Mobility in Regional Trade Agreements’, Joint WTO–World Bank Symposium on Movement of Natural Persons (Mode 4) under the GATS, 11–2 April 2002 for a discussion of temporary movement of persons in other recent bilateral agreements. 22. A. Antkiewicz and J. Whalley, ‘Recent. Chinese Buyout Activity and the Implications for Global Architecture’. NBER Working Papers: 12072, National Bureau of Economic Research, 2006. 23. See J. Whalley and X. Xin, ‘China’s FDI and Non-FDI Economies and the Sustainability of Future High Chinese Growth’, NBER Working Paper No. 1224, National Bureau of Economic Research, 2006. 24. See J. E. Rauch, ‘Reconciling the Pattern of Trade with the Pattern of Migration’, The American Economic Review, vol. 81, no. 4. (September 1991), pp. 775–96.
References Antkiewicz, A and J. Whalley (2006). ‘Recent. Chinese Buyout Activity and the Implications for Global Architecture’. NBER Working Papers: 12072, National Bureau of Economic Research. —— (2006). ‘The Sustainability of Growth in the BRICSAM countries’ (mimeo). —— (2005). ‘China’s Regional Trade Agreements’, World Economy. October, vol. 28, no. 10. Dayaratna Banda, O. G. and J. Whalley (2005). ‘Beyond Goods and Services: Competition Policy, Investment, Mutual Recognition, Movement of Persons, and Broader Cooperation Provisions of Recent FTAs involving ASEAN Countries’. NBER Working Papers: 11232, National Bureau of Economic Research. Hamilton, C., R. Hill and J. Whalley (1985). Canadian Trade Policies and the World Economy. Research Studies, vol. 9, Royal Commission on the Economic Union and Development Prospects for Canada, Toronto, ON: University of Toronto Press. Hill, R. and J. Whalley (1985). ‘Canada–US Free Trade: An Introduction’ in J. Whalley with R. Hill, Canada–United States Free Trade, vol. 11 of the Research Studies, Royal Commission on the Economic Union and Development Prospects for Canada, Toronto, ON: University of Toronto Press. Kelkar, V. L. (2004). ‘India: On the Growth Turnpike’, paper presented at the 2004 Narayanan Lecture at Australian National University, Canberra, April 27.
120
The Asian Economic Revolution and Canadian Trade Policy
Lin, J. Y. (2004). ‘Is China’s Growth Real and Sustainable?’, Asian Perspective, vol. 28, no. 3. Lin, J. Y., Fang Cai, and Zhou Li (2002). China’s Miracle: Development Strategy and Economic Reform, Hong Kong, Chinese University Press. Ng, E. and J. Whalley (2004). ‘Visas and Work Permits: Can GATS/WTO Help or is a New Global Entity Needed?’ (mimeo). Nielson, J. (2002). ‘Current Regimes for Temporary Movement of Service Providers: Labour Mobility in Regional Trade Agreements’, Joint WTO–World Bank Symposium on Movement of Natural Persons (Mode 4) under the GATS, 11–12 April. Panagariya, A. (2004). ‘India in the 1980s and 1990s: A Triumph of Reform’, IMF Working Paper 04/43, International Monetary Fund. Parikh, K. S. (2006). ‘Indian economy: A Shining Star or a Passing Comet?’ in Kumar, Nagesh; Rahul Sen and Mukul G. Asher (eds), India–ASEAN Economic Relations: Meeting the Challenges of Globalization. Singapore: Research and Information Systems for Developing Countries (RIS) and Institute of Southeast Asian Studies (ISEAS). Polanyi, Karl (1944). The Great Transformation: The Political and Economic Origins of Our Time, Boston: Beacon Press. Rawski, T. (2002). ‘Measuring China’s Recent GDP Growth: Where Do We Stand’, China Economic Quarterly, vol. 2, no. 1. Rauch, J. E. (1991). ‘Reconciling the Pattern of Trade with the Pattern of Migration’, The American Economic Review, vol. 81, no. 4 (September): 775–96. Rodrik, D. and A. Subramanian (2004). ‘From “Hindu Growth” to Productivity Surge: The Mystery of the Indian Growth Transition’, NBER Working Paper No. 10376, March, National Bureau of Economic Research. Rodrik D., A. Subramanian, and B. DeLong (2003). ‘India Since Independence: An Analytical Growth Narrative’ in D. Rodrick (ed)., In Search of Prosperity: Analytic Narratives on Economic Growth. Princeton, New Jersey: Princeton University Press. Whalley J. and A. Antkiewicz (2005). ‘A BRICSAM Strategy for Canada?’ in A.F. Cooper and D. Rowlands, Split Images: Canada Among Nations 2005, Montreal: McGill-Queens University Press. Whalley J. (2006). ‘Recent Regional Trade Agreements: Why so many, why so much variance in form, why so fast, and where are they headed?’, Paper prepared for CESifo summer institute, Venice, 2006. Whalley J. and X. Xin (2006). ‘China’s FDI and Non-FDI Economies and the Sustainability of Future High Chinese Growth’, NBER Working Paper No. 12249, National Bureau of Economic Research. Wilson, D. and R. Purushothaman (2003). Dreaming with BRICs. The Path to 2050. Goldman Sachs, Global Economic Paper no.99, October. http:www. gs.com.
Internet sources Association of Southeast Asian Nations (ASEAN): http://www.aseansec.org/4979. htm. Industry Canada, Trade Data Online: http://www.strategis.gc.ca.
John Whalley 121 Minister of Public Works and Government Services Canada. ‘Seventh Annual Report on Canada’s State of Trade. Trade Update’ Foreign Affairs and International Trade Canada, June 2006: http://www.international.gc.ca/eet/ trade/state-of-trade-en.asp. Office of the United States Trade Representative (USTR): http://www.ustr. gov/Trade_Agreements/Bilateral/Singapore_FTA/... Final_Texts/Section_Index. html.
5 Reforming Canadian Medicare: Can an Icon be Redesigned? Åke Blomqvist
5.1 Introduction In the first decade of the twenty-first century, all of the world’s advanced economies can reasonably be classified as ‘welfare states’, in that all of them have tax and transfer programmes that redistribute income from the rich to the poor to a significant extent, and provide economic safety nets for those who, for one reason or another, are not able to reach acceptable standards of well-being without assistance. Differences across countries in terms of the ambition level of their welfare state programmes continue to exist of course, but they have over time become less pronounced. In the years following World War II, questions with respect to development of welfare state programmes were an important part of the social policy debate in Canada, as it tried to steer a middle course between the conservative and the selective approach taken in the US, and the more comprehensive and universal programmes being developed in Europe generally and in the UK particularly. The nature of public involvement in the health care system is one of the clearest examples of the kind of compromise approach that was taken. As in the UK, Canada put in place a system of publicly funded health care that covered everybody under a single comprehensive plan with little or no patient cost sharing. (In the US, by comparison contrast, government involvement was mostly through regulation of private insurance, and selective publicly funded insurance for specific population groups.) However, in contrast to the case in the UK National Health Service (NHS), in Canada patients retained an unrestricted right to choose which doctor and hospital to go to (as they did in the US), and doctors, even hospital-based ones, were allowed to practice privately wherever they chose, with payment on the 122
Åke Blomqvist 123 basis of fee for service. To this day, some of the problems we face with respect to management of the health system arise because of features of this compromise structure that was chosen so many years ago.1 David Laidler, of course, has a deep understanding of the differences between North American and European approaches to the design and management of the welfare state, both because he has lived for a long time in both Canada and the UK, and because of his continuing interest in the relation between economics and individual welfare. For me, he was an ideal colleague off whom to bounce ideas when I first became interested in health economics and comparative health policy. Indeed, he helped arrange my first opportunity to publish a piece in this area some 25 years ago, and I have continued to enjoy our conversations about health policy ever since.2 In the monograph I published in 1979, I discussed some of the design flaws of the Canadian health care system and a number of changes (modelled in part on international experience) that could be undertaken to make it more efficient, without sacrificing what I take to be its fundamental equity-related cornerstone.3 There obviously have been large and dramatic changes in the system since that time. However, the design flaws I identified then have not been addressed; if anything, the problems they are causing have become more severe in today’s environment in which there is a greatly expanded range of medical and pharmaceutical options, and a rapidly aging population. The purpose of the present paper is twofold. The first objective is to summarize what I believe to be the lessons that international experience offers for the task of designing a system of health services funding and provision that overcomes the various types of potential market failure that would exist in an unregulated private one, and to consider how these lessons can be used to make the Canadian system more efficient. In this part of the paper, I implicitly assume that the basic framework will remain as it is, that is, it will continue to consist more or less of a uniform set of monopoly government plans in each province. Second, however, I have over time become less optimistic regarding our ability to implement the types of reform that I believe would be beneficial. Given the way health policy is discussed and managed in our federal system, recent history seems to suggest that the most we can accomplish at present amounts to little more than minor administrative tinkering. For this reason, I now believe the time has come to give serious consideration to two major policy steps that have not been extensively debated in the past. The first is to allow individuals who so desire to opt out of the provincial health insurance plans and obtain coverage through private
124
Reforming Canadian Medicare
insurance instead. The second is to redefine the Canada Health Act and give full responsibility for health policy and financing to the provinces. I briefly discuss these issues towards the end of the paper. The rest of the material is organized as follows. In section 5.2, I review the principal types of market failure that are typically cited as reasons for extensive public involvement in health care in all countries, and summarize the way the Canadian health care system has been organized to respond to these market failures. I then describe briefly what I see as the most promising reforms that might make the existing system more efficient; section 5.3 discusses changes relating to the way health services are produced, while section 5.4 deals with the methods and division of responsibility for funding the provincial plans as well as the nature of the coverage (user fees, scope of coverage) they provide. In section 5.5, I then consider the arguments for allowing individuals to opt out of the provincial plans and for an expanded role for private health insurance. Section 5.6 offers brief concluding comments and suggests that a revamping of the Canada Health Act so as to essentially eliminate the federal role in health policy may be a good first step towards more meaningful reform.
5.2 Market failures in the health care system In an analysis of the scope of market failure in health care, one must take into account that an unregulated private health care system would involve two types of markets: that for health services, and that for health insurance. The reason for this, of course, is that illness shocks strike very randomly. In a given year, the cost of health services rendered to a significant number of individuals will be zero, and for the majority it will be below some small amount like a few hundred dollars. On the other hand, a large proportion of the total health care costs in a given year is accounted for by the services provided to a very small minority with serious illness for whom the cost of treatment can easily exceed their total annual income. A well-functioning health care system, therefore, requires some form of risk pooling, whether through private insurance or public funding, in addition to an efficient system for provision of health services. Indeed, it is probably the fact that the insurance and service provision functions are so interdependent that explains why most discussions of health policy typically refer to the ‘health care system’, rather than to the market for health services or health insurance separately. The problems of market failure that are usually cited as reasons for government intervention relate both to the
Åke Blomqvist 125 insurance and services provision functions of the system, and to the way they interact. 5.2.1 Market failure and information asymmetry By far the most important sources of market failure that would arise in an unregulated market-based health care system would be those stemming from incomplete and asymmetric information among the actors involved in it (consumers/patients, providers of health services, and providers of health insurance). For the sake of completeness, it should be noted that some degree of market failure would also arise as a result of non-priced externalities, namely those associated with treatment and prevention of contagious disease. In advanced countries, however, the portion of total health system costs that relate to contagious disease is not large, and the activities involved are usually handled by separately funded public health departments. The most obvious type of information asymmetry in the market for those health services that are private goods is that between doctors and patients. Doctors know better than the patients how to interpret the latter’s symptoms, what treatment methods and drugs are available, and what the prognosis is under different treatment approaches. Moreover, a large proportion of cases do not involve ‘repeat buying’, so little information has typically been accumulated as a result of the patients’ past experience. Because of the highly personal nature of most transactions involving health services, many patients find the alternative of seeking a second opinion (or price comparisons) awkward. Finally, because most medicine remains such an inexact science, it is difficult to enforce ex post warranties, a solution that is often used in other situations where buyers pay for goods or services that are difficult to evaluate ex ante. While health economists continue to debate precisely how serious these problems are, most of us have little difficulty accepting the proposition that, other things being equal, a market where doctors and hospitals were free to set their own fees (subject only to competition), and to recommend whatever treatment methods they chose, would not produce a very cost-effective pattern of care. That is, most of us probably believe that in such an environment, what has been referred to as ‘supplier-induced demand’ would be a real problem, leading to both high unit prices and an inefficiently high volume of services.4 Another potentially important type of information asymmetry in a hypothetical unregulated health care system is that between insurers, on the one hand, and doctors and patients, on the other. Specifically,
126
Reforming Canadian Medicare
it is the doctor and not the insurer who carries out the diagnosis of the patient’s illness, partly on the basis of information provided by the patient, and it would be difficult and costly for the insurer to independently verify the diagnosis. For this reason, state-contingent insurance contracts that specify a fixed payout based on the patient’s diagnosis are not common in private insurance; instead the typical contract specifies that the plan will pay all or part of the expenditures that actually have been incurred in the course of treatment. As a result, there is the wellknown problem of moral hazard: conventional insurance effectively acts as a subsidy to the utilization of medical care, implying an incentive for patients to utilize more health services than can be justified by their medical benefits. Note that the moral hazard problem exists even if the doctor acts as a perfect agent for the patient; if it is to the doctor’s advantage to exploit his/her information advantage over the patient and provide more services than the diagnosis warrants, the tendency towards over-utilization will be even stronger.5 Finally, there is also some degree of information asymmetry between patients and insurers with respect to factors that affect the patients’ risk of different kinds of serious illness. This is the information asymmetry that gives rise to the problem of adverse selection in health insurance markets, where it may manifest itself in a tendency for generous insurance plans to disappear as they tend to disproportionately attract people who know themselves to be at high risk of illness.6 5.2.2 Health policy and equity Supplier-induced demand, moral hazard, and adverse selection are all factors that may tend to lead to market failure in the sense of reducing the efficiency with which health system resources are likely to be used, so certain forms of government intervention can be justified on the grounds of economic efficiency. However, much of the discussion of health policy in Canada and elsewhere focuses more on the goal of equity, and those objecting to a market-based system do so most vehemently on equity grounds. From the viewpoint of conventional microeconomic theory, the idea that health policy should be conducted so as to promote equity appears to collide with the principle that it is more efficient to do so by redistributing income (general purchasing power) rather than by redistributing particular commodities such as health services (or housing). An especially counter-intuitive version of the equity-based approach to health services provision is that which proposes not only that everyone, regardless of income, should have access to a basic standard of care when ill, but also that those
Åke Blomqvist 127 willing and able to pay for a higher standard should be prevented from doing so. This idea was at one time supported by as eminent an economist as James Tobin (1970). However, he appears to have done so on the basis of an implicit model in which any reduction in risks to health (‘chances of death and disability’, p. 273) lexicographically dominated other commodities, and health care resources were in extremely inelastic supply (so that incremental health services utilized by the rich ‘for trivial purposes’) would imply a corresponding reduction in the supply of services urgently needed by the poor. Neither of these implicit assumptions seems very reasonable today, and for reasons that I will further discuss below, I believe the case for allowing people with high income to spend it on whatever kinds of health care they like, is much stronger now than it may have been in an earlier era. Conversely, I also think that in principle, it is both efficient and consistent with equity to allow people with low income, who so desire to opt for a standard of health insurance that is below that of the plan held by the average person, in return for additional resources for other goods and services, consistent with the notion that transfers in cash are preferred to non-tradeable benefits in kind. In spite of the above, however, there is one sense in which I continue to believe that a tax-financed system that offers every citizen guaranteed access to a basic level of health care may in fact be a second-best efficient element of redistribution policy, in a world with imperfect information. Specifically, in a model of second-best optimal income taxation, if individuals differ not only in their ability to earn income, but also in their probability of being ill, publicly provided health insurance can be shown to increase social welfare if information problems preclude differentiation of the income tax in accordance with the probability of illness.7 Strictly speaking, an information problem of this kind does not in itself imply market failure in a conventional sense, and the principle that transfers in kind give rise to some inefficiency continues to apply. However, in a model of social welfare in which lump-sum subsidies cannot be used to redistribute income towards those with a high risk of illness, provision of a basic package of publicly funded health insurance may increase social welfare by more than enough to compensate for the potential inefficiency associated with the transfer in kind, especially if some form of opting out is allowed (that is, if those who wish to buy a private plan instead are allowed to do so, and are at least partially compensated for giving up their coverage under the public plan). In practice, this idea may implicitly constitute one of the most important arguments in favour of government involvement in health care funding.
128
Reforming Canadian Medicare
5.2.3 International differences in health care systems Taken together, the various factors discussed above add up to a pretty powerful case against a pure laissez-faire approach to the organization of a country’s health care system. As a result, government intervention (through regulation or direct intervention) in health care is pervasive, in essentially every country in the world. While countries are similar in the sense of having the government play a major role in health care, the similarity largely ends there: when one looks at how the government intervenes, one finds a great variety of approaches. While countries obviously differ in the way they organize other sectors (such as agriculture, or education), I cannot think of any major sector in which organizational differences are as large as they are in health care (with the possible exception of religion). Assuming that the relative importance of health in peoples’ preference functions is similar across countries, this is puzzling, since, given the medical technology available, one would think that the most efficient way to organize the health care system ought to be pretty much similar in all countries. Contrary to most other industries, any tendency for health systems across the world to converge to a similar ‘most efficient’ model would seem to be weak, perhaps reflecting the heavily politicized nature of the health policy debate in most countries, and the limited ability of citizens to try out different systems on their own. The diversity in health policy and health system organization applies both with respect to the way health care is financed, and with respect to the organization of the production of health services at the inpatient and outpatient levels. In the following sections, I will sketch ways in which the Canadian system could be reformed in both these dimensions, partly on the basis of the experience that other countries have had with policies that have been quite different from the ones that Canada has adopted since the current system was put in place in the 1970s.8 As a background, the next few paragraphs give a brief summary of the basic principles that have characterized the Canadian system since the 1970s and the 1980s. 5.2.4 The main features of the Canadian system On the insurance side, the Canadian system of funding physician and hospital services consists of a single government plan in each province. In accordance with the rules of the Canada Health Act, the plans are similar in that each one is universal (covers all citizens), comprehensive (in the sense that it covers all physician and hospital
Åke Blomqvist 129 services), and requires no payments at all from patients (in the form of deductibles or co-payments).9 Their coverage must also be portable, meaning that persons are covered even when they fall ill and receive care in another province. They also impose no restrictions on the patients’ choice of provider, in that patients can seek care from any licensed doctor, including specialists, or from any hospital. With respect to the cost of pharmaceuticals, there is more variety, as the Canada Health Act is silent on the subject of covering the cost of outpatient pharmaceuticals. All provinces have government plans that help pay for at least part of the cost of drugs for specific categories of patients (the elderly, those with low income), and some have fairly comprehensive and universal ‘pharmacare’ plans. Many, however, have little or no public funding for the drug costs of the non-elderly and the non-poor. Most of the cost of the provincial plans comes out of general tax revenue, although some provinces impose premiums or dedicated taxes to pay for part of the cost. The role of private insurance is virtually nonexistent when it comes to the cost of physician and hospital services, partly because several provinces (including Ontario and Quebec) have legal rules that explicitly prohibit insurers from offering coverage of hospital and physician services, but private insurance plays a major role in providing coverage of the cost of pharmaceuticals, dental services, and certain other services. While the constitutional responsibility for health policy rests with the provinces, the federal government makes substantial transfers to them which are designated as being ‘for health care’, and which, in principle, are paid only if the provinces abide by the rules of the Canada Health Act when designing and managing their insurance plans. Although government intervention on the insurance side mainly takes the form of direct provision through a single set of provincial plans, with respect to the production of health services, Canada belongs in the ‘public contract’ category in which such services are produced by private agents under contracts with the government plan. This certainly is true for physician services which are almost exclusively provided by independent doctors who work privately in solo or group practice. Technically, it is also true for hospital services, as almost all acute-care hospitals are classified in the category of privately owned non-profit institutions, not as government-owned ones. However, since physicians and hospitals receive all their funding from the provincial governments, it will obviously be the case that the way they operate is heavily influenced by the terms according to which they receive their funding, so in
130
Reforming Canadian Medicare
reality the fact that hospitals are technically not ‘owned’ by provincial government probably has more significance from a political than from an economic point of view. In every province, by far the largest share of physician services, both among family doctors and specialists10 is paid for through feefor-service, on the basis of a fee schedule which has been negotiated between the government and each province’s medical association, and under the terms of the Canada Health Act provinces cannot allow their physicians to charge patients anything beyond the fees specified in the schedule. (There are a few exceptions, in that some services are provided by interns who are paid a salary, and some physicians in teaching hospitals are paid through a mixed system that includes a salary component as well as fee for service; there have also been pilot projects involving alternative payment systems such as capitation, in some provinces, but fee for service continues to dominate.) Although some variety exists in terms of the way hospitals are funded by the provincial governments, funding through an annual negotiated global budget probably remains the most common method, and not all provinces impose ‘hard’ budget constraints on hospitals, in that some have had deficits covered by governments even when they have overspent their budgets. With respect to pharmaceuticals, drugs administered to hospitalized patients are paid for by the hospital, but drugs prescribed by doctors for outpatients are supplied through independent pharmacies and paid for either by the patients themselves out of their own pocket, or by the patients’ insurance plan, private or public. In the main, the features of the Canadian system have remained the same since Medicare in its present form was designed over 30 years ago. While the system has served Canada well in the past, for the reasons noted in the introduction, it now appears to be coming under increasing stress, giving the discussion of efficiency-enhancing reforms more urgency. In the next two sections, I describe certain reform proposals that, in the light of international experience, seem to me to offer the most promise for change in this direction, if we continue to manage health system resources through a set of provincial monopoly plans. (Discussion of opting out, that is, allowing people to substitute a private plan for the public one, is postponed to section 5.5.)
5.3 Reforming Canada’s system of health services provision In this section, I discuss reforms intended to improve the efficiency with which health services are produced, through changes to the rules and
Åke Blomqvist 131 contractual arrangements under which doctors and hospitals deliver care to patients in the public plans. In the recent health economics literature, a distinction is sometimes made between incentives that are intended to influence health care costs and resource use by affecting decisions by patients (principally by requiring patients to pay part of the cost of their own care), and those that influence decisions made by health care providers. The former are termed demand-side incentives, while the latter, which I focus on in this section, are referred to as supplyside incentives.11 5.3.1 Capitation in primary care In looking at the system of health services production, is it useful to make a sharp distinction between different levels of care, in particular, between primary care and that provided at higher (secondary/tertiary) levels.12 Most treatment episodes will include at least one visit to a primary-care provider since, by definition, these are the providers to whom people first go when they have a health problem of some kind. The role of primary-care providers is critical not only because they diagnose and treat (and prescribe drugs for) patients with relatively uncomplicated illness conditions, but also because they must decide when a patient’s diagnosis is such that he/she has to receive care from a specialist and/or in a hospital. Often, the primary-care provider retains a role as an advisor to the patient even when he/she is referred to a specialist or hospital, and during the convalescence phase that follows treatment for serious illness. Traditionally, most primary care in North America and elsewhere has been provided by doctors classified as general practitioners (GPs) or family doctors, though in some countries, such as Sweden, it has been supplied in clinics staffed not only by doctors but also nurse practitioners. In Canada in recent years, a growing share of primary care has come to be provided in hospitals’ emergency departments, and in some other countries (including Japan), hospitals actually maintain large outpatient clinics that provide the same type of care as family doctors would. The notion that the primary-care sector needs strengthening in many of Canada’s provinces has been around for some time, and there is now more and more anecdotal evidence of a shortage of primary-care doctors who are willing to accept new patients, causing more use of emergency departments as a substitute.13 In the long run, dealing with this problem clearly should involve efforts to increase the number of medical students who choose training in family medicine. In the short
132
Reforming Canadian Medicare
run, giving more people access to a regular family doctor can only be based on inducing existing doctors to take on new patients (or inducing more specialists to switch to practicing primary care). A key feature in organizing an efficient system of primary care is the method used to pay for primary-care services. Many countries besides Canada use the fee-for-service method; it is the one used in the basic Medicare plan in the US that covers Americans over the age of 65, and also in several European countries (Germany, France) as well as in Australia and Japan. Generally doctors who treat patients covered by a government plan are only allowed to charge the fee specified by the plan (this is true in Canada, the US, and Japan). In Australia, they can charge above the Medicare rate (a practice referred to as extra-billing) provided they inform the patient in advance. Essentially, each of these countries except Australia thus imposes price controls on primary-care services, which obviously is one way to prevent doctors from taking advantage of the ability they have to control the demand for their own services as a result of the patients’ lack of medical expertise. At the same time, the fact that fees for primary-care services are controlled means that there will not be a market response when shortages arise in local markets, as has happened in the market for family doctors in some parts of Canada. Other methods of paying for primary-care services are salary and capitation. Salary is relatively uncommon, but capitation is the method used in the UK to pay the GPs who provide primary care in that country. In a capitation system, the provider (GP) assumes the responsibility to provide all primary-care services as needed for a given patient over a fixed period of time, in return for a fixed payment from the insurer (the NHS) for that time period. The system implies a restriction on the patient’s choice of primary-care provider, but it is important to recognize that this restriction applies only for the duration of the contract; at given intervals, patients are allowed to choose a new GP if they wish. Obviously, the incentives affecting a doctor who is paid by capitation are drastically different from those under fee for service: the former rewards a high volume of services, while the latter rewards assumption of responsibility for the care of a large number of patients. While information asymmetry between doctors and patients is a potential problem in either case, its effects are different. Under fee for service, it may give rise to ‘overservicing’ of each patient as it rewards the volume of services provided, regardless of the number of patients. Under capitation, on the other hand, it may lead to ‘underservicing’, with doctors being tempted to sign up so many patients that they don’t have enough
Åke Blomqvist 133 time on average to provide each one with proper care. If patients can evaluate the quality of care they are receiving at least to some extent, competition among doctors for patients may alleviate either problem somewhat, but note that for a patient who pays little or nothing out of pocket for medical care, competition for patients will not be effective in alleviating the overservicing problem under fee for service. Capitation as the main method of paying for primary care in a public insurance plan was chosen as a compromise (between fee for service and salary) in the UK at the time the NHS system was put in place after WWII, and it has also been used in the public plans in Italy and the Netherlands as well as some smaller European countries. In recent years it has come into wider use in the US, where different forms of capitation (or mixed systems involving at least an element of capitation) have become increasingly common in managed-care plans in the private sector. To the extent that the public health insurance plans in the US (Medicare for those over 65, and Medicaid for those with low income) have been subcontracting the provision of care to private managed-care plans, capitation has indirectly become an element of those public plans as well.14 To the extent that the lack of an established relationship with a primarycare provider is perceived by many Canadians as a problem, a payment system which rewards doctors for taking on this role certainly makes sense. Moreover, international evidence is consistent with the idea that the supply-side incentives inherent in a capitation system are effective in keeping costs relatively low. While we certainly hear stories from the US which suggest that the supply-side incentives in managed-care plans in that country can give rise to conflicts between patients and insurers about the quality of care, there is little evidence that the health status of individuals in managed-care plans is worse than that of people in conventional plans. On balance, therefore, I believe a strong case can be made for introducing capitation, or at least a mixed payment system with an element of capitation, as the method to pay for primary care in the provincial plans in Canada.15 5.3.2 Capitation, gatekeeping, and fundholding While primary-care services are an important component of any health care system, they typically account for a much smaller share of the total cost of health care than hospital inpatient services. Moreover, in relation to the total cost of outpatient services, the total payments to doctors in some countries are smaller than another component, namely pharmaceuticals. The incentive structure inherent in the methods of
134
Reforming Canadian Medicare
funding hospital services (and the services of the specialist doctors who treat patients in hospitals) and pharmaceuticals thus are major determinants of the system’s overall cost-effectiveness as well. The use of inputs in the health production function like pharmaceuticals and hospital services also tends to involve decisions by primary-care providers, however. For example, many kinds of illness can be adequately treated at the primary-care level, through the use of drugs prescribed by the primary-care provider; in such cases, the physician’s choice of drugs will be an important determinant both of the treatment’s success, and its cost. In cases of certain kinds of more serious illness, there may be a choice whether, or at what stage of the illness, the patient should be hospitalized for more advanced (and typically more costly) treatment than can be provided at the primary-care level. Even when patients are formally allowed to decide on their own whether or not to seek hospital care, in reality most decisions to do so are on the recommendation of primary-care physicians. The same applies to decisions whether or not to seek outpatient care from specialists. Primary-care providers in all the countries mentioned so far influence the cost of pharmaceuticals through their role in prescribing for their patients. In the UK, the role of the primary-care provider in advising the patient with respect to hospitalization and specialist services has also been formalized through assigning them the role of ‘gatekeeper’ whose explicit referral is required in order for an NHS patient to be entitled to specialist care and hospitalization. While this system has been in place for many years, part of the reform package that was introduced in the UK in the 1990s involved strengthening the incentive on GPs to make cost-effective decisions when carrying out their prescription and gatekeeping functions, through a system known as fundholding. Under this system, participating GP practices were given a budget out of which they had to pay for a portion of the cost of the drugs they prescribed to their patients, as well as of the costs of certain kinds of elective procedures undertaken in hospitals following the GP’s referral. Although the terminology will be different, the principle of fundholding will be preserved under the new system of Practice-Based Commissioning that is currently being implemented in the UK.16 A system of gatekeeping and fundholding is a logical extension of the principle of capitation. It gives an incentive to the primary-care givers, who have the medical expertise required to do so, to treat the patients in a cost-effective way, not only with respect to the services they themselves provide, but also with respect to other potentially costly inputs
Åke Blomqvist 135 in the treatment package (pharmaceuticals, hospital, and specialist services). Thus the effectiveness of a system of capitation could be substantially enhanced if it included these features. 5.3.3 Population-based funding and regionalization A second type of reform that I believe could potentially be very helpful in improving the efficiency of health system resource allocation within the framework of a set of monopoly provincial health insurance plans is already underway in all Canadian provinces: increased decentralization (regionalization) of management of the hospital system.17 One beneficial effect of regionalization is that it may contribute to a more efficient and equitable geographic allocation of health services resources, especially if it combined with the principle of populationbased funding. Under this principle, each region is allocated a budget for hospital services that is more or less proportional to the size of the population to be served (perhaps with some adjustments for factors like age structure and others that affect the population’s need for care; the details are less important than the general principle of formula-based allocation, as distinct from one that largely reflects history or political bargaining power). Another obvious advantage of decentralization is that it allows more scope for taking into account local conditions in the allocation process. The role of the regional health boards in the Canadian system differs in some respect from that of their counterparts in two other countries where the decentralization principle either is well established (Sweden), or has been an element of ongoing reforms (the UK since the late 1980s). In the UK, the decentralization process was part of an attempt by the Thatcher government to introduce market-like incentives and competition in the hospital system by means of the so-called ‘purchaser– provider split’. Under the pre-reform system, the NHS hospitals, which are owned by the NHS, were funded through annual budget allocations that were not directly linked to any measures of the quantity or quality of the services they rendered, and their managers were accountable only to the higher-level NHS bureaucracy. One element of the reforms was to give each hospital more autonomy, including more freedom to make their own hiring and firing decisions. More importantly, a new role was created for the District Health Authorities (DHAs, that were a part of the NHS bureaucracy), namely that of ‘prudent purchaser’ of health services for the district’s population. In that role, the DHAs were supposed to negotiate service contracts with the hospitals, with terms that would
136
Reforming Canadian Medicare
include items like the quantity and the quality of the different services that were to be provided, as well as the basis according to which the hospitals would be funded by the DHAs (through which government funding of the hospitals was to be channelled). While the expectation was that DHAs would ‘purchase’ most of the hospital services for their populations from hospitals in their own district, they also had the right to use part of the funds they were allocated to negotiate contracts for particular types of services with outside hospitals, if the latter were able to offer more favourable terms. The system was sometimes referred to as one of ‘internal markets’ for hospital services. In Sweden, the health care system has historically been managed in a highly decentralized manner, with considerable autonomy for over 20 county councils (in a country with a population of no more than about nine million), each one of which had responsibility for managing the government-funded health services system for their population. The health system management in the Swedish county councils is different from the district-level management under the UK reforms, in several important respects. First, ultimate responsibility rests with an assembly of officials who are elected as part of periodic local elections in Sweden (in the UK, DHA managers are appointed by the NHS). Secondly, the county councils are also responsible for funding the health care system, and have the right to levy a proportional local income tax, most of which goes to pay for the health care. (Part of the cost, however, is paid by the central government.) Third, while the DHA managers in the UK were only responsible for hospital services (not primary care), Sweden’s county councils are responsible for organizing both the hospital and the primary-care components of the system. During the 1990s, many county councils in Sweden experimented with management reforms inspired by the UK experiment with a purchaser–provider split, with considerable success in some areas, especially the larger cities.18 The examples provided by the UK and Sweden can serve as models for the role of the regional health authorities that are now part of most of Canada’s provincial health services systems. Because Canada’s hospitals already are privately owned and managed in a decentralized fashion, the regional authorities could naturally assume the role of ‘prudent purchasers’ of health services for their populations, along the lines of the UK DHAs, with a mandate to negotiate contracts about the terms under which the hospitals will supply them. Their ability to improve the efficiency of health services resource use could be further enhanced if their mandate were broadened to include responsibility
Åke Blomqvist 137 for funding and contracting for outpatient physician services as well as hospital services, as it does in the case of Swedish county councils. This could be done in conjunction with a switch to paying for primary care through capitation, as discussed in the previous section. Although regionalization could play a helpful role even if the regional authorities were appointed, the possibility of moving to a system whereby their functions are carried out by elected officials should be explored as well.
5.4 Funding and coverage in the government insurance plan The discussion in the previous section refers to methods of government intervention in the market for health services. Because of the expectation that a purely private and unregulated system of health insurance is likely to entail substantial economic inefficiency as well as be inequitable, governments in almost all countries intervene, directly or indirectly, in that market as well. As in Canada, the intervention in most advanced countries takes the form of offering the citizenry coverage through one or more plans that are organized and/or funded by the government. Government plans differ in a number of important dimensions, however. First, the extent of coverage varies, in terms of the patients’ right to choose from which provider to seek care, what categories of benefits are included, and in terms of what portion, if any, of eligible costs that patients have to pay out of pocket. Second, the government’s share of the cost may be financed in a variety of ways: out of general tax revenue, through payroll deductions, or by specific insurance premiums. Of particular interest in the Canadian context, of course, is the division of financial responsibility between the federal and provincial levels of government. In the following paragraphs, I discuss various alternatives to the way the Canadian system of provincial plans is designed in these respects, and changes that could be made that would improve it. For now, I continue to assume that most insurance is provided in the form of a single monopoly plan in each province. 5.4.1 Paying for health services: Federal versus provincial funding With respect to the method of raising the public-sector revenue required for government health insurance plans, a sharp distinction is usually made in the literature between, on the one hand, the social insurance model in which health insurance is administered
138
Reforming Canadian Medicare
and financed through multiple ‘sickness funds’ to which the insured contribute (typically in the form of a pay-roll contribution), and, on the other hand, funding from general government revenue. As long as social insurance contributions are levied on a compulsory basis, they are of course equivalent to a tax, so that in that sense the debate over the relative merits of social insurance or tax financing really belongs in the realm of public finance, not health economics. Thus, the analysis of what constitutes a good way of financing the public sector’s cost of health care should be no different from the analysis of what constitutes efficient and equitable ways of raising government revenue in general. In a universal system, the same applies to funding through compulsory premium payments. One can also argue that what is perhaps the most hotly debated issue in health policy in Canada, namely what shares of health care costs should be paid by the provincial and federal governments, also is one that in reality has little or nothing to do with health policy. The logic of this argument, of course, is that every dollar that is transferred from the federal to the provincial governments adds to provincial revenue in the same way, whether it is designated as being ‘for health’ or just part of general revenue equalization. As long as it is true that the federal transfers that are designated as being ‘for health’ are smaller than the total amounts that each province spends on health, this logic seems compelling, so the debate on this issue should properly be seen as a distorted version of the more general one that concerns federal–provincial revenue sharing more broadly (the vaunted ‘fiscal imbalance’, in current parlance).19 What is important for health policy, however, is the existence of the Canada Health Act and the way it offers the federal government an opportunity to impose restrictions on the way the provinces discharge their constitutional responsibility for health policy. Having thought about the issue for a long time, I have come to the conclusion that the most beneficial health system reform that could be undertaken in Canada would be to dramatically revamp the role of the Canada Health Act (or even abolish it) and merge the federal transfers ‘for health’ into the general revenue equalization programme. Since I share the belief of the defenders of the CHA that it incorporates values that most Canadians support, I am convinced that each province would continue to uphold the most important principles in the Act even if they were not compelled to do so. At the same time, abolishing the Act would give more freedom to individual provinces to interpret these principles in a flexible way, and to experiment with reforms that, under the present
Åke Blomqvist 139 system, can be blocked by those with a vested interest in the current system under the guise of ‘defending the Canada Health Act’. I will return to this theme below. 5.4.2 The role of user fees Unlike Canada’s provincial insurance plans, the public plans in most other countries require some degree of co-payment from patients for insured services (user fees). This is true about the US Medicare plan, the Japanese social insurance system, and in European countries like France and Sweden. In Australia too, patients are required to pay a portion of the cost out of pocket when they choose treatment as private patients in hospital, or when they receive outpatient treatment from a doctor who does not ‘bulk bill’.20 Among advanced countries, only a few (the UK, Canada) impose no user fees for hospital and physician services. When user fees are relatively large as a percentage of costs (as they are in Japan), they typically are accompanied by some kind of stop-loss provision in the form of an upper limit on the amount of co-payments that an individual or family may be responsible for over a given time period, and there may be special arrangements under which persons who have been classified in a low-income category are exempt. In some countries (France, the US), private health insurance plans are available that will cover the user fees imposed by the public plan; in others (Australia, Japan), such plans are not allowed or do not exist. There is probably no issue that has been more hotly debated in Canadian health policy than that of user fees. In my view, the debate has become largely symbolic and intertwined with the competition between the federal and provincial levels of government for the position as the best defenders, in the eyes of the voters, of our publicly funded system of health care. My own view is that while demand-side incentives such as user fees are less important tools for controlling health care costs than supply-side incentives, limited user fees can play a useful role as part of the system of financing health care. Existing evidence, as I interpret it, on balance supports the idea that user fees can be effective in reducing total health care spending without seriously affecting the ultimate health status of most people.21 Taken together with the fact that tax revenue has an excess burden, this suggests an efficiency argument for financing a portion of total health care costs in this way, at least as long it can be done with relatively low administrative costs. While opponents also regard even limited user fees as inequitable in the sense that they tend to reduce the medical care utilization of
140
Reforming Canadian Medicare
poor people by more than that of the rich (as economic theory would predict), this can be countered by special provisions to exempt the poor. A version of the tax-based user fee scheme proposed by Aba, Goodman, and Mintz (2002) would meet this criterion. 5.4.3 Paying for pharmaceuticals Public plans in other countries also differ from the Canadian provincial plans in another important way: in the nature of the coverage that is provided (in terms of what categories of costs are covered). In particular, most public plans in advanced countries now cover the cost of pharmaceuticals (whether used in hospitals or by outpatients), as well as the cost of hospital and physician services. This certainly is the case in the UK and Australia as well as in Japan, and following a recent controversial reform in the US, that country’s Medicare plan now covers a substantial part of the cost of pharmaceuticals as well (Stuart et al. 2005). The cost of drugs has gradually become a larger and larger share of the aggregate cost of health care in Canada, and the arrangement whereby this component of the cost of illness is treated asymmetrically from physician services is becoming harder and harder to rationalize. The case for a plan that integrates the cost of pharmaceuticals with coverage of hospital and physician services costs becomes especially strong if the providers are paid in such a way that they have an incentive to make prescription decisions in a cost-effective way, through some form of fundholding (see the earlier discussion). 5.4.4 Restrictions on choice of providers One feature of the provincial health insurance plans as currently organized is that they place essentially no restrictions on the patients’ choice of provider: a patient has the right to seek care from any licensed doctor (including specialists) or hospital. The same is true in the public plans in Australia and Japan, as well as in the regular US Medicare plan that covers persons over the age of 65. In countries where insurance is offered through private plans, some of them restrict this right. This is true in most managed-care plans in the US, or more generally, in plans with lists of ‘preferred providers’; in such plans, patients are only covered if they are treated by providers who appear on the insurers’ list. Restrictions of this kind also exist in some public plans. For example, patients who are covered by some of the state-managed Medicaid plans in the US are also limited to specific providers, and, implicitly, the capitation and gatekeeping system in the UK restricts patients’ choice of primary-care provider at a given time, as patients can only seek care
Åke Blomqvist 141 from or through the GP with whom they are registered during a given contract period. Some people, including some health policy analysts, perceive restrictions on the patients’ choice of provider as a major disadvantage. Those who feel this way are also likely to oppose the use of capitation and gatekeeping, since these methods imply such restrictions. They may do so even though they recognize that capitation and gatekeeping are likely to give rise to more cost-effective patterns of care, something from which patients will benefit as taxpayers. Others, however, may view this trade-off differently and, on balance, support a capitation system. This suggests that, ceteris paribus, a good alternative would be to offer individuals a choice between the current provincial plan and an alternative version in which capitation was used. But if individuals were to be given this kind of choice, why limit their options to only two specific types of plans? This line of reasoning thus leads to the issue of whether individuals should be given the choice to opt out of the public plan more generally, and sign up with private insurance plans instead.
5.5 Beyond monopoly: Allowing opting out While the discussion in the previous sections has dealt with reforms within the context of a system that continued the framework of a monopoly public plan in each province, in this section I consider a more radical reform alternative in which current rules against offering private insurance covering hospital and physician services would be abolished. Moreover, I now consider the case where individuals would be allowed to opt out of the public plan. By ‘opting out’, I mean a situation in which an individual may choose not to be covered by the public plan, and is financially compensated for making this choice, at least to some extent. (For example, in the German or Dutch systems, certain categories of individuals have been allowed to opt out in this sense; when they do, they are compensated by not having to pay the payroll tax that partially funds the health care system.) Among the public plans referred to so far, only the US and Australian Medicare plans have some form of opting out. It is not allowed in Japan (where there is little or no private health insurance). In the UK, there is no Canadian-style rule against private insurance covering the same services as the NHS, but those who purchase private insurance for the purpose of covering the cost of private care outside the NHS are not financially compensated for doing so. In the US Medicare plan, opting out takes the form of allowing those in the plan who wish to do so, to enrol in
142
Reforming Canadian Medicare
an approved private managed-care plan instead.22 The premium on this plan is then subsidized by Medicare, through transfer of an amount corresponding to the estimated average that Medicare would have paid for the person had she or he stayed with the public plan. In Australia, individuals can choose to purchase private insurance that will pay when they choose to receive hospital treatment as ‘private patients’, either in a public or a private hospital. If they do purchase private insurance, the government currently pays a subsidy equivalent to 30 per cent of the premium cost if the plan is from a set of approved ones.23 As noted earlier, in Canada opting out is impossible not only in the sense that there is no compensation if an individual decides that he or she does not want to belong to the public plan; in most provinces, alternative private plans that cover the same services (hospital and medical) that the public plan does are not legal. The possibility of allowing opting out in some form has not been a significant part of the debate about health policy in Canada since our system of universal publicly funded health insurance was created in the 1970s. However, the recent Supreme Court of Canada decision in the Chaoulli–Zeliotis case has, at the time of working, to some extent revived the discussion about allowing at least somewhat more of a role for private health insurance for hospital and medical services.24 The opposition to any form of opting out has primarily stressed two arguments. The first one is that a health insurance system with a monopoly plan (a ‘single payer’ system) is much simpler and cheaper to administer than a multiple-payer plan with competing insurers. In a series of influential studies, Woolhandler and Himmelstein have compared administrative costs in Canada and the US, finding that they are much higher south of the border; their most recent study (Woolhandler, Campbell and Himmelstein 2003) suggested a figure for the US that was three times the Canadian one in absolute terms, and almost twice as high in relation to total health care costs (31.0 vs 16.7 per cent in 1999). The second, and perhaps more important, basis for opposing any form of opting out is that it would open up the possibility that different individuals would receive different standards of health care when ill, that is, it would lead to what is popularly known as two-tiered medicine. The idea that those willing and able to pay for a higher standard of medical would be able to do so, in turn, is considered as incompatible with prevailing Canadian values regarding equity. The obvious counterargument, as discussed above, is that the equity objective motive is more efficiently promoted by redistributing more real income to the poor,
Åke Blomqvist 143 rather than by trying to raise the standard of medical care they receive when ill by forcing everyone to belong to the same monopoly health insurance plan. In spite of this, the dynamics of the health policy debate in Canada have been such that no political party has been willing to lay itself open to the charge that it is in favour of ‘an American-style model of two-tiered medicine’. My own view is that in a system with a well-managed monopoly plan, the efficiency losses from not allowing opting out may not be very significant, and in the past I have put other types of reforms (in particular, introducing some form of capitation in primary care) higher on the list of priorities than changes in the rules governing private insurance. Over the past several years, however, it has become increasingly difficult to remain optimistic regarding the ability of our current health care system to respond to the challenges of evolving medical technology and an aging population. As previously noted, in some respects, the system now seems to be performing worse than it did some years ago, for example, with respect to access to a regular family doctor, and in terms of the waiting times required for certain types of elective but medically urgent surgery. While this trend may reflect the fact that there are more effective treatment methods now than was the case some years ago, rather than a deterioration in the system’s ability to provide a given type of service, this doesn’t alter the impression that aggregate health resources could be better managed, and that the incentives on political decision-makers that exist under the system of provincial monopoly plans do not appear strong enough to bring about a reasonable pace of progress in this regard. The question therefore arises: would some degree of competition from private plans be an efficient catalyst that could bring about better performance in the publicly funded plans? I am inclined to believe that the answer may be ‘yes’, and that we should start to think seriously about allowing citizens to opt out, and private plans to compete with the provincial ones. While proposals to allow Canadians to enrol in private insurance plans would be characterized as moving towards ‘an American-style health care system’, it is important to note that a reform along these lines would still leave Canada with a system that would remain consistent with the principle of universality, in the sense that everyone would have the option of continuing to belong to the public (provincial) plan. Only those who chose to opt out would not belong to the public plan. By requiring that anyone who opted out would have to first enrol in an approved private plan, the principle that everyone would have to be insured could be preserved, so that in this sense, a
144
Reforming Canadian Medicare
system that allows opting out would still be fundamentally different from the US one in which a substantial portion of the population has no health insurance at all. 5.5.1 Models for opting out Legislation to allow opting out would have to address a series of fundamental questions. Perhaps the most important one would be the extent to which those who chose to opt out would be compensated for doing so. Under some of the European models in which a large proportion of health care costs is financed through an earmarked payroll tax, those who opt are compensated by being exempted from this tax, as previously noted. In the Australian system, by contrast, there is no reduction in the tax payment obligations of those who opt out; instead, the Australian government pays them a subsidy equivalent to 30 per cent of the premium for the private insurance plan in which they sign up. (In addition, the current Australian system also imposes a small surtax on high-income individuals who do not sign up for a private plan.) Since provincial health insurance plans in Canada are largely financed out of general revenue, compensation for those who opted out of the public plan would also have to take the form of some type of subsidy. It could be designed either along Australian lines (with the subsidy being defined as a percentage of the actual premium), or as an absolute amount independent of the taxpayer’s income, but perhaps dependent on other factors (such as age) that would influence his or her private insurance premium. Even in countries that do allow opting out, certain restrictions are typically imposed on the conditions under which an individual is allowed to do so. In Germany and Holland, only individuals with income above a certain level were earlier given this right, while in Australia, opting out has implicitly been allowed only if the individual has obtained coverage in a private plan that meets certain minimum standards with respect to what services are covered, and the extent to which the patient is required to make co-payments for covered services. (The same is true in the Medicare + Choice plan in the US.) To some extent, these rules may simply reflect a spirit of paternalism, but they can also be interpreted as a reasonable response to the moral hazard that exists in a situation where seriously ill people will not be refused treatment in public institutions even if they cannot pay. However, part of the purpose of allowing private insurers the opportunity to compete with the public plan is to encourage experimentation with different
Åke Blomqvist 145 payment and insurance mechanisms. This suggests that the regulation defining eligible private plans should be kept general and simple, with as few restrictions as possible imposed on their characteristics. For example, plans should be allowed to limit coverage to include services of only specific providers, and to negotiate with providers regarding the way they would be reimbursed and the rates of reimbursement.25 With respect to premium rates, a critical issue is whether they would be regulated and whether insurers would be allowed to discriminate by risk classes and/or refuse enrolment of bad risks; another critical question is whether insurers would be allowed to offer employment-related group plans as opposed to individual ones. In general, concerns about equity would support regulation requiring the insurer not to discriminate, and to offer insurance on an individual basis, not as group plans. While enforcing such regulations is not easy, the Australian example suggests that it can be done reasonably well. At the same time, regulation may indirectly have the result of reducing the intensity of competition, and one may argue that the equity-based case for regulation is less strong in a situation where everyone has access to insurance in the default public plan. The experience of some countries with mixed public–private systems suggests that another important issue will be the rules under which providers who treat patients in the public system may supply services to private patients as well. If payment rates are such that treating private patients becomes more profitable for doctors and hospitals than treating public ones, it is likely that private patients get preferential treatment, in terms of waiting times or in other ways. One obvious way to avoid this problem is for the public plan to not allow doctors and hospitals who treat public patients to also treat private patients. However, this may effectively prevent private insurance from competing in smaller communities. Designing payments mechanisms and monitoring arrangements in the public plan that allow effective competition between the public and private plans, while at the same time ensuring that clients from both plans receive care of high quality, is a task that requires considerable care, but there are instances where it seems to have been done successfully enough to avoid major complaints. 5.5.2 Private insurance and the choice of technology As implied by the preceding discussion, on balance, I favour a reform that would permit individuals to opt out of the public plan, and insurers to offer alternative private plans subject to relatively little regulation, not
146
Reforming Canadian Medicare
only because it would offer consumers more choice, but also because competition with private insurance would put more pressure on the managers of the public plan to adopt various efficiency-enhancing changes. In particular, it might indirectly cause some of the principles that are involved in managed-care plans elsewhere in the world to become more prominent in the Canadian system, something that I think would be beneficial. But there is also another reason why I believe more competition in the health insurance market is becoming increasingly important, namely the dramatic advances that are taking place in our ability to treat various forms of illnesses through more and more sophisticated (and therefore, potentially expensive) interventions. While the expanding opportunities in this regard can obviously improve human welfare greatly, a possibility that one must recognize is that medical technology may become so advanced and potentially resource-demanding that it will no longer be possible to let resource use in the health care sector be governed by the simple principle that, implicitly, underlies the way we currently make many spending decisions: that everything which improves health is worth doing. Below I will refer to this as the ‘lexicographic principle’. As any economist realizes, we are already in a situation where no health care system in the world has enough resources to do everything that may improve health for anyone; implicitly, there is already ‘rationing’, most often in the form of waiting times for certain kinds of care. If one accepts the lexicographic principle, the answer to the question what we should do to improve the situation simply is that we should spend more resources until waiting times are brought to zero. But if we don’t accept it, the problem becomes different. We then have to recognize that there may be some types of treatment that are so costly but yield such small incremental health benefits that they are not worth doing. One is then led to an efficiency criterion such as a critical minimum value for dollars per Quality-Adjusted Life Year (QALY) that is becoming more widely suggested as a guide to resource allocation in the health economics literature. In some countries, attempts have been made to use this criterion in making decisions about what kinds of services will be covered in public insurance plans.26 But a single critical QALY value would only be sufficient to allocate resources efficiently in an economy if each one of its citizens assigned the same value to an incremental QALY, and if the relative valuation of different disease conditions that affect quality of life were the same for every individual. These conditions are not likely to hold in reality, even
Åke Blomqvist 147 approximately. The value that will be assigned to QALYs of different kinds at different ages (for example, for premature babies with serious handicaps, people with different kinds of disabilities or for very old individuals with illnesses that subject them to severe pain) will be very different between people with different cultural backgrounds, and with different religious persuasions. In such circumstances, there is potentially a case for allowing individuals, or communities of individuals, the leeway to enter into different ex ante health insurance contracts with treatment rules that would best correspond to their individual preferences with respect to how they would want to be treated in various circumstances. This is unlikely to happen if everyone has to belong to the same monopoly insurance plan, but may happen at least to some extent in a more competitive insurance market where private insurers are allowed to compete with the public plan. In practice, of course, it may not be easy to write explicit insurance contracts that specify different ‘treatment protocols’ for specific disease conditions. However, in countries where private insurance plans compete, they do have some differences. For example, most people probably understand that if they are insured through a managed-care plan in which doctors have an incentive to hold the costs of care down and may be subject to tight guidelines for drug prescriptions and second opinions before surgery, on average they are likely to be treated less aggressively than if their care is provided by doctors paid on the basis of fee for service under a conventional plan. Or, again, most individuals probably realize that the expected cost of their care if they are near death is likely to be lower if they sign advance directives regarding the circumstances when they want to be resuscitated, and in principle there is no reason why a private plan could not recognize this by offering cheaper plans to those who do. 5.5.3 Opting out: What became of ‘market failure’? In an earlier section, I reviewed the various types of market failure that have been cited in the literature as arguments for government intervention in the markets for health services and health insurance. Yet in this section I argue that allowing consumers access to a relatively unregulated market for health services through private insurance plans would be a good policy. Is there an inconsistency here? To my mind, there is not, in the following sense. Some of the earlier literature on market failures in health care, while correct as far as it went, was misleading in that it did not recognize that institutional innovations in private markets could overcome some of the problems
148
Reforming Canadian Medicare
it emphasized. Specifically, the predictions that competitive markets would fail because of information asymmetry between patients and providers and moral hazard, were essentially based on a model in which providers were paid by patients via fee for service, and insurance was through conventional plans that simply paid their share of the bills ex post. The expectation that competition among providers in that environment would not be particularly effective in keeping fees down or produce an efficient pattern of care, certainly seems reasonable. But when insurance is through various forms of managed-care plans, the situation is different. In these circumstances, the focus shifts to fee structures and contracts negotiated between providers and insurers, so information asymmetry is less of a problem. Managed-care plans also are better able to overcome the moral hazard problem (insured patients’ tendency to over-utilize medical services is likely to be less significant when providers do not have an incentive to support high levels of utilization). If competition among private insurance plans is allowed, the prediction would thus be that managed-care plans would tend to be preferred over conventional plans by most people. The experience in the US health insurance market over the last several decades certainly is consistent with this prediction. While the early literature may thus have underestimated the ability of competitive markets to deal with some of the potential sources of market failure, it may also have overestimated the ability of governments in Canada to design and manage a cost-effective health care system based on provincial monopoly plans. While international comparisons of health care system performance on an aggregative basis are difficult, and every health care system has its critics, it is easy to get the impression that countries like Australia, the UK, France, and Sweden have been able to generate more open and pragmatic approaches to the process of reform and management of their health care systems than the nearparalysis that the Canadian system of divided federal and provincial jurisdiction has produced in this area. The above arguments apply principally to the potential market failures due to provider–patient information asymmetry and moral hazard. However, managed-care plans are not effective in overcoming market failure associated with information asymmetry between patients and insurers regarding risk of illness, and the resulting tendencies towards adverse selection. Also, when individuals differ substantially in risk of illness (as well as in ownership of various types of income-earning assets, including inherited human capital), the equity motive becomes another reason for government intervention of some sort, as discussed
Åke Blomqvist 149 earlier. However, while these factors constitute a compelling reason, to my mind, for government to offer a health insurance plan that everyone can choose to belong to, they do not justify the suggestion that the government plan should be the only plan.
5.6 Conclusion: Prospects for health system reform in Canada In many people’s perception of what it is that they like better about living in Canada than in the US, one characteristic of our health care system plays an important part: the principle that everyone, regardless of income or wealth, has access to needed health care. However, concerns relating to the performance of the health care system consistently rank high on lists of problems that respondents to public opinion polls want to see addressed, and many people believe that in reality, the implicit guarantee of access to medical care of high quality is being eroded, as waiting lists are lengthening and as more and more people are unable to find a regular family doctor. At the same time, the aggregate costs of care keep rising, so that remedying the problem by large-scale increases in public sector health care spending does not seem to be on the cards. In the circumstances, finding ways to make the system more cost-effective ought to be a priority. In previous sections, I have sketched certain reforms that I think could move us in this direction. With respect to the system of health services production, reforms such as regionalization of the system of health services management and introduction of new payment methods such as capitation in primary care, could potentially yield substantial dividends in terms of improved cost-effectiveness. On the insurance side, integration of pharmaceuticals in the benefit package of provincial insurance plans, and introduction of a system of tax-based user fees could do the same. Another way to promote reform and institutional innovation, finally, would be to expose the public insurance plans to more competition by allowing citizens the right to opt out of the public plans and obtain coverage through private insurance plans instead. 5.6.1 Prospects for reform: Has the Canada Health Act outlived its usefulness? Although politicians at all levels appreciate the importance that voters attach to the health care system, recent history does not inspire much confidence in the prospects for meaningful reform to deal with its problems. As should be clear from the preceding discussion, I believe
150
Reforming Canadian Medicare
that to a large extent, the problem is the division of responsibility for health policy between the federal and provincial governments. In my view, therefore, Canada would, on balance, be better off if the current federal–provincial arrangements in health care were abolished and replaced by laws under which the provinces were given full responsibility for managing their health care systems. Under such a change, the federal government would no longer be responsible for enforcing the provisions of the Canada Health Act, and the federal transfers to the provinces ‘for health’ would just become part of the general system of federal– provincial revenue sharing.27 Some of the provisions of the CHA deal explicitly with the penalties that the federal government can impose on the provinces if they are deemed not to have lived up to the obligations the Act imposes on them; these provisions would of course have to be changed if the responsibility for health policy were transferred to the provinces. However, with the exception of the proposal above to allow the introduction of limited user fees (which would violate the Act’s current ‘accessibility’ provision), none of the reforms discussed in this paper would explicitly violate provisions in the CHA, not even the proposal to allow opting out. Thus the basic principles of the Act (which I take to be those of comprehensiveness, universality, and portability) could continue to be the law of the land even if health policy and management became a purely provincial responsibility, and even though enforcement would be through the courts, not by the federal government. Thus my answer to the question in the heading is that no, the CHA has not outlived its usefulness. In particular, its provisions regarding universality and comprehensiveness (and even, perhaps, a weaker form of that referring to accessibility) are simply expressions of the fundamental principle on which the Canadian health care system is based: that every citizen, regardless of income or other characteristics, should have the option to be covered by a publicly funded health insurance plan of good quality. I share the view of most Canadians that this is a good principle, and that the US could become both a more equitable society and even a more efficient economy, if it were to adopt this principle (as almost every other developed country has done). But in contrast to those who elevate our health care system and the Canada Health Act to iconic status, I don’t support this principle for its own sake. Instead I support it because I think it constitutes a relatively effective way to improve (indirectly) the distribution of real income, broadly defined, and a pragmatic response to market failure in private insurance.
Åke Blomqvist 151
Notes 1. The best-known reference on the history of the Canadian health insurance system is Taylor (1987); another readable and interesting account is Naylor (1986). Marchildon (2005) contains a very detailed and updated account of its current state, with historical background. 2. A sample of David’s thinking on social policy in Canada is the introduction to the collection he edited for the MacDonald Royal Commission (Laidler 1985). 3. When my monograph was published (by the Fraser Institute; see Blomqvist 1979), it got instant negative publicity in the form of a particularly vicious editorial in the Globe and Mail, which appeared while I was en route to a sabbatical in Sweden. When I was later sent a copy, it was accompanied by a copy of a letter in my defence that the paper had published a few days after the editorial. The letter, which was written by David, countered most of the criticisms raised in the editorial much more effectively than I could have done myself. 4. The early work of Evans (1974) marked the origin of the vast literature on supplier-induced demand in health care. (For a review, see McGuire 2000). Evans and his co-workers (see Barer, Evans, and Stoddart 1979) subsequently were very influential in supporting the 1984 modifications to the Canada Health Act under which provinces effectively were compelled to eliminate all user fees for hospital and physician services. 5. Although there is a widespread belief in Canada that moral hazard cannot be effectively countered by increasing the consumer’s degree of cost sharing, the evidence from the large experimental study undertaken by the Rand Corporation in the 1970s suggests otherwise; see Newhouse (1993). 6. For a survey, see Cutler and Zeckhauser (2000). A particularly readable and convincing example of the potential quantitative significance of adverse selection is in Marquis (1992), who derived her data from the Rand study. 7. For a theoretical model illustrating this principle, see Blomqvist and Horn (1984). 8. The exposition in these sections draws heavily on ideas that I have been discussing in greater detail elsewhere, going back to my 1979 monograph and later in the volume that I edited with David Brown for the C. D. Howe Institute (1994), and in the C. D. Howe Benefactors Lecture (2002). 9. This is the ‘accessibility’ provision. In an earlier version of the CHA, it only specified that the plan must give ‘reasonable’ access; the current version dating from 1984 specifies that no user charges may be imposed. Another provision of the CHA specifies that provincial plans must be ‘publicly administered’. Although this provision is often used in popular debate to justify opposition to private insurance, or to health services produced in private clinics, the actual wording makes it clear that it refers only to the administration of the public plan; it does not say anything about the role of competing private insurance, nor about the question whether any health services can or cannot be privately produced. 10. Technically, many family doctors are now specialists in ‘family medicine’, but the term ‘family doctor’ is often used to describe doctors in general practice who have not undergone specialist training.
152
Reforming Canadian Medicare
11. Much of the early literature in health economics was principally concerned with demand-side incentives. One of its main conclusions was that, because of information asymmetry between doctors and patients, these incentives were not likely to be efficient as tools to control health care costs or to bring about an efficient pattern of resource use in health care. By extension, many early writers argued that for this reason, conventional microeconomics did not have much of anything useful to say about health policy (see, for example, Evans 1984). However, as analysis of situations with asymmetric information has become more and more prominent in microeconomics, and as the focus in health economics has shifted to analysis of the incentives on providers, this has changed, and health economics is once again firmly established as part of mainstream microeconomics. 12. Since the latter typically is given in a hospital, an almost equivalent distinction is that between outpatient and inpatient care, except that certain types of outpatient care provided by specialist physicians (including certain forms of outpatient surgery) may be considered as being part of secondary, not primary, care. 13. Public concerns with access to a regular family physician are emphasized in the recent report by the Canadian Medical Association (2006). 14. The option to substitute enrolment in a private managed-care plan for Medicare coverage has existed in the US since the mid-1980s; in recognition of this option, the plan is often referred to officially as ‘Medicare+Choice’. 15. A wholesale transition to a system of capitation (or a mixed payment system) for all primary-care providers would obviously be very difficult and costly to organize and administer. For that reason, I think a strong case can be made for offering doctors and patients an option under which they can choose one or the other (this is further discussed in Blomqvist 1994 and 2002). This would also represent a modest step towards introducing some degree of consumer choice among different types of insurance. 16. The process of health system reform, including the fundholding experiment and the so-called purchaser–provider split discussed below, are well described in Robinson (1999). For a review of the evidence regarding the effectiveness of the fundholding experiment, see also Gosden and Torgerson (1997). 17. The regionalization process has by now been going on for more than ten years. For a discussion, see Lomas (1997). 18. For a review of the Swedish system, see Hjortsberg and Ghatnekar (2001). An evaluation of the Swedish reform experiences is Harrison and Calltorp (2000); see also Blomqvist (2001). 19. It is probably the case that the preoccupation with the question what share of health care costs are paid by the two levels of government is a historic relic from the time when the federal contribution took the form of something like a matching grant (that is, the amount of federal transfers for health depended on the amounts that provinces had spent). Since the mid1970s, the transfer system no longer has this feature. 20. ‘Bulk billing’ is when doctors do not give the patients the bill for their services but instead send it directly to the insurance plan. Doctors who do not follow this practice collect their fee from the patient, who then has to file for full or partial reimbursement from the plan.
Åke Blomqvist 153 21. Although the Rand study (Newhouse 1993) was based on data collected over 30 years ago, it remains the mother of all empirical studies on this issue, and I am not aware of any major later studies that convincingly contradict it. An interesting finding from the US is that there has been a trend towards more use of patient charges in managed-care plans as well, even though most of these plans employ strong supply-side incentives in order to control costs; see Gabel (1997), cited in Glied (2000). 22. As noted in an earlier footnote, because of this option the US Medicare plan is sometimes referred to as ‘Medicare+Choice’. 23. For a careful analysis of the Australian health funding system with an emphasis on the role of private insurance, see Colombo and Tapay (2003). 24. See Canadian Medical Association (2006). 25. When managed-care plans were allowed to compete with the public plan in Holland, an issue that became important (and ultimately resulted in court action) was whether the plans should be obliged to accept any provider who wanted to join the plan. (A rule to this effect would make it harder for plans to negotiate favourable conditions from providers.) See Flood, Colleen (2000, p.72). 26. A well-known example was the attempt by the state of Oregon to formalize this type of cost-effectiveness criterion for its Medicaid plan (the plan that provides publicly funded insurance for persons with low income). For a discussion of QALYs in general and the Oregon example in particular, see Kaplan (1995). 27. Another alternative, of course, would be to give full responsibility for health system management to the federal government (as in the UK or the US Medicare plan). While this is a possibility, it is difficult to envisage as dramatic an expansion of the federal government’s role as this would entail, given the nature and history of the Canadian federation.
References Aba, Shay, Wolfe D. Goodman, and Jack M. Mintz (2002), ‘Funding public provision of private health: The case for a co-payment contribution through the tax system’. C. D. Howe Institute Commentary 163, Toronto: C. D. Howe Institute. Blomqvist, Åke (1979), The Health Care Business: International Evidence on Private vs. Public Health Care Systems. Vancouver: The Fraser Institute. —— (2001), ‘International health care models: Sweden’, submitted to the Standing Senate Committee on Social Affairs, Science and Technology, Study on the State of the Health Care System in Canada (The Kirby Committee), available as part of Appendix A to Volume 3 of the Committee’s report at www. parl.gc.ca. —— (2002), Canadian Health Care in a Global Context. Toronto: C. D. Howe Institute (available for downloading at www.cdhowe.org). Blomqvist, Åke, and Henrik Horn (1984), ‘Public health insurance and optimal income taxation’, Journal of Public Economics, vol. 24, pp. 353–73. Blomqvist, Åke G. and David M. Brown, eds (1994), Limits to Care: Reforming Canada’s Health System in an Age of Restraint. Toronto: C. D. Howe Institute.
154
Reforming Canadian Medicare
Canadian Medical Association (2006), ‘It’s about access! Informing the debate on public and private health care’. CMA Task Force on the Public–Private Interface. Colombo, Francesca, and Nicole Tapay (2003), ‘Private health insurance in Australia: A case study’, OECD Health Working Papers, No. 8. Paris: OECD. Culyer, Anthony J., and Joseph P. Newhouse, eds (2000), Handbook of Health Economics. Amsterdam: North-Holland, Vols 1A and 1B. Cutler, David M., and Richard J. Zeckhauser (2000), ‘The anatomy of health insurance’, in Culyer, Anthony J., and Joseph. P. Newhouse (eds), Handbook of Health Economics. Amsterdam: North-Holland. Ch. 11, pp. 563–644. Evans, Robert G. (1974), ‘Supplier-induced Demand’, in M. Perlman (ed.), The Economics of Health and Medical Care. London: Macmillan, pp. 162–73. —— (1984), Strained Mercy: The Economics of Canadian Health Care. Toronto: Butterworths. Flood, Colleen M. (2000), International Health Care Reform: A legal, economic and political analysis. London: Routledge. Gabel, J. A. (1997), ‘Ten ways HMOs have changed during the 1990s’, Health Affairs, Vol. 16, pp. 305–22. Glied, Sherry (2000), ‘Managed care’, in Culyer, Anthony J., and Joseph P. Newhouse (eds), Handbook of Health Economics. Amsterdam: North-Holland. Vols 1A and 1B, Ch. 13, pp. 707–45. Gosden, T., and D. Torgerson (1997), ‘The effect of fundholding on prescribing and referral costs: A review of the evidence’, Health Policy, Vol. 40, pp. 103–14. Harrison, Michael I. and Johan Calltorp (2000), ‘The re-orientation of marketoriented reforms in Swedish health care’. Health Policy, Vol. 50, pp. 219–40. Hjortsberg, Chatarina, and Ola Ghatnekar (2001), Health Systems in Transition: Sweden. Copenhagen: WHO (on behalf of the European Observatory on Health Systems and Policies). Kaplan, Robert M. (1995), ‘Utility assessment for estimating QALYs’, in Frank Sloan (ed.), Valuing Health Care. Cambridge University Press, Ch. 3, pp. 31–60. Laidler, David E. W., ed. (1985), Approaches to Economic Well-Being. Volume 26 in a series of studies commissioned by the Royal Commission on the Economic Union and Development Prospects for Canada (the MacDonald Commission). Toronto: University of Toronto Press. Lomas, Jonathan, (1997), ‘Devolving Authority for Health Care in Canada’s Provinces: 4 Emerging Issues and Prospects’. Canadian Medical Association Journal, Vol. 156 (3), pp. 817–23. Marchildon, Gregory P. (2005), Health Systems in Transition: Canada. Copenhagen: WHO (on behalf of the European Observatory on Health Systems and Policies). Marquis, M. S. (1992), ‘Adverse selection with a multiple choice among health insurance plans: A simulation analysis’, Journal of Health Economics, Vol. 11, pp. 125–53. McGuire, Thomas (2000), ‘Physician agency’, in Culyer, Anthony J., and Joseph P. Newhouse (eds), Handbook of Health Economics. Amsterdam: North-Holland. Ch. 9, pp. 461–536. Morris L. Barer, Robert G. Evans, and Greg L. Stoddart (1979), Controlling Health Care Costs by Direct Charges to Patients: Snare or Delusion?. Toronto: Ontario Economic Council.
Åke Blomqvist 155 Newhouse, Joseph P. and the Insurance Experiment Group (1993), Free For All? Lessons from the Rand Health Insurance Experiment. Cambridge, MA: Harvard University Press. Naylor, C. David (1986), Private Practice, Public Payment: Canadian Medicine and the Politics of Health Insurance, 1911–66. Montreal: McGill-Queen’s University Press. Robinson, Ray (1999), Health Systems in Transition: United Kingdom. Copenhagen: WHO (on behalf of the European Observatory on Health Systems and Policies). Stuart, Bruce, et al. (2005), ‘Riding the roller-coaster: The ups and downs in outof-pocket spending under the standard Medicare drug benefit’, Health Affairs, Vol. 24, pp. 1022–32. Taylor, Malcolm G. (1987), Health Insurance and Canadian Public Policy: The Seven Decisions that Created the Canadian Health Care System. Montreal: McGillQueen’s University Press, second edition. Tobin, James (1970), ‘On limiting the domain of inequality’, Journal of Law and Economics, Vol. 13, pp. 263–77. Woolhandler, Steffie, Terry Campbell, and David U. Himmelstein (2003), ‘Cost of health care administration in the United States and Canada’, New England Journal of Medicine, Vol. 349, pp. 768–775.
6 Tax Incentives for Owner-Occupied Housing – Then and Now Finn Poschmann
Introduction and overview David Laidler’s name is not typically connected to the public finance literature. Yet the published version of David’s dissertation, which appeared in the 1969 Brookings Institution volume, The Taxation of Income from Capital edited by Arnold Harberger and Martin Bailey, bore the title ‘Income Tax Incentives for Owner-Occupied Housing’ (Laidler 1969). David’s thesis included an empirical analysis of the welfare effects on the exclusion of imputed rent (from owner-occupied housing) from the US federal income tax code’s definition of taxable income. In other words, David took on some normative and positive welfare economics issues, squarely in the tradition of Harberger triangles,1 and the volume in which his work appeared became something of a touchstone in the public economics literature of the 1970s and onwards. This brief essay provides an overview of David’s chapter, reports on subsequent work related to the demand for housing, including some relevant updates and caveats, and situates the findings in the more general area of the US tax reform debate – a political debate that, for the moment, is in stasis, owing mostly to tax incentives for owner-occupied housing.
Tax incentives for owner-occupied housing – issues in welfare measurement The flow of service income derived from the ownership of durable goods accrues tax-free to the goods’ owners.2 The most important durable good and the largest asset owned in a typical household is of course 156
Finn Poschmann 157
the home itself, ownership of which is further encouraged by three other tax incentives in the US federal income tax3: favourable treatment (exemption) of capital gains derived from the sale of a principal residence, income tax deductibility of property taxes paid to other levels of government, and income tax deductibility of mortgage interest paid on owner-occupied homes. Laidler (1969) was concerned with only one of these categories, the income-tax exclusion of homeowners’ imputed rent, because at the time (and since), discussions of US tax reforms focused almost exclusively on the mortgage interest deductibility question. This focus implied less attention to what might be thought of as the source of the problem, the initial fact of the non-taxation of imputed rents, at least from the point of view of analysts who favour comprehensive income definitions.4 Laidler sought to measure the degree to which the progressivity of the personal income tax was lessened by the income-tax exclusion, and the associated economic welfare loss following from over-investment in housing stock that the tax subsidy stimulated. Doing so necessitates having estimates in hand for the price elasticity and income elasticity of demand for housing, generating which was the main purpose of the empirical portion of his chapter, relying on census tract data from Chicago, Philadelphia, Los Angeles, and San Francisco. Before turning to those estimates, here is a brief discussion of the welfare model within which they are to be set. Laidler’s assumptions were the existence of a cardinal utility function, constant marginal utility of income over the relevant range, and that income’s marginal utility was the same for all consumers. Provided also that interpersonal comparisons of utility are permitted, this assumption set allows calculation of a welfare-loss triangle – based on a downward-shifted supply curve wherein the downward shift is equal to the price impact of the marketdistorting tax subsidy – that can be aggregated from the individual to the housing market at large. Plotting a supply curve, however, also requires some notion of the marginal private and social costs of housing, which Laidler derived through a simple set of estimates and assumptions about the rate of return on housing capital (a figure near to private long-term mortgage rates – 6%, then as now), depreciation and maintenance, and state and local taxes, all expressed as a percentage of home value, and which summed to 11 per cent.5 The downward shift of the supply curve, the value of the subsidy, is equal to the (imputed) non-taxed return to housing ownership, which also would be 11 per cent, less the non-deductible portion of housing costs – maintenance and depreciation, for a net
158
Tax Incentives for Owner-Occupied Housing
7.5/11 in this estimate.6 Thus the marginal subsidy to home ownership would be 7.5/11 times the owner’s marginal tax rate, which tends to be correlated with income, hence generating the downward influence on tax progressivity mentioned above – the marginal value of the subsidy tends to be positively correlated with income. Now, the overall welfare implication of the subsidy, the overinvestment in housing, depends on consumers’ responsiveness to it, which requires estimates of price and income elasticities of housing demand, and some thoughtful manipulation. As income rises, so too does housing demand, but the relative price of housing falls, owing to the rising subsidy on the margin. Capturing the subsidy alone (for owner-occupied housing) would therefore require knocking out the income effect, which Laidler accomplished by estimating demand elasticities for rental and for owned-housing separately, based on the assumption that in the absence of a subsidy to rental housing, the elasticity of demand for rental housing would capture only the income effect. The difference between the two therefore captures the subsidy’s impact, a useful estimate to have because of the lack of a counter-factual wherein home ownership is not provided a tax subsidy. Laidler pursued these calculations for the purpose of establishing that owners’ income elasticities were indeed higher than renters’, indicating the presence of a subsidy that influenced homeowners housing expenditure. The core empirical exercise, based on the 1960 Census of Housing, posed a number of challenges, owing particularly to the number of variables (such as income) that were not ventilated by tenure status at the census tract level. A series of practical adjustments and some heroic assumptions permitted Laidler to arrive at 1.427 as representative of the income elasticity of housing demand, which he deemed consistent if conservative relative to prior estimates.7 When including some problematic San Francisco data, at −3.057 his estimate for the price elasticity of demand was higher than some earlier estimates, and Laidler elected to use instead the figure −1.5, which he described as ‘a figure firmly grounded in empirical evidence’. Based on housing stock figures from the Census of Housing, Laidler calculated aggregate over-investment in housing as a proportion of existing stock in 1960 (his answer: 17.8 per cent) and thence an annual flow of welfare loss – the subsidy rate times the implicit return to the over-investment, divided by two (because the welfare cost is a triangle, half the rectangle measured by the subsidy times the over-investment), yielding an answer just over $500 million per year.
Finn Poschmann 159
Now, in a market the size of the US, $500 million was not a large number, but it did represent an annual flow, and it was a large number compared with things US citizens might otherwise choose to spend money on in pursuit of public purposes, or in pursuit of at least weakly Pareto optimal tax reforms. And, it should not for a moment be forgotten, the deadweight loss represented a resource wasted in pursuit of subsidizing larger home purchases or more consumption spending on the part of households better able than others to themselves fund such purchases.
Some subsequent findings and caveats Dispensing first with positive aspects, I have not found subsequent results suggesting that Laidler’s income and housing price elasticity estimates were generally off the mark. What is more challenging, however, is the normative debate over the definition of taxable income and the neutral benchmark. The 1970s saw a newly re-invigorated debate over what ought constitute the measure of income and, for example, Feldstein’s (1976) essay constituted a shrewd and aggressive counterpoint to, for another example, the Carter Commission’s (Canada 1969) not-terribly-well-supported claim that ‘a buck is a buck is a buck’. The new reformers saw consumption, not income, as the proper base for taxation, primarily for efficiency reasons. By lightening or eliminating taxes on capital income, the income tax’s over-taxation of savings could be avoided, and the concomitant rise in productive investment would put the economy on a permanently higher level. For these reasons, the fundamental tax reformers would have supported the notion of taxing the consumption value associated with the flow of services provided by durable goods. Notwithstanding a shift in viewpoint, from seeking accurately to measure the non-taxed imputed rent associated with owner-occupied housing, to seeking to capture the value of housing services consumed, analysts still needed to deal with measuring the value of non-traded housing services, and this posed some conceptual quandaries. After all, whether we are trying to get at a comprehensive measure of income of the sort that Gladstone understood many decades before,8 or if we are trying to achieve a more efficiency-oriented consumption base, we are still confronted with the question of ‘why the market rent for the dwelling should be used to measure the homeowner’s implicit income [or consumption], and even less why [a homeowner] should be taxed on that basis’, as Thalmann (2005) bluntly put it.
160
Tax Incentives for Owner-Occupied Housing
This is not necessarily the debate over the normative case for taxing the flow of housing services in general; although there certainly are many doubts on that score, it is the question of whether market rent is the appropriate neutral benchmark. Thalmann (2005) commends Laidler for suggesting a model wherein the implicit return to housing stock was the correct measure, rather than seeking to impute a market rent against which supply costs would be deducted, but goes on to argue that because (in Thalmann’s view) the correct target is tenure neutrality with respect to taxation, the tenure status that agents would prefer before taxes should also be the tenure status they prefer after taxes.9 Yet this is to get a little ahead of things.10 Laidler’s model was useful because it offered a simple and immediate way to use Harberger triangles to measure excess burden. It seems, however, that his calculation and use of demand elasticities could not take account of the fact that changes in taxes would themselves have an impact on effective housing prices.11 And while Laidler did not claim to present a full welfare model, we should note that tax changes would also have an impact on tenure choices and probably savings rates, with at least some welfare implications. One way to think about those implications is to observe that while David drew the supply curve for housing as flat over the relevant range (which some readers may find contentious), the demand curve might have steeper and flatter parts, or kinks as some papers have suggested. Then these wriggles in the demand curve would wriggle differently if the tax subsidy to housing was reduced, in which case the Harberger triangle might have exotic characteristics. There are other questions related to consumer responses to housing subsidies that I find intriguing, such as the concept of over-investment. Many analysts have argued that there is no such thing as over-investment, because the benefits of the tax subsidy to home ownership – whether we are talking about the income exclusion of imputed rent or its mirror twin, mortgage interest deductibility – are fully capitalized in housing (land) prices (Capozza, Green and Hendershott 1999). This is an appealing political argument against the tax-base reformers, because it would suggest the subsidies’ economic distortions were small and that the impact of reducing them would be merely redistribution from homeowners to renters, not a politically promising mission. However, given that I believe that the relevant supply is indeed quite responsive to price, it should not be surprising that I find the full capitalization argument unconvincing. Sinai (1997) suggests that capitalization of about 14 per cent of the subsidy is a more plausible number, and more consistent with observations about the large size of houses in
Finn Poschmann 161
the US relative to similar markets without housing subsidies (Cox 2005). In the end, proponents should be glad if such low number were correct because, were tax benefits fully capitalized and real resource allocation not distorted, then the presumed externalities and other social benefits of subsidies that promote home ownership, which I do not canvas here, could not possibly follow from such subsidies (Bourassa and Grigsby 2000). Another important observation about the US housing subsidy is that while physical investment is distorted, so too are there impacts on household financial asset composition. Mortgage interest deductibility is the proximate cause here, because it encourages households to carry more debt, which current commentators tend to assume finances only more housing investment or more spending on consumer durables or current consumption. This view presupposes, however, that households do not also choose to invest their leveraged wealth in financial assets – and of course they do. Now consider the implications of reducing the mortgage subsidy: households would choose to boost their equity investment in real estate, even if their total investment in residential housing was less than it otherwise would be. Less over-investment in housing would actually result in declining diversification of household asset portfolios.12
Tax concessions for housing and the current tax reform debate As I implied above, the fundamental tax reform mission, base-broadening and rate-lowering in particular, has always been intimately entwined with the quandary over the suite of housing tax incentives that is surely the US’ ultimate middle-class boon. These incentives’ origin is interesting and pertinent, because it is mostly a historical accident, offering hope that it can one day perhaps be undone. The US income tax, as in so many other countries, was driven by the prospective need for war finance, and was conceived as a simple tax on net money income received by a relatively small share of high-income households. Although many practical reasons might have militated against attempting to tax imputed incomes in that context, the exclusion was probably not entirely accidental: Baer (1975) reports that Civil War tax legislation explicitly mentioned the exclusion of owneroccupants’ imputed rent. Mortgage interest deductibility was another matter. Tax rates were low, taxpayers were few, and if they had interest payments, they almost certainly represented business expenses, hence
162
Tax Incentives for Owner-Occupied Housing
there was no strong reason to distinguish interest payments by type (Jackson 2005). With no malice aforethought, therefore, did mortgage interest become tax deductible. Until the Reagan reforms of 1986, there were no limits on the dollar value of deductible mortgage interest, or on the number of residential properties to which it might apply. The 1986 changes and subsequent ones capped the size of mortgages and the number of residences that could generate tax-deductible interest payments, and limited the extent to which deductible financing could finance home improvement or general consumer debt. Nonetheless, the mortgage interest deduction continues to generate tax savings exceeding $80 billion in 2006 (US Congress 2005), leaving the preference a continuing target for base-broadening reformers. Few initiatives have made serious headway, always owing to political opposition, as clearly described by Bartlett (2001). On this point, note that the last major reform effort, George Bush’s President’s Advisory Panel on Tax Reform (2005) was charged in its executive order with pursuing simple, fair, growth-oriented reforms, ‘while recognizing the importance of home ownership and charity in American society’.13 This was clearly an elliptical warning to fundamental reformers that a serious cut at housing subsidies would be perceived as a politically unacceptable attack on homeowning voters.14 This being the case, it is perhaps surprising that the Panel did make an effort at reining in housing preferences and shrinking their attraction to future homeowners. The Panel recommended quite swiftly phasing in lower limits on the size of principal eligible for interest deductibility, switching to caps determined by average regional housing prices after three years and converting from deductibility to a 15 per cent credit (against tax otherwise payable) after four. The Panel was intent on this reform, including it in both the modest and the more-sweeping packages they put forward. The Panel’s recommendations, only ten months old at the time of writing, seem quite dead, and this I conjecture is for three reasons: 1 Former President Bush retained no political capital toe put to use in reform efforts, and Congressional elections loomed; 2 the Panel put forward two options, as just mentioned, leading to confusion over what reform aspects the Panel thought worth seriously pursuing; and 3 the Panel put forward what was almost universally portrayed as a proposal to swiftly end mortgage deductibility (which it was), which left the report with few defenders in any public political forum.
Finn Poschmann 163
In my view, a plausible scheme would be one as put forward by Bourassa and Grigsby (2000), who would announce a distant future date for eliminating mortgage deductibility. In so doing, current mortgages would run their course without mortgage interest payments being exposed to taxation, and new mortgages would enjoy a slowly shrinking time window during which their servicing costs would be favourably taxed. This would quite thoroughly mute the transitional real estate market impacts of the reform, and current taxpayers would perceive no current financial losses. Bourassa and Grigsby would therefore exclude imputed rents from taxable income, primarily for practical reasons15 (see Table 6.1). As to capital gains and real property taxes, most analysts lean towards the Canadian approach, which would not tax gains on principal residences, nor would it permit local tax deductibility. Table 6.1
Pros and cons of retaining current US tax concessions
Concession
Pros
Non-taxation A tax on wealth of imputed net would effectively income be created Problems would arise with setting a single tax rate due to variations in rates of return across space and time Assessment standards vary across jurisdictions Additional complexity of the tax system would be onerous Non-taxation Much of the of capital gains nominal gain is due to general inflation If gains are taxed, then losses should be deductible, thus reducing revenues Additional complexity of the tax system would be onerous
Cons
Overall assessment
Inefficiencies in allocation of capital investments result
Pros outweigh cons: Administrative infeasibility of accurately taxing net imputed income makes it undesirable
Inefficiencies in allocation of capital investments may result
Pros outweigh cons: Efficiency gains do not appear to justify high administrative costs and small revenue gains
(Continued)
164
Tax Incentives for Owner-Occupied Housing
Table 6.1
Continued
Concession
Pros
Cons
Overall assessment
Mortgage interest deduction
The homeownership rate rises The amount of housing constructed increases, thus preventing a housing shortage The capital value of existing homes is maintained
At best, it has only a marginal impact on the rate ofhomeownership (and may benefit only those who would become homeowners anyway, but at a later date) It is unlikely to have a big impact on the number of new starts Capital values could beprotected by phasing in tax reform Higher-income households receive undue benefits Renters are treated inequitably It complicates the tax system unnecessarily It encourages urban sprawl
Cons outweigh pros: Benefits of the deduction are negligible
Real estate tax deduction
The real estate tax is analogous to a very high sales tax, which calls for offsetting treatment in the federal tax code It is often unrelated to ability to pay It is imperfectly assessed It is regressive with respect to household income
Housing deductions are themselves regressive, meaning that they do not offset the regressivity of the real-estate tax The real-estate tax deduction would become more regressive after the mortgage interest deduction is eliminated
Cons outweigh pros: Benefits of the deduction appear to be the opposite of what is intended
Source: Bourassa and Grigsby (2000).
Finn Poschmann 165
With respect to the core options, however, I perceive no constituency in the US for taxing imputed incomes. The alternative is the phase-in for ending mortgage deductibility, as I just described, and it does not seem likely that such a proposal could succeed in a budget any sooner than fall 2009; a likely terminal date for mortgage deductibility would then be 2034, or 70 years after Laidler wrote his dissertation.
Concluding notes Laidler’s tax incentives paper offered a tidy application of evolving models of deadweight loss in an empirical setting, in an area of continuing policy controversy. The fact that he could arrive at a readily comprehensible numeric result, while clearly explaining how he arrived at that result, makes his piece a superb model for contemporary policy analysis. It might give one brief pause to note that the numbers he arrived at were almost certainly wrong, owing to the manner in which demand shifts under a changing set of changing incentives along intensive and extensive margins. That, however, is in the nature of the game, so to speak, and the process – of building simple models and thereafter refining them to reflect improved understandings of economic processes and the interactions between tax and behaviour – is crucial to policy development. And on that score, I think David can take some pride in pointing to work that, owing to a confluence of happenstance, effort, and sound thinking, remains a reference piece in the central policy debate in the US tax environment, some four decades on.
Notes 1. He was attracted to the project by the prospect of working with several superb intellects, including Robert E. Lucas and Eitan Berglas. David’s supervisors, other than Harberger and Bailey, were Margaret Reid and Richard Muth, the latter having established a substantial body of work related to demand functions, household production and the housing market. 2. There are pertinent examples of exceptions to the rule, as highlighted below. 3. And, in many cases, state income taxes. 4. Personal communication. 5. For reasons not much developed in the text, Laidler (1969) also took 11 per cent to represent the marginal social cost of housing supply, which was probably quite reasonable.
166
Tax Incentives for Owner-Occupied Housing
6. Laidler points out in personal communication that this form is quite flexible: with appropriate signs changed, a durable good whose imputed rate of return is subject to a tax subsidy is analytically identical to a stock of real balances that is subject to an inflation tax. 7. Such as Muth (1960); this was Muth’s dissertation, also prepared under Al Harberger’s supervision. 8. And, as was reflected at the time in English tax law’s inclusion of imputed rent in taxable income. 9. A key feature of Thalmann’s paper is his focus on taxes paid by landlords, an issue often forgotten in these discussions. 10. The following draws on Platania (2001). 11. Rosen (1979) brought taxes explicitly into prices, in the view of Platania (2001), avoiding some of these issues. 12. At the broader capital market level, the supply and demand for mortgages and mortgage-backed securities would decline, with uncertain – to me at least – effects on pre-tax interest rates. It is interesting to contemplate the extent to which the monstrous size of the US mortgage-backed securities sector is the product of households’ riding the tax wedge between borrowers’ tax-deductible mortgage payments and tax-free mortgage interest arriving in non-taxable accounts and institutions. 13. See http://www.taxreformpanel.gov/executive-order.shtml. 14. Note also the implicit assumption, not grounded in any durable evidence (Rosen 1989), that housing preferences have a meaningful influence on the rate of home ownership. 15. Which might seem odd, given that they provide a long list of countries that do tax or have in the recent past taxed some measure of imputed rents (525–7); they also note that Wisconsin taxed imputed rent between 1911 and 1917.
References Baer, William C. 1975. ‘On the Making of Perfect and Beautiful Social Programs’, The Public Interest, 39: 80–98. Bartlett, Bruce. 2001. ‘Tax Reform’s Third Rail: Mortgage Interest’. Austin: National Center for Policy Analysis. Bourassa, Steven C., and William G. Grigsby. 2000. ‘Income Tax Concessions for Owner- Occupied Housing’, Housing Policy Debate, Volume 11, Issue 3. Washington, DC: Fannie Mae Foundation, 521–46. Canada. 1969. Report of the Royal Commission on … (Carter Commission). Ottawa: The Queen’s Printer. Capozza, Dennis R., Richard Green, and P. Hendershott. 1999. ‘Taxes and House Prices: Large or Small Effect?’ Proceedings of the 91st Annual Conference of the National Tax Association, 19–24. Cox, Wendell. 2005. ‘International House Sizes’. Available online at http://www. demographia.com/db-intlhouse.htm. Feldstein, Martin. 1976. ‘On the Theory of Tax Reform’, Journal of Public Economics, 6(1–2): 77–104. Jackson, Pamela J. 2005. ‘Fundamental Tax Reform: Options for the Mortgage Interest Deduction’, CRS Report for Congress RL33025. Washington: Congressional Research Service.
Finn Poschmann 167 Laidler, David. 1969. ‘Income Tax Incentives for Owner-Occupied Housing’, in Taxation of Income from Capital, Arnold C. Harberger and Martin J. Bailey (eds), Washington, DC: Brookings Institution. 50–76. Muth, Richard C. 1960. ‘The Demand for Non-Farm-Housing’, in The Demand for Durable Goods, Arnold C. Harberger (ed.), Chicago: University of Chicago Press. 29–96. Platania, Jennifer M. 2001. ‘The Welfare Implications of the Tax Benefit to Homeownership’, Iowa State University, Federal Reserve Board, University of Central Florida, Elon University. President’s Advisory Panel on Tax Reform. 2005. Final Report. Washington: Government Printing Office. http://www.taxreformpanel.gov/final-report/. Rosen, Kenneth T. 1989. ‘The Mortgage Interest Tax Deduction and Homeownership’. Fisher Center for Real Estate and Urban Economics Monograph 1, University of California, Berkeley. Rosen, Harvey S. 1979. ‘Housing Decisions and the U.S. Income Tax: An Econometric Analysis’, Journal of Public Economics 11: 1–23. Sinai, Todd. 1997. ‘Are Tax Reforms Capitalized into House Prices?’ Department of Public Policy and Management Working Paper. Philadelphia: University of Pennsylvania. Thalmann, Philippe. 2005. ‘Equity and Neutrality in Housing Taxation.’ FAME Research Paper No. 147. Available at SSRN: http://ssrn.com/abstract=770605. US Congress. 2005. ‘Estimates of Federal Tax Expenditures for Fiscal Years 2005 to 2009’. Joint Committee on Taxation JCS-1-05. Washington: Government Printing Office.
Discussion John P. Palmer
Finn Poschmann presents a summary of David Laidler’s work on the taxation of imputed income received by homeowners. In the process, he shows that Laidler’s article, which Poschmann aptly describes as a ‘touchstone’ piece, has formed the basis for much of the continued debate and discussion of the issues surrounding tax reform and housing in the US and elsewhere. Like many economists, I was mildly surprised to learn of this work by Laidler, who is best known for his work on monetary economics. And yet we should not have been so surprised: his textbook, Introduction to Microeconomics, was a leading textbook in the field, going through four editions and having been translated into Spanish, Polish, Italian, and Bulgarian.1 Also, as David said during the Festschrift, there is much in common between money and housing since both are stocks and both are affected by rising price levels.2 Studying the effects of tax policy on housing markets would have been much more complex than most of us realize, given that Laidler’s work was done in the early 1960s. The statistical work in his study was not, after all, based on a bunch of data thrown into a spreadsheet. Rather, the data were painstakingly collected by hand, and the calculations were carried out on a then state-of-the-art hand-crank desk calculator. Including more than two or three variables in an ordinary least-squares regression meant spending hours and hours inverting a matrix by hand to estimate the coefficients and their standard errors. Those were the days when people spent much more time thinking about their theoretical model and their testable hypotheses because re-running a regression sometimes took months, not seconds. In 1979, Canada flirted briefly with enacting mortgage interest deductibility as part of the income tax code. Several of us at the time 168
Discussion
169
bought houses in anticipation of receiving both the capital gain and the annual subsidy to owner-occupied housing. It was clear from our actions that the tax treatment of housing would and did have an impact on the demand for housing. Alas for us housing speculators, personally, mortgage interest deductibility never became law in Canada, even though its absence is probably a benefit for the Canadian economy on efficiency grounds. At the same time, though, because property taxes are not deductible from income taxes in Canada (unlike the US, where they are deductible), Canadians are at least approximately taxed on the imputed income received on owner-occupied dwellings via the property tax. So given that Canada does not have mortgage interest deductibility (but the US does) and given that Canada does not have property tax deductibility (but the US does) it is pretty clear that the Canadian tax treatment of housing is much more likely to promote allocative efficiency than is the US tax regime.3 Unlike Poschmann, I am not so readily convinced that the assumption of a horizontal supply curve for housing is reasonable. Resources are scarce, especially land for urban housing. Increasing the demand for housing in the large, dense urban markets that Laidler studied surely bumps up against land and zoning constraints, thus forcing housing prices higher. These effects on the supply side of the market tend to be either ignored or assumed away too readily in much of the work referenced by Poschmann. But if the supply effects are serious, then increasing the demand for housing does not do so much damage to resource allocation; indeed, if the supply curve is vertical, the increased demand becomes readily converted into rents (in the economic sense) to land owners, and the distortions in the economy will be small. And there is some pretty strong evidence that all the subsidies to homeownership and owner-occupied housing in the US have had little impact on overall homeownership. Although the deduction for home mortgage interest is often justified on the grounds that it is necessary for promoting home ownership, it is unclear to what extent rates of home ownership depend on the subsidy. According to the Census Bureau, there are more than 123 million homes in America, with a home ownership rate of 69 percent. There are many countries that do not allow any home mortgage interest deductions for tax purposes, including the United Kingdom, Canada, and Australia. The rate of home ownership in the United States is higher than that in some countries (approximately
170
Tax Incentives for Owner-Occupied Housing
66 percent in Canada), lower than that in others (approximately 70 percent in Australia), and comparable to that in still others (the United Kingdom). Thus, it appears that the level of subsidies provided in the United States may not be necessary to ensure high rates of home ownership.4 Even if the overall homeownership rates are not substantially different across these countries, however, it is quite likely that there is still an over-investment in housing in the US to the extent that people there build and buy larger homes with more land than they would buy, were it not for these tax write-offs. And that raises the sticky question of the transition. Once government policies have been implemented with the effect (intended or otherwise) of generating rents, it becomes increasingly difficult to eliminate these programmes without causing a political backlash which is more than a little daunting for most politicians. The result is that even a proposal to gradually phase out mortgage interest deductibility as suggested by Poschmann, or to gradually phase out property tax deductibility (which would be more in keeping with taxing the imputed rents to owneroccupied housing) is not likely to make its way into the tax code.
Notes 1. Cf. ‘David Laidler’: http://en.wikipedia.org/wiki/David_Laidler. 2. Also see n.6 in Poschmann’s article. 3. There is one minor exception in Canada, and it is probably also efficiencyenhancing: individuals may write off a portion of their housing expense to the extent that they use their homes to generate private business income. Also, keep in mind that these efficiency arguments are all based on partial, not general, equilibrium analysis. 4. President’s Advisory Panel 2005, Chapter 5, p. 72.
References Laidler, David. 1969. ‘Income Tax Incentives for Owner-Occupied Housing’, in Taxation of Income from Capital, Arnold C. Harberger and Martin J. Bailey (eds). Washington, DC: Brookings Institution. 50–76. Poschmann, Finn. 2006. ‘Tax Incentives for Owner-Occupied Housing – Then and Now’, this volume. President’s Advisory Panel on Tax Reform. 2005. Final Report. Washington: Government Printing Office. http://www.taxreformpanel.gov/final-report/.
7 The Political Economy of Inflation Targets: New Zealand and the UK C. A. E. Goodhart
7.1 The Reserve Bank of New Zealand Act, 1989 One of the several occasions on which my path, as a monetary economist, crossed that of David Laidler relates to the adoption in New Zealand of an inflation target, which was effectively introduced in 1988 and then confirmed by the Act of 1989. There is now a detailed record of the events surrounding the passage of this Act, in Singleton, Grimes, Hawke and Holmes, ‘The Reserve Bank of New Zealand Act, 1989’, (2006), and a somewhat more wide-ranging paper by the same authors, ‘Twenty years of modernisation: The Reserve Bank of New Zealand’ (2005). In this latter paper, the authors note that ‘key reforms have often been tested on leading thinkers internationally before implementation. (For instance, aspects of the reforms contained in the 1989 Reserve Bank Act were run past each of Charles Goodhart, David Laidler and Alan [sic] Meltzer)’. In the rather more detailed account of the passage of the Act (2006, forthcoming), the authors note, (fn. 54) that [c]ommentary was also sought from Professor David Laidler (a highly respected ‘monetarist’, then at University of Western Ontario), who was unavailable. However the Bank had discussed relevant issues with Professor Laidler the previous year. The Bank’s file notes record Professor Laidler acknowledging that with the existing state of knowledge and changing relationships between money, GDP and prices, binding monetary policy rules were not feasible; the only alternative was discretionary monetary policy combined with close public monitoring. The Bank’s proposed approach was consistent with this view. The Bank later discussed the issues with another 171
172
The Political Economy of Inflation Targets
prominent visiting monetary economist, Professor Allan Meltzer (Carnegie-Mellon University). A subsequent Bank response to a Parliamentary Select Committee notes that Professor Meltzer was supportive of the monetary policy institutional approach proposed by the Bank. I was more fortunate than David on this occasion, since I was able to give more time to participating in the 1987/88 discussions in New Zealand in advance of the passage of the Act. I shall only discuss this fairly briefly, in some large part because it has been admirably recorded in Singleton et al. (2006), before turning to the main subject of this chapter, which is the adoption of both Inflation Targeting (1992/93) and Operational Independence for the Bank of England (1997) in the UK. One of the most interesting facets of the 1989 Reserve Bank of New Zealand (RBNZ) Act is that one of the main motives for it did not come from monetary policy or monetary analysis at all. Instead, intense dissatisfaction had developed with the intervention, meddling and direct (micro) management with all aspects of the economy by the previous (National) government, led by Sir Robert Muldoon. This was, perhaps, particularly marked in the public sector industries, and was almost total with respect to the RBNZ, which had been historically a subservient central bank. Muldoon acted as his own Minister of Finance; ‘He implemented a system of widespread interest controls, coupled with a wage-price freeze and reverted to a fixed exchange rate. The Bank advised against adoption of these policies, favouring a shift to less interventionist policies with greater reliance on interest rate flexibility, but was overruled by the Minister’ (Singleton et al. 2005, p. 5). Another interesting feature of the New Zealand scene was the economic role reversal of the two main parties, with the (supposedly right wing) National party having adopted interventionist, statist, direct controls. So, in opposition, the (supposedly left wing) Labour party espoused more liberal, market-friendly policies. When they came back to power, in 1984, the Prime Minister, David Lange, and, even more so, his Minister of Finance, Roger Douglas, determined not only to dismantle the existing system of direct controls, but to try to construct a new structural system which would prevent the readoption of such direct controls in future. As noted in Singleton et al. (2006, pp. 11–12), [o]fficials were working [in 1983] on desired monetary control mechanisms to respond to these issues. The election [in 1984] took the
C. A. E. Goodhart 173
field of endeavour further, however. The new Minister of Finance, Roger Douglas, verbally communicated to officials that he wished to ‘Muldoon-proof’ future monetary policy, meaning that he wished to implement a new institutional structure for monetary policy that would prevent mis-use of the monetary printing presses. Later, in the 1988 Budget, he explained this objective as ‘to make certain that no future politician can interfere with the Bank’s primary objective of ensuring price stability, or manipulate its operations for their own purposes, without facing the full force of public scrutiny ….’ While there appears to be no formal documentation of the Minister’s initial request to work on this topic, the request to ‘Muldoon-proof’ monetary policy is well-entrenched in the institutional memory of the Bank and the wider policy community. That the request did originate from the Minister is clear from a July 1986 memorandum from the Governor to the Minister stating ‘You have expressed an interest in changing the legal/constitutional position of the Reserve Bank so as to increase the Bank’s autonomy, in order to enhance its ability to promote stable and consistent policies for New Zealand in those areas for which the Bank has particular responsibility.’ There was, however, a problem. As Singleton et al. (2006, pp. 12–14) reported: [a]t the same time as Bank officials were working on new monetary policy regimes, Treasury officials were working on a wide range of public sector reform policies, both for core government departments and for other state entities. A key aspect of these reforms were policies to be applied to State Owned Enterprises (SOEs). Treasury and the State Services Commission (SSC) considered it important that the Bank comply with the rest of the state sector in adhering to the new public sector management ethos, preferably along the newly established SOE model. Through the period of financial deregulation in 1984 and early 1985, Reserve Bank officials concentrated primarily on monetary policy issues rather than on contemplation of new institutional structures along the SOE model. However once the main deregulation policies had been implemented, internal attention shifted towards these broader institutional matters. The Bank received a copy of the 8 May 1985 Treasury report to the Minister of Finance ‘State Owned Enterprises’ which included appendices on ‘Framework for State Owned Enterprises’ and ‘Accountability
174
The Political Economy of Inflation Targets
Procedures for State Owned Enterprises’. This paper did not explicitly mention the Bank. However, it is important for laying the groundwork for what became major issues between the Bank and Treasury in the formulation of the Act. The paper’s framework for SOEs emphasised three main points: clarity in objectives, competitive neutrality and managerial responsibility. It argued that applying these principles to relevant agencies would provide a much clearer basis against which the Government as shareholder can monitor their performance and take appropriate corrective action. Application of the principles entailed inter alia focusing on the principles of the enterprise’s operations, performance targets to be met, and establishment of a clear dividend policy. Once objectives had been determined, the framework envisaged operational decisions being left with managers who would be accountable for results. Treasury and SSC would be responsible for analysing the results and reporting them to Government. One key aspect of the framework, which would foreshadow a long debate between the Bank and Treasury, was its consideration that it was desirable to attack problems of monopoly power and to treat SOEs as far as possible the same as comparable private sector firms. The SOE framework would do this by removing any legislative or regulatory barriers to entry to a market, so that the enterprise faced competition or the threat of competition. The problem is that central banks have developed into monopoly providers of base money. An explanation of why this has happened is given in Laidler (2003/2004, pp. 45–6); also see Goodhart (1988) and Congdon (1996). It was, therefore, difficult to fit the RBNZ into this latter competitive framework. In this context, Roderick Deane, the Deputy Governor, asked Peter Ledingham to write a survey (think-piece) on its institutional structure in the light of the 1985 Treasury note. He came up with three options for monetary policy – (i) free banking (i.e. no monetary policy); (ii) some Government-imposed stable standard of value (i.e. exchange rate or price level or inflation targeting); and (iii) some Government-imposed targets aimed at all the objectives of economic policy – he concluded that only the second made sense. The first was impractical, and the third could lead to different agencies pursuing objectives with different weights. Further, while acknowledging that there may be a short term trade-off between inflation
C. A. E. Goodhart 175
and other objectives, Ledingham concluded that there was little or no long-term trade-off. Thus, for stability, monetary policy should be assigned to the inflation objective. He added: ‘Moreover, proper accountability for the conduct of policy requires a fairly strict assignment of instruments to objectives.’ Singleton, et al. (ibid. p. 17) It was thus Ledingham who gains the credit for insisting that the RBNZ should have a single target for price stability,1 but the question of how that should be made operational was left for discussion; and the need to specify an agreed, quantitative target, which should be wholly, or partly, decided by government, came out of the wider SOE framework. At that point (in November 1986) a major spanner got thrown into the works. An economist at Treasury, Paul Atkinson, came up with an entirely different blueprint, ‘Monetary Reform: An SOE Framework for the Reserve Bank’, the merits of which he managed to persuade his Treasury colleagues. This combined constraining the Bank’s ability to expand the note issue at the same time as encouraging it in other respects to act as a competitive commercial bank, which could be allowed to become bankrupt in the normal way, (if so, the Minister would take over). Most other functions of a central bank, including foreign exchange management, debt management, research and forecasting, and supervision – if the latter was to be retained at all – were to be undertaken in a separate Services Division, directly funded by government, in practice a government department. This was a remarkable throwback to the Ricardian (Monetarist) ideas of the Currency School of the nineteenth century. Indeed, ‘The Treasury proposal envisaged the Bank being set up in 3 departments: an Issue Department, a Banking Department, and a Services Department’; this was almost a carbon copy of the 1844 Bank of England Act. As may be imagined, ‘Internal Reserve Bank reaction to the Treasury document was negative’ (ibid. p. 30). This ‘led to an intense period of debate about the finer points of the autonomy proposals’ (ibid. p. 30). At this time, the Board obtained external views on the Bank’s own proposals for autonomy and on Treasury’s proposals. Commentaries were provided by Sir Frank Holmes (who attended the special meeting in April), Roderick Deane (who attended the special meetings as a consultant) and Charles Goodhart (formerly Advisor to the Governor of the Bank of England and, at the time, Professor at London School of Economics).
176
The Political Economy of Inflation Targets
It was at this point that David was also asked for advice, but was unavailable. One of the main issues, in the Bank/Treasury dispute, was whether the single objective, or target, should relate to measurable, intermediate ‘outputs’, such as the note issue or the monetary base, or to ‘outcomes’ such as the inflation rate. It was, on this front, quite fortunate for the RBNZ position that, formerly apparently stable, established relationships between monetary variables and nominal incomes had collapsed so patently in both the USA and the UK in previous years. Hence the Bank could claim support for its position from (moderate) monetarists (like David) as well as more eclectic economists. After Labour had been returned a second time, in August 1987, Roger Douglas delegated to Peter Neilson, his Associate Minister of Finance, the job of sorting out the fractious discussions between Bank and Treasury. As Singleton et al. (ibid. p. 43), record: Mr Neilson was kept informed of internal discussions during the preparation of the various proposals for the Bank’s restructuring. He showed no signs of accepting the Treasury’s radical restructuring proposals, accepting that money base measures had little information content, especially during deregulation, and had no credibility advantages either in the markets or more widely. He saw the need to retain flexibility regarding monetary policy implementation mechanisms provided there could also be accountability regarding their implementation. Accordingly, he indicated to the Bank (in early 1988) that he wished to work within the broad framework of the Bank’s proposals. However, he also indicated that he was keen to see officials from Treasury and the Bank come to a jointly acceptable position if that were possible. With the Bank’s proposals having then become the centrepiece of the proposed government Bill, attention then turned to details. Among these was the specific nature of the target. As Singleton et al. (ibid. p. 45) note: [a]t a meeting between Bank officials and the Associate Minister in early 1988 (reported to the February 1988 Board), Mr Neilson suggested a number of consequential amendments that shaped the final form of the Act. He wanted more precision, with clear quantification of the final (inflation) target, and possibly also of some intermediate targets. Two months later, on 1 April 1988, the Minister of Finance (Roger Douglas) announced in a television interview that monetary
C. A. E. Goodhart 177
policy was to be directed to reducing inflation to ‘around 0, or 0 to 1 per cent’ over the following couple of years. The object of this statement was to lower inflation expectations by making it clear that he would not settle for inflation of around 5–7% p.a. This type of target was in keeping with the targets sought by Mr Neilson for the new Reserve Bank Act. I played a minor role in this exercise of designing the other details, notably on the question of potential overrides to the quantitative target. I was also quite unhappy that the sole formal sanction, with the Governor being held personally accountable, was to be dismissal for failing to meet the agreed target. This struck me as too much of a nuclear trigger-point, which would often be (felt to be) unfair if rigorously enforced and lack credibility if not rigorously enforced. Instead I argued for a monetary incentive which would relate the Governor’s salary to his/her success in achieving the target, several years in advance of Walsh’s (1995) similar more academic proposal. This almost got into the Bill, but was eventually rejected by Treasury on a Public Relations (PR) argument that it could lead to headlines along the lines of ‘Governor raises own salary by restrictive policies throwing workers out of jobs’. I still think that the use of monetary incentives for public service is a good idea; if it works in the private sector, why not in the public? Having decided on the main structure of the Bill that they wanted to present to Parliament, the next stage for Ministers (David Caygill taking over from Roger Douglas as Minister of Finance) was to open it up for discussion among their own backbenchers (the caucus), and with the Opposition and the public. The main concern of caucus was whether the government was relinquishing too much economic power to unelected officials, a concern that was assuaged by noting that the proposed Act did not give goal independence2 to the RBNZ. The Select Committee of the House of Representatives received 23 submissions from the public (Singleton, et al., 2006, p. 50). The previous RBNZ Act (1964) had required monetary policy to have ‘regard to the desirability of promoting the highest level of production and trade and full employment, and of maintaining a stable internal price level’ (Section 8). The main concern of the submissions from the public related to the single price stability objective, dropping the real output/unemployment objective.3 This was the gist of two submissions from well-known monetary academics (David Sheppard and Jan Whitwell, and Paul Dalziel) and from the Manufacturers Federation (Manfed) (Singleton et al., 2006, pp. 50–2).
178
The Political Economy of Inflation Targets
The latter has, of course, remained a concern to those opposed to (strict) inflation targeting. The response, again of course, is based on the concept of the vertical Phillips curve, in that the medium- or longer-run monetary policy cannot improve real variables (growth, employment and so on) though bad policies (hyperinflation, financial crises and bad deflation), may impair those same variables. Thus the best basis that monetary policy can provide for growth is to maintain price stability. I gave evidence along these lines, rebutting the prior evidence of David Sheppard. The National Party’s leading member on this Committee was Ruth Richardson (the shadow Minister of Finance and later the actual Minister of Finance). She was sympathetic to the Bill, including the priority it gave to price stability, and managed to persuade the National Party caucus to support it, despite the glowering opposition of Muldoon. Following the Select Committee process, the Bill was discussed again, and passed unanimously through all remaining stages, by Parliament in December 1989. (Sir Robert Muldoon was absent from the House due to a prolonged illness.) The Act came into force on 1 February 1990. The Opposition stated it would amend certain minor aspects of the Act if it came to power, but in the event chose not to do so. Singleton et al. (ibid. p. 53) This history is consistent with a frequent feature of a move to operational independence for the Central Bank, in that this is commonly done by the nominally more left-wing party when in government, as also in South Africa and the UK. They receive a larger credibility bonus. Moreover, it is extremely difficult for the more right-wing opposition to disavow such a measure. The resulting near-unanimity then further reinforces the credibility of the overall commitment to the new regime.
7.2 Developments in the UK, 1988–92 One of the more dramatic events during the Thatcher years (Prime Minister for 11 years from 1979 until 1990) was the resignation speech of Nigel Lawson on 31 October 1989. This was the culmination of a long battle over economic and European strategy (retold in Thatcher, The Downing Street Years, and Lawson, The View from No. 11). An entirely unexpected element of that speech was the revelation that Lawson had recommended, in November 1988, that the Bank of England be made ‘independent’.
C. A. E. Goodhart 179
At that time, the Bank continued to be one of the most subservient central banks among developed countries. All the major strategic issues of macro/monetary policy, including setting short-term interest rates, continued to be made by the Chancellor of the Exchequer, in consultation with the Prime Minister. The Bank was forbidden to publish its own, separate, forecasts, since, so it was argued, the only interest of commentators and journalists would be to probe the differences and distinctions between the Bank’s and the Treasury’s forecasts. Indeed, the Bank sent over the key paragraphs of its Quarterly Bulletin to the Treasury in advance for them to vet, suggest revisions, and on rare occasions veto. The Bank was an agent of the Government and an Adviser to the Chancellor, though Treasury officials who worked directly to, and personally with, the Chancellor generally had the inside track in offering advice, on monetary policy as well as on other issues. The Bank’s main strength, in discussion with the Chancellor and HMT (H. M. Treasury) in those days lay in its command over technical and market issues, see Goodhart (2004). Even so, the amount of leverage, in support of its own viewpoint, that the Bank could exert still depended greatly on the personal rapport between the Governor and the Chancellor (and also the Prime Minister). This varied considerably over time. The relationship between Governor Richardson and Chancellor Healey in the 1970s had been close; that between Richardson and Chancellor Howe in the early 1980s was reasonably cordial, although that between Richardson and Prime Minister became frosty; that between Governor Leigh-Pemberton and Chancellor Lawson was correct, but somewhat distant, since Nigel Lawson considered his own abilities as a monetary economist superior to those of the Bank, both Governor and staff. It was symptomatic of his relationship with the Bank that his nascent plans for making the Bank independent were kept secret from (all those in) the Bank, and instead worked up by a small group in HMT (The View from No. 11, p. 868). The proposal to make the Bank independent did not derive from any admiration of the Chancellor for the quality and skill of the Bank; indeed neither the Prime Minister nor the Chancellor were great admirers of the Bank. So what was the point of the exercise? Politicians had seized control of the power to set interest rates from central banks in the 1930s, and even more decisively in World War II. But they had not found the use of such powers to be generally comfortable. In particular, increases in interest rates were, and remain, deeply unpopular with debtors, particularly manufacturers in the
180
The Political Economy of Inflation Targets
company sector; it raised their costs, tended to appreciate exchange rates and lowered demand. The various lobbies for the private company sector were, and remain, continuously vociferous in their support for ‘easy money’. Similarly, house-owners with mortgages, and prospective house-buyers, are directly harmed by interest rate increases. This group, young marrieds, are politically important. By contrast the segment of the population that benefits from higher interest rates, the old and the retired, are poorly organised and are ineffective as a lobbying group. So, raising interest rates was, and remains, a deeply unpopular political act. This was, however, somewhat less of a problem during the Bretton Woods period of pegged, but adjustable, exchange rates. In those years (1946–71) fiscal policy was the instrument of choice in the UK for maintaining the domestic level of demand in equilibrium, while monetary policy (short-term interest rates) was mainly directed at trying to sustain the balance of payments. Thus, interest rates were raised at times of reserve outflow, pressure on the exchange rate and balance of payments crisis (and lowered during periods of calm on the external front). Such external ‘crises’ were very obvious events, and thereby gave politicians an excuse and justification for raising interest rates. That crutch became much less effective after the generalised switch to floating in the early 1970s. Thereafter the justification for raising interest rates had to depend increasingly on the need to prevent worsening domestic inflation. But the link between raising interest rates now, especially as this initially raised producers’ input costs, and dampening inflation in future was less immediately obvious than the link between international interest differentials and capital flows. Moreover there were, at that time in the UK, many who believed that the use of interest rates (monetary policy) to control domestic inflation was misguided. Up to a point, it was argued, monetary policy would have little, or no effect on expenditure decisions, whereas beyond some (unknown) level, increases in interest rates could cause a crisis and collapse. Instead, that line of theory suggested, a better way forward was to apply prices and incomes controls directly. Moreover this could be supported by direct controls over bank lending; these could also be allocated, in favour of manufacturers and exporters, and against services, imports and personal consumption, in such a way as appealed to the basic mercantilist instinct that was, and remains, commonplace. A further twist to this argument was given by Kaldor’s claim that productivity gains were more easily achievable in manufacturing, than
C. A. E. Goodhart 181
in primary production or services, (Verdoorn’s Law), and his consequent proposal for a Selective Employment Tax, see Kaldor (1966). Such prices and income policies were tried repeatedly in the 1970s and proved to be failures. Moreover, such direct controls, including those on bank lending, were increasingly perceived as generating inefficiencies and harming growth. Although such controls were, once again, adopted by Prime Minister Heath in 1972/73, when the new Conservative Government, under Mrs Thatcher, came to power in 1979, they had decisively rejected their use. But that meant that the use of restrictive monetary policy had to operate via interest rate increases, and that remained as politically unpopular as ever. There was, from time to time, a perceived mechanism for trying to defuse that unpopularity. There is a duality between the short-term interest rate ruling in the market and the quantity of base money available to banks. In order for interest rates to be X per cent, the availability of base money has to be Y; and when the central bank makes Y base money available, the interest rate will be X per cent. So the idea was, could not policy be organised so that the central bank controls the volume of base money, essentially its own liabilities, and then the resultant interest rate will be ‘market-determined’; and who could blame the politicians for a ‘market-determined’ rate? This line of reasoning had much appeal for monetarists and right-wing politicians. Indeed, when Volcker and the Fed feared that Congress would not allow interest rates to rise high enough to defeat inflation, they moved in this direction towards control of the ‘non-borrowed monetary base’ in October 1979. On all this, see Bindseil (2004). A problem was that much of the institutional structure of money markets and the banking system had been built on the historic arrangement, whereby the central bank set the official short-term interest rate and accommodated all high-frequency fluctuations in demand for base money. The implicit agreement within the financial system was that this would continue. Reversing the procedure, whereby the base would be held fixed, and short-term interest rates allowed to vary freely, risked the danger of chaotic high-frequency fluctuations in interest rates and disturbances to prior financial relationships, for example, the overdraft system, at least until the institutional structure of the system adjusted, see Foot et al. (1979). Despite this view, strongly held within the Bank of England, there was sufficient impetus for a move towards monetary base control (mbc) for there to be an official enquiry and discussion of the relative advantages and disadvantages of mbc in the UK in the early years
182
The Political Economy of Inflation Targets
of the incoming Conservative Government, see Goodhart (1989, pp. 322–6). As noted (ibid. p. 324), [s]ince the arguments, pro and con, in the United Kingdom depended largely on claims about how banks, and other agents, in the financial system might behave in the hypothetical conditions of a change in the regime to (some version of) mbc, it was not really possible to demonstrate the superiority of one set of arguments over the other. The protagonists on either side in the United Kingdom, who met to discuss it under official auspices, in the improbable venue of Church House in Westminster on 29 September 1980, stuck generally to their prejudices. One argument that did, however, sway some of those in positions of power and influence was that it would be difficult to steer the system clearly though the transitional learning period: ‘thus we in the UK have very little idea of the size of cash balances the banks would wish to hold if we were to move to a system of monetary base control’ (Lawson 1981); moreover the ratio of £M3 to base money was not stable or predictable, so there was ‘little or no point in trying to use the MBC system to control £M3’ (Walters 1986, p. 123). These considerations, combined with the convinced opposition to mbc from the Bank of England, the commercial bankers and the City of London, persuaded the monetarists not to push more strongly for mbc, although remaining unpersuaded of the contrary case, in the early years of the MTFS [Medium Term Financial Strategy], eg in 1980/1/2. Despite their unwillingness on such grounds to move to mbc, and reverse the ordering of causation (whereby the official choice of shortterm interest rates makes the high-powered money base endogenous4), Conservative governments continued to hanker after procedural adjustments which might make the short-term interest rate appear somewhat more market determined. As noted in Goodhart (2004, op. cit., pp. 363–5), there were two extraordinary occasions during these decades when Conservative governments tried to suggest that short-term interest rates instead were market-determined, and not set by the monetary authorities. The first occurred towards the end of 1972 in the context of government negotiations on pay and price controls, and the imposition of a standstill on prices, incomes and dividends in November.
C. A. E. Goodhart 183
The government wanted to avoid the accusation that, when other costs were fixed, it had increased the interest-cost burden on borrowers. Hence it introduced a formula, in October 1972, whereby minimum lending rate would be related to the outcome of the previous weekly Treasury bill tender. So for a time the senior officials in the Bank played out an elaborate charade, in which they decided what minimum lending rate they wanted and then adjusted tender, and market, conditions to achieve that, (aided on occasions by calls for (repayments of) Special Deposits) At least the purpose of the 1972 exercise was patent, i.e. to obfuscate the fact that short-term interest rates were set by the monetary authorities. The next episode, in 1980, was even more confused. It arose in the aftermath of the decision not to proceed further with (immediate) moves to Monetary Base Control (MBC), after the publication and debate on the Green Paper on Monetary Control (1980). One of the hopes of advocates of MBC was that the central bank would set the growth rate of the base, and that (short-term) interest rates would then become ‘market-determined’ As a kind of consolation prize for MBC monetarists (for not getting MBC), the authorities agreed that operations could become somewhat more market-oriented with less reliance on discount window lending; a note on ‘Methods of monetary control’ was issued on 24 November 1980. The authorities would still set interest rates (with a view to hitting their monetary target), but would, it was suggested, disguise what their interest rate objective was, (the unpublished band), and pretend that it was all the market’s doing. Frankly this was confused and silly. In practice, it had little effect, apart from being a pretext for the Bank to introduce some reform and widening of British bill markets, which it wanted to do anyhow for its own purposes. See ‘Monetary control: next steps’, 12 March 1981, and reproduced in BEQB 21 (1981), pp. 38–9. The unpublished bands, etc., never transpired; the authorities went on announcing administered changes in minimum lending rate, and the monetarist overtones in the supposed ‘new methods’ of 1980 rapidly became a dead-letter and forgotten. Despite all the various efforts to pretend that it was the financial market, and not the authorities, that was responsible for determining interest rates, there was, in the event, no escape. The authorities could not dodge taking responsibility; but monetary restraint implied responsibility for thoroughly unpopular actions.
184
The Political Economy of Inflation Targets
One of the major problems is the long (and variable) lag between raising interest rates and its having a negative effect on inflation, which may not occur for over a year; indeed because of its immediate impact on producers’ costs, the initial response may even be positive, (the wellknown price puzzle of Vector Auto Regression (VARs)). If the relationship between interest rates and inflation had been instantaneous, then politicians could justify their increases in interest rates by pointing to concurrent worsening inflation. But, given the lags involved, increases in interest rates should be related to forecast excess inflation (relative to target). Forecasts are, however, notoriously unreliable. It is not easy for a politician to try to justify an immediately unpopular policy by reference to an uncertain (and potentially manipulable) forecast. There was always, therefore, likely to be a tendency for politicians5 to delay taking unpalatable monetary action – raising interest rates – until excess inflation was already apparent in the system, by which time the dynamic of worsening inflationary expectations would already be taking hold. The combination of such lags and the political unpopularity of raising interest rates was a recipe for ‘too little, too late’. The reluctance to bite the (interest rate) bullet was, of course, particularly marked in advance of elections when politicians’ time discount rate rises sharply. This was first captured by W. Nordhaus in his (1975) article on ‘The Political Business Cycle’. A theoretically fancier, and more general version of this same syndrome was subsequently published by Kydland and Prescott in their (1977) article on ‘Rules rather than discretion’, which won them the Nobel Prize in 2004. The concept of ‘time inconsistency’ entered the vocabulary of officials, if not of politicians. Then there was the further suggestion of Barro and Gordon in their 1983 (a and b) articles that politicians not only had short, high discount rates, but also would aim for a level of output (employment) above the long-term sustainable rate, because that too was electorally beneficial in the short run. Although there are doubts whether the accusation that politicians try to cause a temporary (and unsustainable) boom prior to elections can be empirically sustained (Alesina and Roubini 1997), the view that politicians could not be trusted to operate monetary policy in the best long-term interests of the general public began to take hold during the 1980s, not only among academics and commentators, but also among politicians themselves (who could recognise the political pressures preventing interest rate increases in advance of elections or at times of political unpopularity). The credibility of political promises to control inflation was waning.
C. A. E. Goodhart 185
Moreover, the central banks that had been most successful in restraining inflation during the disastrous decade of the 1970s were exactly those who had the greatest independence from politicians in determining interest rates, the Bundesbank, Swiss National Bank and the Federal Reserve System. Since independent central bankers not only had a longer tenure and were not at risk of being removed by democratic election, but also gave a higher weighting to controlling inflation (and less to stimulating output), than politicians (Rogoff 1985), theory would suggest that they should be more credible and more effective in controlling inflation. There is no doubt that it was this line of argument that Lawson (and to some large extent Lamont too, subsequently) was following, and that the Bundesbank was the proposed exemplar for the prospectively independent Bank of England. Indeed, when rebuffed on this attempt by Mrs Thatcher, a major advantage, according to Lawson, for entering into the European Exchange Rate Mechanism was that it directly subjected British monetary policy to Bundesbank control. In his Minute on this matter to Mrs Thatcher, he does not mention New Zealand nor does he mention a direct inflation target. Instead he mentions monetary targets, which the Bank might commit to achieve, (Lawson 1992). There was a problem with this. Monetary targets had been tried already in the UK and the USA in the early 1980s, had proven too uncertain and unstable and had been largely abandoned (Goodhart 1989). In so far as there had been much success in the UK in the later 1980s, this was partly due to the contentious, some thought cosmetic, policy of ‘overfunding’. While the Bundesbank continued to set monetary targets, and its successor the European Central Bank has maintained a medium-term ‘monetary pillar’, it did not strive officiously to hit such targets, whenever it felt that there might be short-term disturbances to the demand-for-money function. The Bundesbank missed its monetary targets as often as it hit them. Neither the Bundesbank’s targets nor the ECB’s monetary pillar were treated as over-riding commitments. Rather they were used as important information variables, to be combined with other sorts of information, for example, on the path of costs and prices, in reaching a balanced overall judgement on how to adjust policy to achieve price stability (Issing et al. 2005). But this would have led to a dilemma in the UK context. If the Bank of England were to be tied to a tight commitment to achieve a monetary target, it would have, as the available evidence suggested, either failed or caused damage to the real economy in the attempt to succeed. But if the Bank were allowed a very loose target, with discretion to ignore the
186
The Political Economy of Inflation Targets
monetary data from time to time, it would have made it insufficiently transparent and accountable, at least for British tastes (though not for Continental preferences). As will be appreciated, this dilemma was resolved with the switch from monetary targets to inflation targets. The above debate on the economic merits of trying to turn the Bank of England into a simulacrum of the Bundesbank never really developed, because Mrs Thatcher rejected it, indeed dismissed the proposal out of hand, on political grounds. She argued, a common argument for her to deploy, against it primarily on the grounds of ‘unripe times’. She believed that a change of institutional structure to bolster the fight against inflation, when that was not currently going well, would be seen as a confession of failure. She noted (The Downing Street Years, p. 706) that It is on the face of it extraordinary that at such a time – November 1988 – Nigel should have sent me a paper proposing an independent Bank of England. My reaction was dismissive. Here we were wrestling with the consequences of his diversion from our tried and tested strategy which had worked so well in the first Parliament; and now we were expected to turn our policy upside down again. I did not believe, as Nigel argued, that it would boost the credibility of the fight against inflation. In fact, as I minuted, ‘it would be seen as an abdication by the Chancellor when he is at his most vulnerable.’ I added that ‘it would be an admission of a failure of resolve on our part.’ I also doubted whether we had people of the right calibre to run such an institution. As I told Nigel when he came in to discuss his paper, I had thought in the late 1970s about having an independent central bank but had come down against it. I considered it more appropriate for federal states. But in any case there could be no question of setting up such a bank now. Inflation would have had to be well down – to say 2 per cent – for two or three years before it could be contemplated.6 But there was another arrow in her quiver. This was the argument that, willy-nilly, a government could not delegate responsibility for inflation to an ‘independent’ central bank. Whatever happened, the government would continue to be held responsible. Thus she wrote (ibid. p. 707): In fact, I do not believe that changing well-tried institutional arrangements generally provides solutions to underlying political problems – and the control of inflation is ultimately a political problem.
C. A. E. Goodhart 187
This issue, of granting independence to the Bank of England, was revived by Norman Lamont7 in the autumn of 1993, but turned down once again by John Major on similar grounds to the above. Thus, Major wrote (The Economy: Rags to Riches, on p. 675, also see p. 153): Norman brought one other idea to me that autumn that was unwelcome: he wanted to grant independence to the Bank of England. I disliked this proposal on democratic grounds, believing that the person responsible for monetary policy should be answerable for it in the House of Commons. I also feared that the culture of an independent bank would ensure that interest rates went up rapidly but fell only slowly. There were other reasons for opposing operational independence.8 A somewhat uncharitable interpretation of the above ‘democratic’ argument is that politicians, especially Prime Ministers, do not like giving up the power to influence the course of interest rates. There are two more general concerns. The first is the same as surfaced in New Zealand, and has occurred from time to time ever since, which is that a central bank primarily dedicated to the achievement of price stability will be too deflationary, underestimate growth potential and be insufficiently prepared to stabilise fluctuations in output. The counter-argument, relying on a vertical (or even forwards-sloping) medium and longer-term Phillips curve, is dismissed by such antagonists on a variety of grounds, for example, hysteresis, or the non-existence of a stable Phillips curve relationship. This argument has some perennial resonance with the public and with left-leaning academics and politicians, especially when growth becomes slow, as in the eurozone in recent years. The second argument is that there are various arms of macro/monetary policy, notably fiscal policy, competition policy, labour market policy, exchange rate policy, as well as monetary policy. So, the argument continues, these should, for the best outcome, all be coordinated into an optimal, integrated whole. Clearly then the best institution to coordinate overall policies is the Ministry of Finance (HMT in the UK) under the aegis of the relevant Minister (Chancellor). This argument was clearly in the self-interest of such a Ministry (HMT), and has been deployed by such officials from time to time. But it had little resonance with Ministers (recall that both Chancellors Lawson and Lamont backed such independence; it were the Prime Ministers, Thatcher and Major, who objected), nor with most academics, commentators and the general public. Moreover the delegation of the
188
The Political Economy of Inflation Targets
decisions on interest rates to an operationally independent central bank has not, in the event, generated significant problems of coordination; there is an exception to this in the eurozone, but here the problem relates essentially to the absence of any central fiscal authority (with fiscal policy being retained by the separate nation states, whereas monetary policy is centralised in the European System of Central Banks). In practice, the monetary authorities assume that the government’s fiscal plans, for public sector expenditures and tax rates will be carried out, while the Treasury assumes that the targeted inflation rate (and associated time path for interest rates) will also be achieved. There have been a few minor concerns; for example in advance of a new Budget when fiscal policy is generally expected to be changed, the central bank may have to assume a continuation of existing policies, since it is usually not made privy to Budget secrets before announcement. Thus, under such circumstances, it may have to make its forecasts on a somewhat implausible basis. But on the whole, the delegation of operational policy to the central bank has caused, as far as can be observed from the outside, minimal problems to the coordination of overall policy (always excepting the particular problems of the eurozone). Following the rejection, by Mrs Thatcher, of Lawson’s proposal to grant more independence to the Bank of England (November 1988), attention then turned to the question of whether the UK should join the ERM. This battle largely took place within the ranks of the Conservative government, pitching Mrs Thatcher and the euro-sceptics against most of her senior Ministers. The Bank was on the sidelines, and not much involved in the crucial discussions. Once the decision to enter had been taken, monetary policy issues became dominated by the tensions between the need on the one hand to maintain interest rates high enough to be consistent with ERM membership, and low enough on the other hand to prevent a worsening economic downturn and to assuage an increasingly large band of euro-sceptic Conservative back-bench Members of Parliament. In this context, there was no call to discuss inflation targeting or (operational) independence for the Bank of England. All of that would change in a flash once the UK was forced to exit from the ERM on (Black) Wednesday, 16 September 1992.
7.3 From an Inflation Target to Operational Independence: The UK, September 1992 to May 1997 When the UK was forced out of the ERM on 16 September 1992, the Conservative Government ‘lost not only credibility, but also a policy’
C. A. E. Goodhart 189
(Lamont, In Office, 1999, p. 274). He went on to note (ibid. pp. 274–5) that The urgent task was immediately to begin the construction of a new policy. I wanted to show the aims of Government policy remained the same although we needed different means for achieving them. The object of policy had always been low inflation and recovery. Many people at that time, not just in the Treasury, thought that there was a great risk that inflation would rise again through the sudden depreciation of sterling. The great danger I foresaw was that outside the ERM, particularly as we were a Government under pressure, decision-making would now be seen as entirely political. One solution would have been to make the Bank of England independent, but I knew the PM would not go that far. The steps I was proposing would ultimately lead to independence for the Bank of England. Thus began a rapid search for a new policy. It had to be done quickly, because politics demanded that, but it also required care. The reconstruction of policy was done step by step. The Prime Minister, John Major, agreed. He wrote (op. cit. p. 667): After Black Wednesday we could not return to the pre-ERM strategy of monetary targets and ‘taking account’ of the exchange rate; it was vital to develop a new strategy. I had a meeting with Sarah Hogg, Norman Lamont and Terry Burns from the Treasury on 22 September at which we discussed the options. Our objective was a low-inflation economy. Norman developed the theme in meetings at the Treasury, while Sarah fed in more ideas from Number 10 and liaised closely with Terry Burns and Treasury officials. Eddie George, the Governor of the Bank of England, which would have a more public role, was closely involved throughout. We agreed to set a target for the rate of inflation. The arguments, and background papers, presented at such meetings have not yet been made available to the public, so we do not know which protagonists and arguments won the day. Moreover, once the decision had been taken, the politicians took personal responsibility for the new development in policy, as is indeed their constitutional duty, rather than emphasising the provenance and prior application of the idea elsewhere.9
190
The Political Economy of Inflation Targets
By that date, however, the idea of setting an inflation target as the anchor for monetary policy had begun to catch on. The New Zealand example had begun to be noted, for example, by Peter Norman in his Economics Notebook article of 6 November 1989, on ‘New Zealand sets an example’, Financial Times, p. 21.10 Rather closer to home, the Canadians had also adopted an inflation target in early 1991 (see J. Crow, Making Money (2002), especially chapter 8 on ‘Creating Inflation Targets’). The experience, so far, in New Zealand had been good, and inflation targetry had a number of clear advantages, especially in comparison with the previous attempt at monetary targeting. The latter was perceived as having failed whereas inflation targeting was novel, and so far appeared to have been reasonably successful. Monetary targets could be, and were (McCallum 2001, Laidler 2003) described as being the same as inflation targets, but with the (considerable) additional noise of residual variations in demand-for-money (velocity) targets. Monetary targets related to statistical abstracts about which the public cared little, and understood less, whereas inflation was not only easily understandable, but also a subject of direct and immediate public concern. The range and variety of possible monetary targets – such as M0, M1, M2, M3, M4 – was not only confusing, but the claims, and counter-claims, of adherents to one, or another, of these aggregates were so strident (and critical) that the general public tended to become dismissive. The range and variety of possible inflation indices is also very wide, but the extent of general overlap appears greater, and the technical discussions of which might be best has never had the virulence of the debates among the monetarists. For all these reasons, an inflation target was much more easily understood, communicable and capable of political presentation, and defence, than a monetary target. Nevertheless when Lamont moved to an Inflation Target in the autumn of 1992, he did not feel that he could, at the same time, just abandon monetary targets. It is now commonly forgotten that he sought to buttress his inflation target with companion monetary targets. Thus, he wrote (op. cit., p. 276): Inflation itself is a lagging indicator of the economy. Inflation may fall, rise or stabilise, but it will not tell you what is going to happen in a year’s time. For that reason it was necessary to define the role of monetary aggregates in policy. Some monetarists would have liked us to have had a series of targets for the money supply,
C. A. E. Goodhart 191
such as M4, broad money (bank deposits plus money in circulation) and M0, narrow money (notes and coins in circulation). I attached considerable weight to M0 and regarded it as proxy for retail spending, but it was only a contemporaneous indicator, not what economists call a ‘leading indicator’. It too did not forecast the future, only the present. Broad money was meant to be a leading indicator, although it had proved somewhat unreliable in the past. I therefore continued to set a target for M0 of 0–4 per cent for 1992/93. For M4 I set what I called ‘monitoring ranges’, but I also indicated that we would take asset prices and inflationary expectations derived from market instruments and house prices into account in interest rate decisions. There was no easy answer. Lamont was, of course, correct that inflation was a lagging indicator, but it is, to say the least, contentious to believe that the appropriate response was to retain monetary targets; instead, the correct response was to develop (and to publish) inflation forecasts, as Svensson (for example, 1997) and others have noted. Of course, the role for monetary variables as inputs into such forecasts of future inflation remains a much debated issue (see, for example, Issing et al. (2005)). Be that as it may, the continuing role that Lamont had prescribed for monetary aggregates soon disappeared from view. There was yet another major operational advantage of inflation targets over monetary targets. Central banks had historically striven to make their short-term interest rate effective in money markets, and to use such interest rate variations to achieve certain outcomes, for exchange rates and foreign exchange reserves under the Gold Standard and Bretton Woods regimes, to achieve price stability under a floating exchange rate regime. Until the emergence of monetarism in 1950s and 1960s, they had paid little attention to the monetary aggregates as important macro variables (as contrasted with their importance as micro variables indicating the condition of the banking sector). Central banks saw such aggregates as the, largely endogenous, outcome of interest rates, incomes, inflation, animal spirits and so on. The whole institutional set-up in financial markets was conditioned on the assumption that central banks set interest rates and accommodated the response of the banking sector.11 So there was a concern among central banks that any tight commitment to a monetary target might require destabilising structural changes, and/or probably not be deliverable. The switch from monetary
192
The Political Economy of Inflation Targets
targets to an inflation target allowed central banks to operate in their standard, traditional fashion. In the UK at least, the Bank of England had been continuously wary, and on a few occasions opposed, to the government’s initiatives towards monetary targetry in its various guises. Now the switch to an inflation target brought the Bank alongside the government as an enthusiastic supporter of the new approach. Yet another virtue of an inflation target, as it had been established in New Zealand, and copied elsewhere, was that it dealt with the ‘democratic deficit’ issue in a neat fashion. The government either set the inflation target unilaterally, or agreed it with the central bank governor, leaving the central bank with just the technical, or operational, task of deciding what interest rate was best to achieve that mandated rate. This protected the central bank because it could claim that its choice was a technical issue in pursuit of a democratically set objective, and, having been a party to the choice of objective, the Minister could hardly criticise the Bank’s decisions (the means to achieve the ends that had been politically determined) again except on tricky, technical grounds. This division between ‘goal’ independence and ‘operational’ independence12 was crucial to dealing with this issue. Unfortunately John Major did not appreciate this argument, despite Lamont’s promptings. Thus Lamont wrote (ibid. pp. 322–5): Another issue about the Bank was more fundamental. I had come to the conclusion that the logical development of my new policy ought to be to make the Bank of England independent. I had not started out as Chancellor in favour of independence, but a number of factors changed my mind. Firstly, my experiences with the Prime Minister over interest rates convinced me that politicians would always want to interfere with interest rate decisions for political reasons. Even Mrs Thatcher, in my limited observation, was not averse to interfering in the decisions of Nigel Lawson. Secondly, I had over the course of the Maastricht negotiations been able to reflect on the performance of those European central banks that were independent. There was no denying that Europe had become a zone of monetary stability and low inflation. While I disagreed with the idea of a European single currency, I also believed that it would only earn its credibility if it were run by central bankers. The worst of all solutions would be a rainbow coalition of European politicians. The US and Japan also had successful independent central banks, although in their case independence was based on a different model. Britain was one of the
C. A. E. Goodhart 193
few countries where interest rate decisions remained firmly in the hands of the politicians. Some people opposed independence because they believed monetary policy ought to be about more than the control of inflation, but I did not. I had believed that low inflation combined with effective supply side policies were the most any Government could do to promote growth. Furthermore, sound money was the right of every citizen. The idea of a currency that held its value ought to be above politics, so entrusting the value of money to the technicians seemed to be an idea that ought to be acceptable to all political parties. The idea of making the Bank of England independent had grown stronger with the passage of time. Nigel Lawson’s resignation speech and Sam Brittan’s articles made a powerful case. I first raised the issue of independence with the PM in September 1991, when I put forward a scheme based on the model of New Zealand, with the Governor of the central bank being given an inflation target. Failure to meet it could result in his dismissal. But the PM was worried about opposition from the tabloids. He did not believe the issue could be separated from the European debate, or that the House of Commons would stop holding the Government accountable for interest rate charges. My proposals meant that Parliament and Ministers would still have the key role of establishing the objective for the Bank. Central bank independence had surprisingly little support in the Treasury. I suspected that some of them too liked having a say in interest rate decisions. Slightly to my surprise, Terry Burns was not at all in favour. All sorts of surprisingly poor arguments were put forward by the Treasury. What would happen to supervision of the banking sector if the Bank were independent? How would there be accountability to Parliament? It was even put to me that no one would want to take on the exposed task of being Governor. None of the arguments was insuperable; some of them were completely unconvincing. But the PM came up with numerous objections. We couldn’t give any hint while Maastricht was going through the House. Wouldn’t the Government be thought to have lost confidence in itself? The PM also said people were frightened how Labour would handle monetary policy and he didn’t want to remove that fear. I said there were some indications that Labour might move in the direction of independence, but the PM wouldn’t budge. Reluctantly I had to forget the idea.
194
The Political Economy of Inflation Targets
But, without Central Bank independence, a purely political commitment to an inflation target would lack credibility, particularly in the context that the Conservative government found itself in 1992 following the failure of its ERM commitment. So there was a patent need to find some alternative basis for establishing some credibility. A number of steps in this direction were then taken, and mostly announced in Lamont’s Mansion House speech of 29 October 1992. These included, as noted in Seldon (Major: A Political Life, 1997, p. 337): A monthly monetary report of the regular Chancellor–Governor of the Bank of England meetings to set out the basis of policy decisions; clear explanations for any changes in interest rates, to reassure the markets about what the government was thinking; a forecasting panel of seven independent economists (dubbed ‘the wise men’) to assess the economy and supplement the Treasury’s forecasts for the economy; the Bank to publish a quarterly assessment of the progress of counter-inflationary policy. Of these the most important contribution to credibility was given by the requirement for the Bank to publish an Inflation Report. As Lamont noted (op. cit. p. 277): I also asked the Bank to produce a quarterly inflation report assessing the Government’s progress in reaching its inflation target. This novel and unprecedented outburst of glasnost would act as a discipline on the Government and make it more difficult for interest rate decisions to be political. Some saw it as leading naturally to Bank of England independence. This was a complete reversal of the prior prohibition on the Bank’s ability to publish its own forecast; moreover, its virtue was to reside in the Bank’s presumed independence of judgement, so the previous practice whereby the Treasury, acting (supposedly) in the name of the Chancellor, vetted the Bank’s Quarterly Bulletin, and asked occasionally for changes, was suspended in the case of the Inflation Report. Sam Brittan noted (Financial Times, 2 November 1992) that Far and away the most important announcement is that the governor of the Bank of England is to provide a regular report in the Bank’s Quarterly Bulletin on progress towards the inflation target. It will be made by the Bank rather than the Treasury in the hope that
C. A. E. Goodhart 195
this will increase public confidence in the objectivity of the analysis, although the chancellor and his Treasury team will still see prepublished drafts. A toenail is thus being put into the bowl marked ‘Independent central bank’. The inspiration has clearly been New Zealand, where the Reserve Bank has a contract with the government to pursue a given rate of inflation which it can only override in carefully pre-stated circumstances. But the decision on setting interest rates remained with the Chancellor, and the Bank could not appear to be indicating what he should do. So the forecast was based on unchanged interest rates, recording the Bank’s view of what would happen to inflation if the Chancellor left interest rates as they were. This was the genesis of the constant interest rates assumption in the inflation forecast. With hindsight, the Bank could have put more weight on the path of forward rates implied by the yield curve, (to which it switched in 2004). But the techniques to estimate such implied forward rates were not then well developed and, with limited credibility and a recent regime change, forward expectations of future short rates would not have been well anchored. In other respects, however, the Bank, under the leadership of the Governor, Eddie George, and its then Chief Economist, Mervyn King, seized the opportunity to introduce a number of well-designed initiatives in the new Inflation Report. In particular, the expected future path of inflation, and of real output, were presented as a probabilistic distribution, a fan chart (and not as a point forecast), red for inflation (nicknamed the ‘river of blood’) and green for output. The Report had enough details to satisfy commentators, but not so much as to conflict with the Treasury’s more comprehensive forecast. It was widely applauded, and this innovation became copied in almost all inflation targeting countries, and some that were not. The opportunity to publish the Inflation Report gave the Bank a form of presentational independence, since it could indicate whether the existing level of interest rates would, in its view, deliver the inflation target. And that gave some credibility to the government’s inflation target. But it was still thought, and rightly so, that Chancellors would still make decisions to adjust rates for reasons that could have a political flavour.13 Consequently not only Lamont, but most commentators continued to advocate further moves towards operational independence. Perhaps the most influential was the report by a body of (international) experts chaired by Lord Roll, whose monograph, Independent
196
The Political Economy of Inflation Targets
and Accountable, was published in October 1993. This was supported by the all-party House of Commons Treasury Committee, whose report (December 1993) likewise called for greater independence for the Bank. Moreover, a centrepiece of the Maastricht treaty for the adoption of a single European currency and a European System of Central Banks, which was then (1992) being debated, was that all the participating Central Banks should be fully independent from government, as noted by John Plender in a long article in the Financial Times (25 September 1992). Meanwhile academic evidence was mounting that more central bank independence was correlated with lower, and more stable, inflation without any adverse effect on either the level, or volatility, of output (see in particular Alesina and Summers (1993)). On 27 May 1993, Lamont resigned as Chancellor, when Major proposed to demote him in a Cabinet reshuffle (Lamont, In Office, p. 372), and Major appointed Ken Clarke to succeed him. Clarke is pugnacious and self-confident. That combination led him simultaneously to use his continuing prerogative to set interest rates, sometimes disagreeing with the advice of the Bank, and at the same time to allow that advice (and the disagreement) to become public knowledge. Details of such disagreement, when the Chancellor overruled the Bank are presented in Cobham (1997). Once he had taken office, in May 1993, he not only decided that the Minutes of the monthly meeting, between himself and the Treasury on the one side and Eddie George and the Bank on the other, should be published, but also that the Governor’s opening statement at the meeting would be written into the Minutes verbatim.14 While this strengthened the presentational independence of the Bank, the willingness of Chancellor Clarke to go his own way underlined the continuing opportunity for political manipulation of interest rates. Ken Clarke was enjoying his stint as Chancellor, including his command over interest rates. Whatever his earlier views, he was no longer inclined to press for operational independence for the Bank of England, at least in public.15 John Major was firmly against it, mainly on a variety of political grounds. If there was, therefore, to be any early further move towards operational independence for the Bank of England, it had to come from an incoming Labour government. The inflation target, set by Lamont in 1992, had worked well, indeed considerably better than initially expected. There was no serious question of abandoning it, though a new Labour Chancellor would want to reconsider the parameters. Thus in February 1997, the then Shadow Chancellor, Gordon Brown, announced that a Labour government would also aim for an inflation
C. A. E. Goodhart 197
target of 2.5 per cent, or less (The Economist, 1 March 1997). Rather the issue was whether to proceed further with operational independence for the Bank. The Labour Party, and Gordon Brown, while still in opposition, were more cautious on this latter issue. The relevant section on their manifesto reads We will match the current target for low and stable inflation of 2.5 per cent or less. We will reform the Bank of England to ensure that decision-making on monetary policy is more effective, open, accountable and free from short-term political manipulation. The Economist (1 March 1997) reported: Moreover, the [current] regime lacks credibility, because the monthly meetings of Kenneth Clarke, the Chancellor, and Eddie George, the governor, have become personalised confrontations, dubbed by the media the ‘Ken and Eddie show’. Mr Brown does not want a Gordon and Eddie show. Instead, he plans to reform both the Bank and the Treasury to increase the range of advice available to both. A new monetary policy committee would be set up at the Bank, chaired by the governor, but including several other heavyweight full-time advisers, to decide the Bank’s monthly view. The Treasury’s ineffective panel of independent forecasters would be replaced by a Council of Economic Advisers to offer private advice to the chancellor. Yet it is hard to believe that either of these changes would make much difference. The weakness of the current regime is not that the chancellor or governor is short of advice. Nor is it a problem that they sometimes disagree: why meet if you never differ? The real problem is that the chancellor retains the power to set interest rates, and that, so long as he does so, he can set them for political rather than economic reasons. Unfortunately, Mr Brown does not propose to let the Bank set interest rates – ‘operation independence’, in the jargon. True, he says that he hopes his reforms will make independence more likely, by helping the Bank to re-build its reputation, which has been damaged by the perception that Mr Clarke has outwitted Mr George in past disagreements. And Mr Brown is probably sincere in wanting to move the Bank a step closer to independence, not least because Bank independence is a necessary pre-condition for joining a single
198
The Political Economy of Inflation Targets
European currency, on which Mr Brown is keen. But whether his ideas for the Bank actually give it more independence will depend on who is on the monetary committee and whether Mr Brown heeds its advice – and luck. Gordon Brown thus indicated, quite clearly, that he intended to let time pass, to observe the advice and performance of the to-be-established Monetary Policy Committee of the Bank, before granting it operational independence. Yet, in the event, he moved extremely quickly to make this change after their electoral victory in May 1997, to the surprise of almost all observers. So what caused this volte-face? My own guess (and it remains a guess, since he plays his cards close to his chest), is that he had always planned to do this, but that innate political caution made him reluctant to announce such a step before the election, just in case it might possibly be made into an issue by the Conservatives.16 He had accepted the principle that ‘operational independence’ was a good idea, possibly influenced by the advocacy of his personal adviser, Ed Balls, who had, so it is said, learnt to support the idea while a student at Harvard. If the basic principle is valid, why wait? In any case, prior planning for such a step had gone on for some time in secret between officials at HMT and the Bank beforehand; so enough details of the proposal were ready to hand for the new Chancellor to pull the trigger immediately after the election. What was even less advertised in advance were the accompanying measures, to transfer banking supervision from the Bank to a newly established Financial Services Authority (FSA), and management of the gilt-edged market to a new Debt Management Office (DMO). Neither had been foreshadowed in the Manifesto or in a speech by Gordon Brown (though the winding up of the Securities and Investments Board (SIB) and its appendages of the Self Regulatory Organisations (SROs), and their replacement by a statutory FSA, had been a part of announced policy). No account of the arguments, for and against, either of these functional transfers has yet been made public, though some of the arguments for such a change, for example, focus and absence of conflicts of interest, are clear enough, and had been previously made by Sam Brittan (e.g. 1992a and 1992b). Similarly, several of the arguments against separating banking supervision from a central bank are also obvious. Some were made by Forder and Oppenheimer in their ‘Personal view: How to reform the Bank of England’ in the Financial Times (1995).17 Many of the arguments, pro and con, are summarised in Goodhart (2000/2002).
C. A. E. Goodhart 199
Be that as it may, the structure of the new regime for conducting monetary policy was now in place as of May 1997.
7.4 Postscript The history of monetary reform in New Zealand and the UK is essentially a story of the interaction of economic occurrences with the development of theory to explain such events, spiced by the accidental effects of politics and personalities. The key theoretical advance, prior to the events recorded here, was the formulation of the theory that, in the medium and longer term, the Phillips curve was vertical. This translated into the policy proposition that the over-riding objective of a central bank had to be the achievement of price stability. Initially this was given operational effect by the adoption of monetary targets, again in the period prior to our historical account. The robust strength of the longer-run relationship between the monetary aggregates and nominal incomes, the attractions of the Quantity Theory, and the powerful advocacy of Milton Friedman and other leading monetarist economists made this course natural. But the short-run relationship between the various M’s and inflation was precarious, and the structure of the monetary system was such that central banks controlled the short-run interest rate, not any M (including the monetary base). So the record of monetary targetry was mixed, reasonably successful in Germany and Switzerland, less so, and effectively abandoned, in the US and the UK. That said, had the search for a regime change in New Zealand occurred, two or three years earlier, my guess is that they would also have gone for a monetary target, but by 1987/88 this was begun to be seen as problematical. The move there, from a monetary to a direct inflation target, did not arise from academic theory; it came about, almost accidentally, as a means of quantifying the fundamental objective of price stability, and one that had much more appeal to politicians and public on presentational grounds than the arcane and complicated mix of alternative M’s (the fights between adherents of monetary base, narrow money and broad money did considerable damage the cause of monetarism). The various advantages, theoretical and structural, of an inflation target, relative to a monetary target, already outlined in Section 3, were not fully appreciated at the time. But such (theoretical) advantages would not have counted for much if inflation targetry had proven unsuccessful. In practice, however, the experience of the three countries
200
The Political Economy of Inflation Targets
mentioned here, following the adoption of inflation targetry – New Zealand in 1988, Canada in 1991 and UK in 1993 – has been good. See Tables 7.1 and 7.2 and Figures 7.1 and 7.2. Of course there are qualifications. Non-IT countries also had a much better inflation performance in the 1990s and the 2000s; assessing whether IT countries did significantly better than their nonIT comparators has been the subject of several empirical papers (see, for example, Cecchetti, Flores-Lagunes and Krause 2006, and Cecchetti and Krause 2002). The great ‘moderation’ in the volatility of inflation and, though to a lesser extent, of output growth remains quite hard to explain convincingly. Nevertheless inflation targetry has been successful. No country that has adopted it has sought to reject it.18 The decision in New Zealand to give the RBNZ operational independence was even more accidental. It arose from a marriage of political happenstance and underlying economic theory. The political priority was the desire to reverse, and to prevent future intervention by politicians in the detailed running of the economy, à la Muldoon. This involved Roger Douglas in seeking to set basic objectives for public sector entities, contracting with their Chief Executives on a target for such an objective(s), and then letting them get on with the job. But a central bank is not a standard public sector industry, and the delegation of a key instrument of public policy is not quite the same Table 7.1
Annual inflation rate (consumer prices), 1970–2004 Canada
Average (1970–2004) Pre IT Post IT Standard Deviation Pre IT Post IT
Table 7.2
New Zealand
4.95 6.79 1.83 3.38 2.93 0.72
7.63 11.67 2.26 5.70 4.18 0.96
UK 7.14 9.56 2.52 5.55 5.44 0.71
Annual growth rate (GDP), 1970–2004 Canada
Average (1970–2004) Pre IT Post IT Standard Deviation Pre IT Post IT
3.36 3.94 2.74 2.25 2.30 1.99
New Zealand 2.49 2.07 3.51 2.41 2.63 1.62
UK 2.29 1.97 2.91 2.08 2.47 0.73
C. A. E. Goodhart 201 Inflation target implemented
25 20 15 10 5 0
19
70 19 72 19 74 19 76 19 78 19 80 19 82 19 84 19 86 19 88 19 90 19 92 19 94 19 96 19 98 20 00 20 02 20 04
Percent change over previous year
30
–5
Year
Canada Figure 7.1
New Zealand
United Kingdom
Consumer prices
Inflation target implemented
Percent change over previous year 19 70 19 72 19 74 19 76 19 78 19 80 19 82 19 84 19 86 19 88 19 90 19 92 19 94 19 96 19 98 20 00 20 02 20 04
10 8 6 4 2 0
–2 –4
Year Canada
Figure 7.2
New Zealand
GDP at constant prices
United Kingdom
202
The Political Economy of Inflation Targets
as setting targets for the railroad system or the health service. So I doubt whether operational independence for central banks would have had such common acceptability if it had not been further bolstered by concerns about credibility, and fears about the manipulation of interest rates for short-term political gain. There is now a vast literature on all this, which has already been briefly reviewed in sections 7.2 and 7.3. This marriage provided an excellent compromise between the need for overall strategic political and democratic control on the one hand, and for longer-term and independent tactical operation on the other. Thus the relevant politician (Minister of Finance/Chancellor) would set the specific objective for the central bank, which the central bank would seek to achieve by the unconstrained use of its main policy instrument, the short-term interest rate. The central bank only has operational independence, not goal independence; it is also accountable to the Minister, Parliament and public for its actions in pursuit of its set objective. Nevertheless, in my view, such delegation of operational independence remains on somewhat shakier grounds than inflation targetry. Politically there were many, like Margaret Thatcher and John Major, who doubted whether such delegation was desirable or even really feasible. Economically the concept of a vertical longer-run Phillips curve is not generally accepted; there is always a powerful and vocal lobby (usually among manufacturers) for easy money; whenever growth flags and unemployment rises there are those who blame this on the central bank’s excessive commitment to price stability. Moreover, should an independent central bank’s policy seriously discommode the political authorities, the latter can always revoke that independence. For example, there were recent concerns in Japan in early 2006 that the attempt by the Bank of Japan (BoJ) to return to a more normal stance of monetary policy, by raising interest rates at some near future time, might jeopardise either the government’s priority of lowering the public sector deficit and/or the strength of the recovery there. In this context, there have been some vague threats of rescinding the BoJ’s independence, though nothing came of that. The conflict in Poland, also in early 2006, between central bank governor L. Balcerowicz and the government is another example. In other papers (Goodhart 2002, Friedman and Goodhart 2003, Goodhart and Meade 2004), I have expressed the hope that central bank operational independence will come to be regarded as a pillar of the Constitution, akin to its role to judicial independence. But there is
C. A. E. Goodhart 203
a long way to go before that state is reached. For the time being, central bank independence remains, perhaps, a somewhat fragile and delicate plant, despite its strong theoretical and empirical underpinning.
Notes My thanks are due to Arthur Grimes for help with the record of events in New Zealand, and to Jonathan Ng for research assistance. 1. Though Grimes (1987) supported this in an internal paper, stressing the time inconsistency analysis. Arthur was also a Ph.D. student at the newly founded Financial Markets Group at LSE between 1985 and 1988. 2. The distinction between goal independence and operational independence was articulated by Debelle and Fischer (1994) and again by Fischer in Capie et al. (1994). 3. There was no support, either from the public submissions or in the Select Committee, for the Treasury’s prior proposal of a constraint on the note issue combined with a commercially competitive RBNZ. 4. It is a curiosum that monetary textbooks, at least until recently, typically assume the counter-factual, that is, that the central banks choose the highpowered monetary base and that the short-term interest rate is endogenous. That, often brilliant, academics can persist in such patent error for so long is hardly a good advertisement for our profession. 5. Politicians also have differing agenda depending on their office. Prime Ministers, and US Presidents, are generally more conscious of the political consequences of raising interest rates, whereas Chancellors, and Secretaries of the Treasury, are inevitably more aware of the medium term consequences of not raising interest rates. It would be an interesting research exercise to go through the Public Records Office to find out how often British Prime Ministers had asked Chancellors to defer their proposals to change interest rates before 1993. My expectation would be no requests to defer proposed cuts, and on average one, or two, such requests to defer raises every year. Thus Norman Lamont, having been Chancellor, noted, (In Office, p. 322), that ‘my experiences with the Prime Minister over interest rates convinced me that politicians would always want to interfere with interest rate decisions for political reasons’. 6. In personal correspondence, Norman Lamont has given his support to this last argument, arguing that it would have been very difficult for a newly independent central bank to have brought down inflation from 10% (the level in 1990) to 2%. I think that there would have been an enormous backlash against the high interest rates, negative equity, etc., being imposed by a newly independent bank. There would have been strong pressure from the Labour Party, and indeed others, for politicians ‘to assume their responsibility’ and there would have been accusations of ‘reactionary’ ‘deflationary’ policies.
204
The Political Economy of Inflation Targets
7. Lamont had tried to make the case for independence previously in September 1991 (In Office, pp. 323–4), but the case for doing so then was complicated by the UK’s involvement in the European exchange rate mechanism (ERM). 8. Ken Clarke had an ambivalent attitude towards granting independence to the Bank of England. Major (op. cit., p. 682) records him as being in favour, but, after his experience as Chancellor from 1993 to 1997, he reached the view that he was better placed to determine interest rates than an independent Monetary Policy Committee (MPC), see Financial Times, 21 July 1997, p. 16. 9. Almost my first job at the Bank of England in 1968/69 was to explain why, and how, we had adopted a target for Domestic Credit Expansion (DCE) which had been required of the UK as one of the conditions for an IMF loan. Thus I wrote (Goodhart 1997), that Much of this was a novel concept to British economists, especially the key role of an (assumed predictable) demand-for-money function, and antipathetic to many – recall that the Radcliffe Report had denied the stability, or even the usefulness as a concept, of velocity. So my first role at the Bank was to try to explain the concept, and role of DCE, both within and without the Bank. [I was the main author of Bank of England (1969).] Actually, we had to go further. To protect British amour propre, there had to be some pretence that we, in the UK, had thought up this wonderful new wheeze, rather than had it foisted upon us, out of weakness, by the IMF. The game continues to be played in the same way. 10. Though this propagated the error that One novelty in the New Zealand plan, is that the Reserve Bank governor, who rejoices in the improbable name of a central banker of Mr Donald Brash, will be put on a performance-related pay scheme in which his remuneration and prospects will be linked to his success in bringing about price stability. At a recent private gathering of officials from Britain and New Zealand, Sir George Blunden, the Bank of England’s deputy governor, remarked that this proposal had prompted other central bankers to look at Mr Brash ‘not only with envy but also with awe’. 11. This, however, was misinterpreted in the 1960s and 1970s by monetarist economists. If one assumes that the central bank sets interest rates without any feedback from relevant economic variables, for example, forecast inflation, output gap, exchange rate and so on, the above procedure can be shown to be wildly destabilising (whereas a fixed monetary aggregate target is ultimately stable). However the idea that the interest rate decision is taken without feedback from key economic variables is patently absurd. The nature of the feedback is now encapsulated in estimated reaction functions, of which the Taylor reaction function is best known.
C. A. E. Goodhart 205 12. It is one of the misfortunes of the ECB that they did not allow the political authorities in the eurozone to help determine the precise choice of the inflation target. 13. In order to reduce the temptation for a Chancellor to choose dates for announcing interest rate changes that were politically propitious (i.e. down just before elections, up just after elections), there was for a time the risible arrangement that, whereas the Chancellor would still make the interest rate decision, the Bank would choose the date of its announcement. 14. In personal correspondence, Ken Clarke wrote to me that I did persuade John Major in an unguarded moment to agree that I would publish the minutes of the monthly meetings between me, my full ministerial team, senior Treasury officials and the Bank. I insisted that genuine minutes would be produced. Unfortunately, it was impossible to persuade the financial press that they were genuine and accurate until the first time that these minutes showed a disagreement between me and the Governor. I remember saying to the Select Committee that I wished the press would realise that the minutes were an important source of information and would stop regarding them as the ‘Ken and Eddie Show’. Of course, I came to regret this chance remark because the description of the meetings stuck, but they were a very important innovation in opening up the basis of monetary policy and information to the markets in this country. I did not allow other political considerations to affect my own judgement, but I continued to fail to persuade the outside world of that. 15. Again, in personal correspondence, Clarke wrote, I had always believed that right of centre parties should favour the independence of central banks and I made my own unsuccessful attempts to persuade John Major to consider it. He remained deeply hostile to the idea and strongly suspected that the Bank had an in-built bias towards raising interest rates …. I determined to make as much institutional change and to introduce as much transparency as possible, both to prepare the way for Bank independence and to free myself of purely political pressure, for so long as I had to take monetary decisions myself. I was actually quite determined to take decisions on interest rates based solely on the aim of lowering inflation and inflationary expectations and to resist all other short-term political pressures. 16. Though Norman Lamont had privately urged both Tony Blair and Gordon Brown to adopt operational independence for the Bank in the period between his return to the backbenches and the 1997 General Election (see In Office, p. 326). 17. There is no serious dispute that the Bank should remain the lender of last resort to guarantee the liquidity of the banking sector in moments of crisis. Given the heavily discretionary nature of the supervision process, it is hard to see how the lender of last resort and supervisory functions can meaningfully be vested in different bodies. The Bank of England could scarcely be expected
206
The Political Economy of Inflation Targets
to take orders from some outside quango on the supply of emergency funds to a bank in distress. 18. There was quite a tough test in Canada where the incoming Liberal government had campaigned against, what they termed, the deflationary policies of John Crow, the then Governor of the Bank of Canada, and the outgoing Conservative government in 1993. Although Crow was replaced by Thiessen, nevertheless, the Liberal government reaffirmed the strategy of published inflation targets.
Bibliography Alesina, A. and N. Roubini (1997), Political Cycles and the Macroeconomy (Cambridge, MA: MIT Press). Alesina, A. and L. Summers (1993), ‘Central Bank Independence and Macroeconomic Performance’, Journal of Money, Credit and Banking, vol. 25, no. 2, pp. 151–62. Bank of England (1969), ‘Domestic Credit Expansion’, Bank of England Quarterly Bulletin, vol. 9(3), pp. 363–82. Barro, R. J. and D. B. Gordon (1983a), ‘Rules, discretion and reputation in a model of monetary policy’, Journal of Monetary Economics, vol. 12, no. 1, pp. 101–21. —— (1983b), ‘A positive theory of monetary policy in a natural rate model’, Journal of Political Economy, vol. 91, no. 4, pp. 589–610. Bindseil, U. (2004), Monetary Policy Implementation: Theory, Past and Present, (Oxford: Oxford University Press). Brittan, S. (1992a), ‘Advertise for Governor’, Financial Times, 26 October. —— (1992b), ‘Not accountable yet’, Financial Times, 2 November. Capie, F., Fischer, S., Goodhart, C. and N. Schnadt (1994), The Future of Central Banking (Cambridge University Press), pp. 262–308. Cecchetti, S. G., Flores-Lagunes, A. and S. Krause (2006), ‘Has monetary policy become more efficient? A cross-country analysis’, The Economic Journal, vol. 116 (511), April, pp. 408–33. Cecchetti, S. G. and S. Krause (2002), ‘Central Bank Structure, Policy Efficiency and Macroeconomic Performance: Exploring empirical relationships’, Federal Reserve Bank of St. Louis Review, vol. 84(4), (July/August), pp. 45–60. Cobham, D. (1997), ‘The post-ERM framework for monetary policy in the United Kingdom: Bounded credibility’, The Economic Journal, vol. 107, no. 443, (July), pp. 1128–41. Congdon, T. (1996), ‘The role of Central Banking in Economic Development’, The Review of Policy Issues, vol. 2(2), Spring, pp. 73–90. Crow, J. (2002), Making Money (Etobicoke, Ontario: Wiley Canada, Ltd). Debelle, G. and S. Fischer (1994), ‘How independent should a central bank be?’ in J. C. Fuhrer, ed., Goals, Guidelines and Constraints Facing Monetary Policymakers (Boston, MA: Federal Reserve Bank of Boston), pp. 195–221. Foot, M. D. K., Goodhart, C. A. E. and A. C. Hotson (1979), ‘Monetary base control’, Bank of England Quarterly Bulletin, vol. 19, no. 2, pp. 149–59. Forder, J. and P. Oppenheimer (1995), ‘Personal view: How to reform the Bank of England’, Financial Times, 1 August, p. 15.
C. A. E. Goodhart 207 Friedman, M. and C. A. E. Goodhart (2003), Money, Inflation and the Constitutional Position of the Central Bank (London: The Institute of Economic Affairs (IEA)). Goodhart, C. A. E. (1988), The Evolution of Central Banks, (Cambridge, MA: MIT Press). —— (1989), ‘The Conduct of Monetary Policy’, Economic Journal, 99 (396), June, 293–346. —— (1997), ‘Whither now?’, Banca Nazionale del Lavoro Quarterly Review, vol. 203, (December), pp. 385–430. —— (2000), ‘The Organizational Structure of Banking Supervision’, Financial Markets Group, London School of Economics, Special Paper No. 127, (October), reprinted in Economic Notes by Banca Monte dei Paschi di Siena, vol. 31, (1), pp. 1–32. —— (2002), ‘The Constitutional Position of an Independent Central Bank’, Government and Opposition, vol. 37 (2), (Spring), pp. 190–210. —— (2004), ‘The Bank of England, 1970–2000’, Chapter 17 in R. Michie and P. Williamson, eds, The British Government and the City of London in the Twentieth Century (Cambridge: Cambridge University Press). Goodhart, C. A. E. and E. Meade (2004), ‘Central Banks and Supreme Courts’, Moneda y Crédito, vol. 218, pp. 11–54. Grimes, A. (1987), ‘Reputation and credibility in monetary policy’, mimeo, Reserve Bank of New Zealand. Issing, O., Gaspar, V., Tristani, O. and D. Vestin (2005), Imperfect Knowledge and Monetary Policy, The Stone Lectures in Economics, (Cambridge University Press). Kaldor, N. (1966), Causes of the Slow Rate of Economic Growth in the UK (Cambridge University Press). Kydland, F. E. and E. C. Prescott (1977), ‘Rules rather than discretion: the inconsistency of optimal plans’, Journal of Political Economy, vol. 85, (June), pp. 473–91. Laidler, D. E. W. (2003), ‘Two views of the lender of last resort: Thornton and Bagehot’, Cahiers d’Economie Politique, vol. 45, pp. 63–78; reproduced in In Retrospect (Edward Elgar: Cheltenham, UK, 2004). —— (2004), ‘Monetary Policy without Money: Hamlet without the Ghost’, in C. Goodlet, D. Longworth and J. Murray, eds, Macroeconomics, Monetary Policy, and Financial Stability: A Festschrift in Honour of Charles Freedman (Bank of Canada), Session 2, pp. 111–42. Lamont, N. (1999), In Office (London: Little, Brown and Company). Lawson, N. (1981), ‘Thatcherism in practice: a progress report’, speech to the Zurich Society of Economics, 14 January, H. M. Treasury Press Release. —— (1992), The View from No. 11 (London: Bantam Press). Major, J. (1999), ‘The Economy: Rags to Riches’, Chapter in John Major: The Autobiography (London: HarperCollins Publishers Ltd). McCallum, B. T. (2001), ‘Monetary Policy Analysis in Models without Money’, Federal Reserve Bank of St Louis Review, vol. 83 (4), July, pp. 145–64. Nordhaus, W. D. (1975), ‘The Political Business Cycle’, Review of Economic Studies, vol. 42, pp. 169–190. Rogoff, K. (1985), ‘The Optimal Degree of Commitment to an Intermediate Target’, Quarterly Journal of Economics, vol. 100(4), pp. 1169–89.
208
The Political Economy of Inflation Targets
Roll, E. (Chairman of an independent panel) (1998), Independent and Accountable: A new mandate for the Bank of England (London: Centre for Economic Policy Research). Seldon, A. (1997), Major: A Political Life (London: Phoenix). Singleton, J., Grimes, A., Hawke, G. and F. Holmes (2005), ‘Twenty years of modernisation: The Reserve Bank of New Zealand’, in N. Courtis and P. Nicholl, eds, in Central Bank Modernisation (London: Central Bank Publications), pp. 167–82. —— (2006, forthcoming), ‘The Reserve Bank of New Zealand Act, 1989’, Chapter 5 in Innovation in Central Banking: A History of the Reserve Bank of New Zealand, by the same authors (Auckland: Auckland University Press). Svensson, L. E. O. (1997), ‘Inflation forecast targeting: Implementing and monitoring inflation targets’, European Economic Review, vol. 41 (6), (October), pp. 1111–46. Thatcher, M. (1993), The Downing Street Years (London: HarperCollins Publishers Ltd.). Treasury and Civil Service Committee, (1981), Monetary Policy: Report (London: HMSO). Walsh, C. E. (1995), ‘Optimal Contracts for Central Bankers’, American Economic Review, vol. 85, (March), pp. 150–67. Walters, A. (1986), Britain’s Economic Renaissance (New York: Oxford University Press).
Discussion Charles Freedman
It is always a pleasure to read and comment on papers by Charles Goodhart. They are invariably interesting, insightful and well written. And this paper is no exception. My only difficulty with the paper is that there is very little in it with which I disagree, an uncomfortable situation for a discussant. So what I propose to do this is to focus on those issues in the paper that in my view are worthy of further comment, expand to some extent on some of Charles’s arguments and assertions, and add a few comparisons with the Canadian experience with inflation targeting. Let me begin with the New Zealand experience. While the link between the changes to the Reserve Bank of New Zealand Act and the general movement to public sector accountability in New Zealand is well known, I had not fully appreciated until I read the paper that in some ways accountability concerns were as important a driving force in the introduction of inflation targeting in New Zealand as the desire to achieve and maintain low inflation. More typically, the motivation for the adoption of inflation targeting has come from one of two sources. In some countries, such as Canada, it has been seen as a way of moving from a history of relatively high inflation to a much lower rate of inflation. The second motivation has come from the forced movement of some countries from a fixed exchange rate to a floating exchange rate, typically as a result of outflows of international capital, and their resulting need to find an alternative nominal anchor for monetary policy. With the loss of their exchange rate anchor and their inability to use monetary aggregates as an anchor because of their perceived instability, inflation targets have become the preferred alternative for such countries. And, thus far, the inflation targeting regime has been successful in both industrialised and emerging economies, with no 209
210
The Political Economy of Inflation Targets
country abandoning the framework because of an inability to achieve the desired objectives of policy. The New Zealand experience is interesting in a number of dimensions. The way that they viewed their targets initially was what I call ‘hardedged’. That is, as Charles notes, the governor was accountable in principle for failure to meet the agreed 0–2 per cent target range and subject to the potential disciplinary action of dismissal. Thus, a 1.9 per cent inflation outcome was acceptable but a 2.1 per cent inflation outcome was not. And, consequently, as inflation rose towards the upper band of the target range, the Reserve Bank (RBNZ) would take aggressive action to prevent it from moving outside the target range. Now the difference between 1.9 and 2.1 per cent is not terribly important from an economic point of view. So structuring the arrangements in such a way as to treat the upper band as hard edged gave rise to the need for aggressive policy actions that might otherwise not have made a lot of economic sense. In contrast, the initial Bank of Canada arrangements explicitly contemplated the possibility of inflation moving outside the range and made clear that if this happened, the Bank would take action to gradually bring inflation back to the centre of the target range. The key word here is gradually. The view in Canada was that it made no sense to impose excess volatility on output in order to bring inflation back into the target range in short order. A related point in the New Zealand experience was the way in which the RBNZ operated policy in order to maintain a very tight control over the inflation rate. During the first five years or so of their inflation targeting experience, the main mechanism by which the RBNZ influenced inflation was through movements of the exchange rate, making use of the direct linkage between exchange rate movements and prices. This involved appreciable movements in the policy interest rate that were designed to bring about the desired exchange rate movement. And, as a consequence of the relatively sharp movements in interest rates and exchange rates, the amplitude of output fluctuations was also rather large. The implicit loss function of the RBNZ thus appeared to put relatively little weight on the variability of the output gap (the deviation of output from potential). The way that this showed up in practice was the use by the RBNZ of a relatively short policy horizon, on the order of four quarters. This approach contrasted with that used in Canada and most other inflation targeting countries. These countries put more weight on dampening the amplitude of output fluctuations (i.e., more weight on the output gap term in the implicit loss function), and therefore used
Discussion
211
a longer policy horizon in determining the appropriate policy actions (in the case of Canada, six to eight quarters). To use later terminology, most countries were flexible inflation targeters while New Zealand came closest to being a pure inflation targeter. Over time, as it became apparent that the output fluctuations being generated by the policy approach were excessive and unnecessary, the RBNZ changed its approach to that used by other central banks. I would like to make one other comment on the idea that central banks should not allow inflation to move outside the announced target range. In choosing a target range, whether explicit as in Canada or implicit as in the UK (where movements greater than +/−1 per cent away from the target give rise to a letter from the Governor to the Chancellor of the Exchequer), central banks expected to be outside the target range a certain proportion of the time. In the case of Canada, the initial expectation was that the range would include about two-thirds of inflation outcomes. In fact, the proportion of inflation outcomes within the range has turned out to be much higher than expected at the outset. While some have argued that the reason for this unexpected outcome is that there have been fewer shocks to the system over the inflation targeting period, I would put considerable weight in explaining this result on the effect of the inflation targeting framework itself in anchoring the system and changing the dynamics of the inflation process. In any case, in assessing the success of the various inflation targeters, researchers should be careful not to treat the proportion of outcomes within the range as an unequivocal measure of success. A 100 per cent result may mean that the central bank has been putting too much weight on achieving inflation outcomes very close to target and too little on dampening output movements. Turning to the Bank of England, I found Charles’s discussion of the various initiatives to give greater independence to the Bank of England, both before and in the early years of inflation targeting period, a fascinating insight into the politics that lie behind decisions of this nature. It is of interest to note that some of the current literature on the so-called preconditions for inflation targeting in emerging economies treat instrument independence (or as Charles refers to it, operational independence) as a prerequisite for the adoption of inflation targeting. And yet for the entire period from September 1992 to May 1997, monetary policy actions to achieve the inflation target were the responsibility of the Chancellor of the Exchequer and not of the Bank of England. Nonetheless, in most academic articles about inflation targeting, the UK is treated as a full-fledged inflation targeter from 1992 on. What is the
212
The Political Economy of Inflation Targets
difference between the general assessment of what is required to be a full-fledged inflation targeter and this treatment of the UK? I suppose that it is the view that giving the government of most emerging economies the responsibility for interest rate adjustments would not result in a policy that is truly aimed at achieving the inflation target, whereas in the UK (at least in the period under study), the Chancellor did by and large take seriously the inflation targets as the objective of policy when making interest rate decisions. One by-product of the unusual path taken in the UK with respect to the institutional relationships between the government and the central bank is that the central document issued by the Bank of England was called the Inflation Report. This made sense in the circumstances, since the Bank of England initially was reporting its views on the inflation outcome that was likely if interest rates remained unchanged. In fact, it would be more reflective of the current reality for the document to be called the Monetary Policy Statement or Monetary Policy Report, as is done in New Zealand and Canada, respectively. Unfortunately, because of the Bank of England’s example, a number of inflation targeting countries have chosen to use the term Inflation Report rather than Monetary Policy Report. Another by-product of the initial arrangements, as Charles points out, was the use of unchanged interest rates in the base case forecast. There has been considerable debate in recent years about the logic of using constant interest rates, or indeed market expectations of future interest rates, as the basis of the central bank forecast. And some central banks are now using in their base case forecast, an endogenous interest rate path that achieves the target rate of inflation at the end of the policy horizon. Moreover, two of them, the Reserve Bank of New Zealand and the Norges Bank, even make public the endogenous interest rate path. I expect that more countries will move in this direction over time. Charles also notes the importance of the insight that it is expected or forecast future inflation that has to be the objective of policy and not current inflation. In the mid-1980s, at the Bank of Canada, we did a study of various potential targeting arrangements in which we compared money, nominal spending and exchange rate targets as ways of anchoring monetary policy. Prices and inflation were not even candidates for the target in this study because the model that was being used exploded if the central bank tried to target current prices. The insight that a central bank had to target future prices or inflation, not their current level, was crucially important in moving thinking on this issue forward.
Discussion
213
Another issue of interest discussed by Charles is the question of the democratic deficit and the way that the inflation targeting regime deals with it by assigning responsibility for the numerical goal of inflation either to the government or to the government in conjunction with the central bank. There are relatively few cases of central banks with goal independence, although the ECB and the Czech National Bank are exceptions. In practice, there seems to be relatively little difference in behaviour across countries in which the goal is set by government, those in which it is a joint decision of the government and the central bank, and those in which the central bank itself sets the target. But I agree with Charles that involvement of the government in the setting of the target is very helpful in dealing with the political perceptions of the process. As well, from an economic point of view, having the government involved in the process and clearly supportive of the targeting framework and objective can be very helpful in persuading the public and financial markets that the government is less likely to abandon the policy framework at times of economic difficulty. In hindsight, one area where inflation targeting central banks could have done better at the outset of the regime was in the rhetoric that they used to describe the framework and policy goals during the early years of inflation targeting. The way that central banks initially described the policy made it sound as if they put very little or no weight on dampening output fluctuations in their loss function. The reality of course was that almost all inflation targeting central banks pursued what later became known as flexible inflation targeting, but they talked as if they were ‘inflation nutters’, to use Mervyn King’s term. Perhaps it was essential to use the more hard-line rhetoric at the time in order to build the credibility of the new framework, but it was unfortunate because it left the impression that central banks cared about nothing but inflation, and this impression was used as one of their main arguments by opponents of the new policy framework. As there was increased understanding of the way that central banks operated under a flexible inflation targeting regime, and as experience began to show that low inflation could coexist with a solid economic performance, this opposition faded away. One interesting difference, between New Zealand and the UK on the one hand, and Canada, on the other, was the political environment at the time of the introduction of the targets. In both New Zealand and the UK, it was the left-wing party that was responsible for the new arrangements, while in Canada it was the right-wing party, the Progressive Conservative party that was in power when the targets were adopted.
214
The Political Economy of Inflation Targets
Moreover, the more left-wing party, the Liberal party that won the election a couple of years later, had opposed the targets while in opposition. However, by the time that they came to power the inflation rate was already down to about two per cent and it would have been both bad economics and bad politics to have abandoned the targets at that time. Indeed, the renewal of the targets every three or five years since that time has basically been a non-issue politically, with very little attention paid to it in the media. To conclude, Charles has offered us a paper that illustrates the way that economic developments, economic theory and political circumstances can interact in bringing about changes in policy regimes. Such a study is very much to be welcomed, given the reality of the role of politics in important institutional and framework changes, and the lack of discussion that it usually receives in economic studies.
8 Monetary Union Proposals in North America and Southern Africa: Do the Same Q&A Apply? J. Clark Leith
Introduction Monetary unions are hotly debated in contemporary economics. Unlike many issues in the discipline where the consensus is overwhelmingly on one side (e.g., that there are gains from trade), there is no consensus about the net benefit of monetary union. The debate over whether or not Canada should seek to join in creating a North American Monetary Union (NAMU) aptly illustrates the absence of consensus. David Laidler has been the front and the centre in that debate.1 Similar policy choices have been debated in several other parts of the world: Europe, the former Soviet bloc, the Persian Gulf, the Caribbean, Latin America, Asia, and Africa.2 Within Africa there are the existing CFA franc monetary unions in Central and West Africa, as well as the Rand Common Monetary Area (CMA)3 of Lesotho, Namibia, and Swaziland with South Africa. The African Union (AU) has declared that an African Monetary Union will be achieved by 2021, preceded by monetary unions in various regions of Africa,4 one of which would be the countries of Southern African Development Community (SADC) which consists of the Rand CMA plus ten other members. SADC is itself committed to the creation of a customs union by 2010, a common market by 2015, and a monetary union by 2016.5 Botswana is a member of SADC, but had withdrawn from the Rand CMA to create Botswana’s own currency, called the Pula, in 1976. Given the AU and SADC agendas, one of the issues on Botswana’s policy agenda must be whether to pursue any sort of monetary union. Would it be in Botswana’s interest to rejoin the Rand CMA? Would a wider monetary union of the SADC member states be in Botswana’s long-run interest? 215
216
Monetary Union Proposals in North America and Southern Africa
During the past two decades, I have been involved for extended periods as an economic policy advisor in Botswana, sometimes in the finance ministry and sometimes in the central bank. This question has inevitably been hovering in the background as I think about the macroeconomic policy issues facing Botswana. However, I must emphasise that what follows in no way represents official views in Botswana. Fortunately for the policy analyst in Southern Africa, a substantial literature on monetary unions, both theoretical and applied, already exists. Yet that leaves the analyst to ponder: are the relevant questions the same; and if so, are the answers the same? Canada’s close economic relationship with its larger neighbour US has considerable potential as an analogy in considering Botswana’s even closer economic relationship with its relatively much larger neighbour, South Africa. Hence, one place to start in looking at the question of Botswana’s potential entry into monetary union in Southern Africa is the debate about Canada and a NAMU. And, it follows that David Laidler’s contribution to that literature might provide important insights into the policy choice facing Botswana and the other nations of Southern Africa. In this paper I will sketch the key features of the established literature as a prelude to outlining key issues in the Canadian case. I will then turn to the Botswana case. The bottom line will be summed up in the conclusion.
Economic benefits and costs of monetary union The principal gain from a common currency is the lower cost of transactions on both current and capital account between residents of one monetary jurisdiction and another. The gains from a common currency depend then on the magnitude of the transactions between residents of the two jurisdictions. But a common currency means giving up the ability to pursue an autonomous policy in matters of money and the exchange rate. Whether the net effect of a currency union is positive or negative depends on whether or not the prospective members constitute an ‘optimum currency area’.6 If the economies are similar in structure, face similar disturbances, and have mechanisms that facilitate smooth adjustment to disturbances, then common monetary and exchange rate arrangements will affect the economies similarly. In these circumstances, a common economic policy response is called for. It follows that the economic benefits of sharing a common currency are likely to outweigh the reduced policy autonomy, and the economies are said to constitute an optimum currency area.
J. Clark Leith 217
To elaborate on these features, consider first the benefits. Where residents of one monetary jurisdiction and another enter into economic relationships, different monies mean greater costs to the participants. The costs may be as straightforward as the cost of changing money from the currency of the buyer to the currency of the seller. There may be more subtle costs such as uncertainty about the exchange rate at some future date, or even the risk that it may not be possible to complete the transaction legally at that future date. For some, with financial interests in different monetary jurisdictions, it may be necessary to keep accounts in multiple currencies. These extra costs would largely disappear if there were a common currency.7 The gain from the elimination of these extra costs will be greater, the greater the economic relationships between the residents of the prospective monetary union. Second, economies with similar structures are likely to be affected in similar ways by economic disturbances. For example, an important disturbance is a change in the terms of international trade. If prospective partners have similar structures of international trade, then a change in the world prices of particular exports and/or imports will affect their real incomes similarly, and a common policy response would be appropriate. If, however, there are different structures of international trade, then an adjustment appropriate for one partner need not be appropriate for others.8 Third, for economies experiencing similar disturbances, a common policy response would be appropriate. But if the economies of prospective partners in a common monetary order typically face different types (or severity) of shocks, then a common policy response is unlikely to be optimal for the partners. For example, if one economy experiences periodic severe droughts while others do not, the appropriate policy response will not be the same for all. Fourth, when there are disturbances that affect different parts of a common monetary area differently, adjustment may occur automatically in ways that do not require specific policy action. For example, movement of labour from a region experiencing a drop in demand to one experiencing an increase, could offset what might otherwise be unemployment in the contracting area and wage pressure in the expanding one. But if the various jurisdictions of a monetary union do not permit labour mobility, this mode of adjustment would be precluded. Another mode of automatic adjustment may be fiscal. In a monetary union that also constitutes a fiscal union, a union-wide social safety net may serve as a shock absorber. Regions adversely affected by a shock have their incomes supplemented, while expanding regions
218
Monetary Union Proposals in North America and Southern Africa
pay more taxes. However, such fiscal arrangements are by no means an inherent part of a monetary union.
Policy autonomy and accountability The foregoing economic analysis, however, is incomplete. It presumes that an appropriate policy response to a disturbance would yield a better economic outcome. Hence, loss of policy autonomy would be undesirable. Yet there is an important strand in the literature which suggests that the discipline of an ‘agency of restraint’ (Collier, 1996) might be preferable to policy autonomy. If policy has a persistently perverse effect on the performance of an economy, then the loss of policy autonomy would in fact be a gain (Bayoumi and Eichengreen, 1994, p. 7). Thus, a common monetary area could restrain fiscal profligacy by imposing market discipline on member governments. Masson and Pattillo (2004) are less convinced: monetary union per se need not enforce fiscal discipline on a state. Supplementary rules may be required. The negative side of policy autonomy is also embodied in what Laidler and Robson (2002) refer to as the ‘lazy manager’ hypothesis: slower productivity growth in Canada is compensated for by adjustment of the exchange rate. Courchene (1999), Courchene and Harris (1999), and Harris (1999) argued that weakness of the Canadian dollar might discourage productivity improvements in Canadian firms producing tradeables. If Canada were to be part of a monetary union with regions experiencing more rapid productivity growth, then Canadian firms would be forced to compete by introducing more productive techniques. In a similar vein is Grubel’s (1999) argument that the exchange rate permits perverse policies to persist. In both cases, the discipline of a common currency could, it is alleged, improve Canada’s economic performance. Underlying the issue of whether surrender of policy autonomy would be an economic gain or loss is the political question: what are the forces explaining the choice to maintain a separate currency, and to fix or float the exchange rate? Helleiner (2006) reviews the sources of political support for Canada’s long-running floating exchange rate, and argues that domestic interests, preferences of state policymakers, and the international context all help explain the choice. These debates about policy autonomy point to a more general issue raised by Laidler. Formation of a currency union implies more than simply the adoption of a common currency. It alters the nature of the monetary order.
J. Clark Leith 219
Any monetary order … involves, among other things, a regulatory and supervisory framework for the banking system and other financial institutions, a set of institutional arrangements within which monetary policy is conducted on a day to day basis, not to mention the political mechanisms through which the goals of monetary policy are chosen, the relationship between fiscal and monetary policy is managed, and the accountability of policy makers to the public at large is defined and enforced. Laidler (2003, p. 15) Evaluation of any currency union proposal must consider these dimensions as well. For Canada, entry into a NAMU would be ‘No Small Change’ in the monetary order. It would sharply reduce the accountability of ‘the makers of monetary policy … to the voters for the goals they set and their performance in achieving them’.9 It could also profoundly alter the responsibility for matters such as lender-of-last-resort, deposit insurance, supervision of financial institutions, and competition in the financial sector. Laidler (2003) argues convincingly that such changes in Canada’s monetary order weigh decisively against Canadian participation in a NAMU.
Case of Botswana: Economics10 If there is a case for Botswana to join a monetary union, the most obvious candidate would be to rejoin the Rand CMA. The Southern African Customs Union (SACU), which consists of Botswana, Lesotho, Namibia, South Africa, and Swaziland, would then coincide with the Rand CMA. Over 80% of Botswana’s merchandise imports are from SACU partners, and a little under 10% of merchandise exports go to SACU partners.11 South Africa is by far the largest economy in Southern Africa, generating 92% of SACU GDP, and over 70% of the SADC GDP (Table 8.1). South Africa also has a higher per capita income than most countries of the region. Only Mauritius and Botswana have similar GDP per capita in PPP terms. Many other SADC countries are much poorer: Congo, Madagascar, Malawi, Mozambique, Tanzania, and Zambia have PPP GDP per capita running at about $1000 or less. Botswana is a very open economy. Current account credits in 2005 were about two-thirds of GDP, and debits a little over 50% of GDP. If Botswana were to rejoin the Rand CMA, or perhaps join a SADC
220
Monetary Union Proposals in North America and Southern Africa
Table 8.1
SADC member GDP and PPP GDP per capita 2004 GDP US$ billions
SACU Botswana Lesotho Namibia South Africa Swaziland Other SADC Angola Congo, D.R. Madagascar Malawi Mauritius Mozambique Tanzania Zambia Zimbabwe Weighted average Simple average
8.97 1.31 5.71 212.78 2.40
Share SACU (%) 3.9 0.6 2.5 92.0 1.0
19.49 6.63 4.36 1.88 6.03 6.09 10.85 5.40 4.70
Share SADC GDP/cap PPP (%) US$ (000) 3.0 0.4 1.9 71.7 0.8
9.14 2.41 6.82 10.29 5.18
6.6 2.2 1.5 0.6 2.0 2.1 3.7 1.8 1.6
2.00 0.65 0.79 0.59 11.05 1.14 0.62 0.87 1.90 8.32 3.82
Source: World Bank, World Development Indicators.
currency union, the principal benefit would be saving on the cost of transactions with South Africa. The Botswana balance of payments statistics and foreign exchange market transactions do not provide the detail necessary to identify the volume of Rand transactions as a whole. However, if we assume that the current account debits and credits are in the same proportion as the trade in goods with SACU, then we can calculate the proportion of GDP that involves current account transactions with SACU, which is almost entirely South Africa. Using the shares of Botswana’s merchandise imports and exports with SACU to represent the current account debits and credits as a whole, we calculate that about 47% of GDP involves transactions with SACU. If the extra cost of carrying out those transactions were, say, 1%, then the saving from joining the Rand CMA would be 0.47% of GDP each year. The structure of Botswana’s economy is strikingly different from South Africa’s, as Figure 8.1 demonstrates. The most conspicuous difference is Botswana’s very large mining sector, and South Africa’s substantial manufacturing sector. Although South Africa is often cited as a minerals-based economy, the share of the South African minerals sector at 6% of GDP is notably less than Botswana’s 40%.
J. Clark Leith 221
45% 40% 35% 30% 25% 20% 15% 10% 5%
Botswana Figure 8.1
Soc & pers’l serv
Gen’l gov
Fin & busin serv
Transp, posts & telecom
Tr, hotls & rest
Constr
Water & electr
Manuf
Mining
Agric
0%
South Africa
Structure of Botswana and South African economies
Sources: Botswana Central Statistics Office, Gross Domestic Product by Type of Economic Activity, 2004/05. Statistics South Africa, Quarterly gross domestic product by industry at current prices, 2005.
South Africa’s minerals sector is more diversified than Botswana’s. The bulk of Botswana’s mineral production is diamonds, followed by copper-nickel.12 South Africa has substantial production of platinum, gold, and coal for export,13 as well as synthetic petroleum from coal, mostly for domestic use. South Africa’s diamond production is now smaller in absolute terms than Botswana’s, and ranks as South Africa’s fifth largest four-digit export. South Africa has a substantial manufacturing sector, at nearly 17% of GDP (about the same share as Canada), in part due to decades of protection under apartheid, and in part because of the working of the SACU. South Africa also has larger non-traded service sectors such as trade, hotels, and restaurants at 13% and financial and business services at 19%.
222
Monetary Union Proposals in North America and Southern Africa
It is the combination of Botswana’s less diversified and more open economy that is the central economic issue in considering a possible monetary union involving South Africa. Even if the disturbances affecting the two economies are identical, these differences would mean different economic impacts. The major disturbances affecting both Botswana and South Africa have been drought and changes in the international terms of trade. There are, however, subtle differences. The climate of the northern part of South Africa is similar to Botswana’s desert climate with frequent droughts, but parts of South Africa have a Mediterranean climate that is less prone to drought, and which favours more temperate agriculture. The biggest shocks affecting both Botswana and South Africa have been in their international trade. Again, there are subtle differences. The world price of Botswana’s major export, diamonds, changes very little from year to year. This has been accomplished until recently by accepting the DeBeers-led marketing arrangement where production, not price, is adjusted to meet changes in world demand.14 Further, the nature of the Botswana Government’s profit-sharing arrangement with DeBeers has meant that the bulk of the volatility of diamond earnings has been absorbed by government, not by mineral sector employment. In contrast, the well-known volatility of world prices of gold and platinum has been absorbed by South Africa’s producers. The recent high world prices for gold have been a lifeline for South African miners facing depletion of deposits and rising extraction costs. Another major difference between Botswana and South Africa has been the exchange rate policy. Since Botswana’s withdrawal from the Rand CMA, the Pula exchange rate has been fixed, first against the Rand, then against the US dollar, and for some time now against a basket consisting of the Rand and international currencies. The Pula exchange rate is currently a crawling band regime, with a daily crawl against a basket of the Rand and the SDR, in which the Rand continues to have a heavy weight.15 South Africa’s exchange rate policy, since the collapse of the Bretton Woods system, has been a float, modified for two extended periods by a dual exchange rate arrangement to separate non-resident capital account transactions from the rest. In addition, exchange controls on residents have been used with varying degrees of severity, some of which remain to this day. The floating exchange rate has seen the Rand move from less than one Rand per US dollar in the 1970s and early 1980s to 13 per US dollar in December 2001, and then recover to less than
J. Clark Leith 223
six per US dollar in early 2005 before settling around seven per dollar at the time of writing. This volatility of the Rand against the US dollar and, more generally, the major currencies of the world, has had a serious impact on the members of the Rand CMA. While the rapidly falling Rand was welcomed as an engine to promote export growth, the subsequent recovery of the Rand has adversely affected many export sectors, including gold and textiles. This impact has been particularly serious for Lesotho, which has long sent a significant portion of its labour force to work in South African mines, and which has been relying on growth of textile exports to replace migrant worker employment. For Botswana, given the Pula exchange rate basket in which the Rand has a heavy weight, the foregoing volatility of the Rand against the major world currencies, has meant considerable volatility of prices of exportables (largely denominated in US dollars) relative to importables (largely denominated in Rands), and domestic goods and services (denominated in Pula). Given Botswana’s trade pattern, there is no way for Botswana to avoid this volatility. If Botswana were part of the Rand CMA, the volatility would appear largely in the export sector. If Botswana were pegged to the SDR, the volatility would appear largely in the import sector. The weights in the basket simply determine the distribution of the volatility: the heavier the weight of the Rand, the more the volatility is borne by the Pula/US dollar exchange rate. Botswana has relied heavily on shock absorbers to adjust to the disturbances outlined above. First and foremost, substantial Government balances in the central bank, and foreign exchange reserves managed by the central bank, have enabled both Government expenditure and foreign payments to continue with scarcely a hiccup in the face of the disturbances. In the 1980s, this was particularly important in dealing with the effect of droughts, which adversely affected the food supply of a substantial portion of the population, and the ups and downs of diamond sales. Since the establishment of an independent currency in 1976, the Botswana Government has never had to borrow from the central bank to finance its expenditures, and short-term borrowing on international markets to supplement the foreign exchange reserves has never been required.16 The more usual adjustment mechanisms identified as important for optimum currency areas, such as labour mobility and fiscal safety nets, do not exist as part of either the Rand CMA or the SACU. Hence, if Botswana were to rejoin the Rand CMA it would be critical for the Botswana Government to maintain a significant reserve as its own fiscal safety net.
224
Monetary Union Proposals in North America and Southern Africa
Case of Botswana: Policy autonomy and accountability If Botswana were to rejoin the Rand CMA (or if SADC were to establish a currency union), there would be significant changes in the monetary order. For Botswana, it would certainly mean giving up the exchange rate as a policy instrument, and would mean abandoning national decision-making authority on monetary policy. The loss of policy autonomy would be substantial. Botswana has been able to offset the squeeze on US dollar and Euro export earnings that befell industries in the CMA as the Rand appreciated dramatically from early 2002 to 2006. Botswana achieved this in part by the Pula exchange rate basket which gives a significant weight to the SDR, and in part by devaluing when the overvaluation of the Pula became serious. Something akin to the Grubel (1999) argument, that the exchange rate has been used to alleviate the adverse effects of government policy, has arisen in the Botswana policy debate. Some have argued that allowing the Pula exchange rate to adjust has given the Government an easy way out from its failure to promote rapid productivity growth, particularly among various state-owned public utilities. The validity of that argument depends crucially on what the counter-factual might have been. The example of provincially- and state-owned public utilities in federal countries suggests that the Botswana state-owned public utilities would have behaved much the same with or without a wider monetary union. There are, however, two more subtle issues which arise from the policy autonomy question. First, for a small open economy like Botswana, policy autonomy for a central bank may be attempting more than is feasible. Second, the development of a capital market in a way that contributes to sustained rapid growth may be constrained by the size of the local currency financial market. One of the central features of a small open economy is substantial and rapid pass-through of foreign prices to the domestic market for tradeables. The implication of this for monetary policy is that if policy autonomy is to be used by the central bank, it should target non-traded goods prices, rather than attempt to achieve a CPI inflation target.17 If these shocks to world prices of tradeables pass through to domestic prices of tradeables, and are met by a monetary policy response, real output instability would follow. (See Devereux, Lane, and Xu, 2006.) But if the central bank insists on following the worldwide central bank fashion of CPI inflation targeting, the exercise of policy autonomy would be perverse. The economy would be better off without policy
J. Clark Leith 225
autonomy, and one of the principal gains from remaining outside a common monetary area would disappear. As an economy becomes more developed, it is widely recognised that its financial markets develop greater breadth and depth. There is also a literature dating back to McKinnon (1973) and Shaw (1973) that suggests that the development of financial markets contributes significantly to economic growth. It is this simultaneous feedback from economic development to financial market development, and from financial market development to economic development that creates the potential for a virtuous circle of growth. However, one of the essential characteristics of the market for any financial product is the liquidity of that specific market. If trades are few and far between, and buy/sell margins are wide, the role of the financial system in allocating funds from low productivity uses to higher productivity uses is thwarted. Yet where a particular small economy has its own currency, its financial market is inherently small and cannot offer the full range of instruments found in larger financial markets. Thus, one of the gains for a small country joining a monetary union would be access to a larger more sophisticated financial system. Whether or not more changes might be involved would depend on the specifics of the new monetary order. The simplest case would be for Botswana to rejoin the Rand CMA. The current Rand CMA is a relatively loose monetary union, with the Reserve Bank of South Africa de facto determining goals for, and day-to-day implementation of, monetary policy. Nevertheless, unlike the Euro zone, each of the smaller members has its own central bank, issues its own currency, and conveys the public impression that it pursues its own independent monetary and exchange rate policies.18 Accountability of the South African Reserve Bank, enshrined in the constitution, is via regular consultation between the Reserve Bank and the minister responsible for finance. In addition, the Reserve Bank’s annual report is submitted to the South African parliament, and there are periodic discussions with the parliamentary standing committee on finance. Effectively, the only influence that the smaller members of the Rand CMA currently have in either setting monetary policy goals, or in calling the South African Reserve Bank to account, is via meetings of the four central bank governors prior to the meetings of the South African Reserve Bank’s Monetary Policy Committee. In 2005, the governors of the Rand CMA central banks commissioned a study of the benefits and costs of the creation of a common central bank. Apparently the study was favourable to the creation of a Euro-style central bank, and awaits decisions by the national governments.19
226
Monetary Union Proposals in North America and Southern Africa
This suggests that the smaller members are not even considering withdrawal. Rather, it appears that to the smaller members, the balance of benefits in the form of reduced transaction costs, successful inflation targeting, and the prospect of some increased role in monetary policy decision-making, outweigh the costs of lost policy autonomy. This new situation is relevant for Botswana’s choice in the next decade. As Laidler (2003) notes, Canada’s choice re a NAMU is quite different from Britain’s choice re the European Monetary Union (EMU). The nature of the monetary order for Canada in a NAMU would have to be negotiated, and the prospect for a significant Canadian voice and accountability are slim. The EMU, on the other hand, is a well-defined and fully functioning monetary order within the European Union to which Britain is already fully committed. For Botswana, already a long-time member of the SACU, the successful achievement of inflation targeting by the South African Reserve Bank, and the creation of a Rand CMA central bank in which all members have a voice could conceivably tilt the balance in favour of rejoining the Rand CMA. Supervision of financial institutions in the Rand CMA countries is currently left to individual member countries. However, a substantial expansion of the Rand CMA, with many financial institutions operating in several member countries, would increase the potential for contagion if a financial institution were to collapse. This might suggest creation of a common organisation to supervise financial institutions in the Rand CMA, as was done in the CFA franc zone countries following the 1994 devaluation. The potential for effects on fellow members would also apply to borrowing by governments, thereby suggesting the need for rules covering that dimension. For Botswana in a Rand CMA, this would be superimposed on the routine consultation between the central bank and the finance ministry that is tasked with coordination of fiscal, monetary, and exchange rate policies. Of course, if Botswana were part of the Rand CMA, coordination between the central bank and the finance ministry on the exchange rate itself would cease only to be replaced by consideration of the real effective exchange rate. Most important of all, if Botswana were to join any currency union, the mechanisms for setting monetary policy goals and the political accountability of policymakers for their performance would change. Botswana’s central bank has considerable de facto autonomy in setting monetary policy, with an annual monetary policy statement drawn up by the Governor and her staff and approved by the Board. The legal framework provides for the annual report of the central bank to be
J. Clark Leith 227
sent to the Minister responsible for finance within three months of the end of the year, and which the Minister must then lay before parliament. The Act also provides for the possibility that the Minister may ask the President to determine the policy to be followed by the central bank.20 However, this provision has never been used in the 30 years since the currency was created. Exchange rate policy is also determined by the President ‘on the recommendation of the Minister after consultation with the Bank’.21 While there is room for improvement in Botswana’s accountability mechanisms, an important feature stands out. Accountability is firmly rooted in the domestic polity.
Conclusion Both Canada and Botswana face a similar dilemma. Each has a much larger neighbour, with which substantial transactions take place. The gains from a currency union in the form of reduced transaction costs would be significant. Those gains would be offset by restricted autonomy in setting monetary and exchange rate policies, both of which have been important in absorbing asymmetric shocks. To the extent that policy autonomy has yielded negative outcomes, the reduced policy autonomy might be regarded as a gain, rather than a cost. This does suggest some contrasts between Botswana and Canada. For example, while no one could reasonably argue that the authorities in Botswana require an agency of restraint to curb profligate spending, a monetary union would block Botswana’s central bank from following an ill-suited monetary model. The absence of a monetary union has also limited access by economic agents in Botswana to the more sophisticated South African financial sector. While neither of these problems necessarily requires a monetary union to solve, rejoining a reformed Rand CMA could prove to be a path of least resistance for Botswana. On balance, the economics of the Botswana case does not weigh decisively either for or against monetary union, as Laidler found in the Canadian case. The question of accountability for goals and outcomes thus becomes decisive. A monetary union, by its very nature, turns over to the broader union responsibility for setting goals, and accounting for success or failure of policy in pursuit of those goals. For Canada, the potential for significant influence in setting the goals or in calling the monetary authorities to account would be extremely limited, and therefore the verdict is firmly against a monetary union. For Botswana, if the Rand CMA becomes an EMU-style monetary union, with well-defined roles
228
Monetary Union Proposals in North America and Southern Africa
for the smaller members in setting goals and enforcing accountability for achieving goals, then the verdict would no longer be firmly against such a union.
Notes The author is indebted to Keith Jefferis and the conference participants for valuable comments on the initial version. 1. Some of his relevant papers are Laidler (1999), Laidler (2003), and Laidler and Robson (2002). 2. BIS Paper No. 17 (2003) reveals the wide range of regions where regional currency areas are on the agenda. 3. The Rand Common Monetary Area was formerly called simply the Rand Monetary Area. To avoid confusion, I will refer simply to the Rand Common Monetary Area throughout, regardless of what the formal name was at a particular historical date. 4. For a thorough review of possible evolution of African exchange rate and monetary regimes, see Masson and Pattillo (2005). For a quantification of the distribution of the likely benefits and costs of a West African monetary union, see Debrun, Masson, and Pattillo (2005). Masson (2006) convincingly shows that the expansion of trade attributable to a currency union would not reverse the net losses from most African currency unions. 5. The goals are set out in SADC ‘Regional Indicative Strategic Development Plan’, Section 4.10.5 (Targets), and reaffirmed at a SADC Summit, 18 August 2006. Jefferis (2006) reviews the issues facing any monetary union of the SADC region. 6. The starting point for the modern literature is Mundell (1961). A helpful survey of the literature in the mid-1970s is Tower and Willett (1976). A more recent study that includes a brief summary of the highlights of the literature is Bayoumi and Eichengreen (1994). 7. The costs need not disappear entirely. Indeed, a common currency does not necessarily mean that it is possible to carry out transactions at lower costs between different parts of a common monetary area than between monetary areas. Thus, it may be more costly to send money to one’s family in a distant village within the same monetary area than to move the same amount between the financial centres of different monetary jurisdictions. 8. As Laidler (2003) points out, economic exchange is based on differences, which implies that the potential monetary partners are unlikely to constitute an optimum currency area. 9. Laidler and Robson (2002), p. 12. 10. For more detail on the economy of Botswana see Leith (2005). 11. Botswana is landlocked, with most transport avenues passing through South Africa. Many goods imported into Botswana are identified as coming from South Africa but in fact have a relatively low South African content. Because of the SACU common external tariff, there is no particular reason to distinguish what portion is of non–South African origin. What is relevant for present discussions is that imports from South Africa are invoiced in Rands.
J. Clark Leith 229 12. Shares of value added by sub-sector are not available, but export data are. In 2005, diamonds were 85.8% of major mineral exports, with copper-nickel at 11.6%. Gold and soda-ash made up most of the rest. 13. South Africa’s top three exports at the four-digit SITC level in 2005 were platinum (10.2% of total), gold (8.9%), and coal briquettes (6.3%). The high share of platinum is recent. In 2003, gold was the largest at 12.6%, and platinum was only 3.1%. 14. For a good description of the international diamond market, see Spar (2006). 15. The crawling band arrangement was introduced at the end of May 2005, although the band between the central bank’s buy/sell rate is still relatively narrow. 16. Government balances in the central bank at end 2005 amounted to 24% of GDP while foreign exchange reserves were 64% of GDP. 17. In addition, some non-traded goods and services prices may be regulated and are therefore not affected directly by monetary policy. 18. For example, the central banks’ web sites include reference to their own monetary policies and exchange rate policies. 19. T. T. Mboweni, Governor of the South African Reserve Bank, referred to the study as follows: ‘A study was conducted in 2005 under the auspices of the CCBG [Committee of the Central Bank Governors] outlining the costs and benefits of a common central bank for the CMA countries. However, decisions in this regard will be taken by the political leaders of the CMA countries, rather than by central bankers.’ (Speech, 7 October 2006). 20. Bank of Botswana Act, Section 54. 21. Bank of Botswana Act, Section 21.
References Bank for International Settlements. 2003. ‘Regional Currency areas and the use of foreign currencies’, BIS Papers, No. 17 (May), Basel. Bayoumi, T. and B. Eichengreen. 1994. One Money or Many: Analyzing the Prospects for Monetary Unification in Various Parts of the World, Princeton Studies in International Finance, No. 76 (September). Collier, P. 1996. ‘The Role of the African State in Building Agencies of Restraint’, Ch. 12 in New Directions in Development Economics, ed. M. Lundahl and B. J. Ndulu, Routledge, London. Courchene, T. J. 1999. ‘Alternative North American Currency Arrangements: A Research Agenda’, Canadian Public Policy, XXV, No. 3 (September) 308–14. Courchene, T. J., and R. G. Harris. 1999. ‘From Fixing to Monetary Union: Options for North American Currency Integration’, C. D. Howe Institute Commentary, No. 127 (June). Debrun, X., P. Masson, and C. Pattillo. 2005. ‘Monetary Union in West Africa: who might gain, who might lose, and why?’, Canadian Journal of Economics, Vol. 38, No. 2 (May) 454–81. Devereux, M. B., P. R. Lane, and J. Xu. 2006. ‘Exchange Rates and Monetary Policy in Emerging Market Economies’, Economic Journal, Vol. 116, No. 511 (April) 478–506.
230
Monetary Union Proposals in North America and Southern Africa
Grubel, H. G. 1999. ‘The Case for the Amero: The Economics and Politics of a North American Monetary Union’, Critical Issues Bulletin, Fraser Institute (September) 50 pp. Harris, R. G. 1999. ‘The NAMU Debate’, Canadian Public Policy, XXV, No. 3 (September) 320–23. Helleiner, E. 2006. Towards North American Monetary Union: The Politics and History of Canada’s Exchange Rate Regime, McGill-Queens, Montreal. Jefferis, K. R. 2006. ‘The Process of Monetary Integration in the SADC Region’, Journal of Southern African Studies. Laidler, D. 1999. ‘Canada’s Exchange Rate Options’, Canadian Public Policy, XXV, No. 3 (September) 324–32. —— 2003. ‘Canada’s Monetary Choices in North America and Britain’s in Europe – Economic Parallels and Political Differences’, EPRI Working Paper Series, 2003-03 (November) UWO. Laidler, D. and W. B. P. Robson. 2002. ‘No Small Change: The Awkward Economics and Politics of North American Monetary Integration’, C. D. Howe Institute Commentary, No. 167 (July). Leith, J. C. 2005. Why Botswana Prospered, McGill-Queens, Montreal. Masson, P. 2006. ‘Currency Unions in Africa: Is the Rose Effect Substantial Enough to Justify Their Formation?’ mimeo. Masson, P. and C. Pattillo. 2004. The Monetary Geography of Africa, Brookings, Washington, DC. Mboweni, T. T. 2006. ‘South Africa’s Financial Markets Within The Southern African Sub-Region’, speech delivered on 7 October in Windhoek, Namibia. McKinnon, R. I. 1973. Money and Capital in Economic Development, Brookings, Washington, DC. Mundell, R. A. 1961. ‘A Theory of Optimum Currency Areas’, American Economic Review, Vol. 51 (September) 657–65. Shaw, E. S. 1973. Financial Deepening in Economic Development, Oxford University Press, New York. Spar, D. L. 2006. ‘Markets: Continuity and Change in the International Diamond Market’, Journal of Economic Perspectives, Vol. 20, No. 3 (Summer) 195–208. Southern African Development Community. Undated. Regional Indicative Strategic Development Plan, Gaborone. Tower, E., and T. D. Willett. 1976. The Theory of Optimum Currency Areas and Exchange Rate Flexibility, Princeton Special Papers in International Finance, No. 11 (May).
Name Index Allen, W. 43 Atkinson, P. 175 Bagehot, W. 61, 81 Bailey, M. 156 Ball, E. 198 Baring, F. 81 Blomqvist, Å. 122 Bouey, G. K. 11, 51, 62 Brittan, S. 193, 194, 198 Brown, G. 196–8 Burns, T. (Lord) 189, 193 Caygill, D. 177 Chretien, J. 75 Clark, A. 42–3, 50 Clarke, K. 196, 197 Clouston, E. 90 Coyne, J. 7, 9, 45 Crow, J. 11, 14, 17, 23, 190 Dalziel, P. 177 Deane, R. 174, 175 Dodge, D. xii, 26 Douglas, R. 172–3, 176, 177, 200
Holmes, F. 175 Howe, G. 179 Howitt, P. 73–9 Johnson, J. 89 Kaldor, N. 180, 181 Keynes, J. M. 44, 47, 63 King, M. A. 37, 195 Kydland, F. 64 Laidler, A. viii–ix Laidler, D. viii–xii, 1–2, 4, 12, 22, 30, 35, 41, 42, 65, 73–9, 80, 96, 123, 156–70, 171, 172, 174, 176, 215–19, 226, 227 Lalonde, M. xii Lamont, N. 185, 187, 189, 190, 191, 192, 194, 196 Lange, D. 172 Laurier, W. 103 Lawson, N. 178, 179, 185, 186, 187, 188, 192, 193 Ledingham, P. 174–5 Leijonhufvud, A. 4 Lucas, R. E. 165
Edgeworth, F. 44 Fischer, S. 74 Forder, J. 198 Freedman, C. viii, 37 Friedman, M. 51, 61, 86, 199 George, E. 189, 195, 196, 197 Gladstone, W. E. 159 Goodhart, C. A. E. 52, 55, 82, 171–214 Harberger, A. 156 Healey, D. 179 Heath, E. 181 Hogg, S. 189
Major, J. 187, 189, 192, 196, 202 Macdonald, J. 103 Martin, P. 17, 75 McChesney Martin, W. 61 McKinnon, R. I. 225 Meltzer, A. H. 171, 172 Muldoon, R. 172, 173, 178 Neilson, P. 176, 177 Norman, P. 190 Nordhaus, W. 184 Oppenheimer, P. 198 Pigou, A. C. 47 Plender, J. 196 231
232
Name Index
Prescott, E. 64 Rasminsky, L. Richardson, R. Richardson, G. Robson, W. A. Roll, E. (Lord)
9, 45, 62 178 179 xii 195–6
Schwartz, A. J. 86 Shaw, E. S. 225 Sheppard, D. 177, 178 Simons, H. 54 Smith, A. 62
Thatcher, M. 181, 185, 186, 188, 192, 202 Thiessen, G. xii, 17, 22 Thornton, H. 81, 96 Tobin, J. 127 Towers, G. 7 Volcker, P. 181 Vreeland, E. 90 Walker, E. 87 Whitwell, J. 177 Wicksell, K. 44, 46
Subject Index African Union countries 215–30 Association of South East Asian Nations (ASEAN) 100–21 Bank of Canada vii, ix, x, xii, 3, 6, 7–32, 36–8, 46, 53–60, 73–9, 91–3, 209–14 Bank of England 81, 95, 178 Bank runs 82, 86, 91 Barings bank 82 Black Wednesday (1992) 188 Botswana 215–30 Bretton-Woods system 7, 180, 222 Canada Health Act 124, 128, 129, 138, 140–50 Canadian trade policy 100–71 Canadian health care 122–54 Canadian banking system 80–99 Canadian Bankers’ Association 85 Canadian Economics Association xii Canadian trade policy 30–2, 100–21 Capitation 131–5 Carter Commission (Canada) 159 C. D. Howe Institute viii, x, xii, 35 China 100–21 Conservative governments, 1979–97 (UK) 135, 178 Currency School 175 C. D. Howe Institute 35 De Beers 227 Dollarisation 31 Donner Prize xii, 35 European Monetary Union (EMU) 226, 227 Exchange Rates 6–14, 25, 30, 180, 222–4 Expectations 21, 41 Federal Reserve System 82, 83 Fiat money 1, 26
General Agreement on Tariffs and Trade (GATT) 104 Gold Standard 6 Goodhart’s Law 76–7 Goods and Services Tax (GST) 64, 74 Harberger triangles 160 Health care systems 122–55 Housing policy 156–70 Imminent failure hypothesis 88–9 India 100–21 Inflation targeting x, 1, 2, 14–21, 41–72, 73–9, 171–214, 224 International comparisons of health systems 130, 131, 136 Lazy manager hypothesis 218 Lender of last resort 44, 80–99 Long Term Capital Management 82 Lucas critique 50 Macdonald Commission xii, 105 Market failure 124–6, 147–9 Medicare 130, 142, 144 Monetary unions 215–30 Monetary Policy Committee 35 Monetary gradualism 51 National Monetary Commission 90 New Keynesian economics 58 New Zealand economic policy North American Free Trade Agreement (NAFTA) 100–21 Organisation for Economic Cooperation and Development (OECD) 14, 50, 106 Phillips curve 19, 49, 56–64, 199 Pigou effect 47 President’s Advisory Panel on Tax Reform (US) 162–3 233
234
Subject Index
Progressive Conservative Government (Canada) 7–9
South Africa 219–30
Quality-Adjusted Life Years (QALYs) 146–7
Tax reform 162 Tax-deductible interest payments 160–3, 168–70 Time inconsistency 184
Radcliffe Committee 62 Reserve Bank of New Zealand 171–214 Royal Society of Canada xii
University of Western Ontario x, xii Vector auto-regressions (VARs) 184
Singapore 100–21 Smoot-Hawley tariff 104
World Trade Organization (WTO) 115