Exchange rate flexibility, target zones, and policy coordination

Exchange rate flexibility, target zones, and policy coordination

World Development, Vol. 15, No. 12, pp. 1437-1443, 1987. Printed in Great Britain. 0305-750X/87 $3.00 + 0.00 (~) 1987 Pergamon Journals Ltd. Exchang...

633KB Sizes 1 Downloads 86 Views

World Development, Vol. 15, No. 12, pp. 1437-1443, 1987. Printed in Great Britain.

0305-750X/87 $3.00 + 0.00 (~) 1987 Pergamon Journals Ltd.

Exchange Rate Flexibility, Target Zones, and Policy Coordination JOHN WILLIAMSON*

Institute for International Economics, Washington, D C 1. I N T R O D U C T I O N At its meeting in April 1986, the Interim Committee "asked the [Fund's] Executive Board to consider further whether there are any modifications in the exchange rate system that could contribute to enhancing exchange rate stability and the mutual consistency of economic policies without sacrificing the essential flexibility of the system" (Interim Committee, 1986, para 5). This was interpreted by a part of the press as an official rejection of target zones, which was odd inasmuch as the target zone proposal can be regarded as a technique to limit misalignments and enhance policy consistency while retaining the very real benefits that exchange rate flexibility can bring. The present paper seeks to explain why I regard this intermediate regime between fixed and floating exchange rates as offering the most attractive combination of the features of the two systems. Section 2 of the paper explains briefly what I understand by the target zone proposal. Section 3 explains how the four dimensions of flexibility built into the proposal could satisfy all the real social functions that floating can further. Section 4 argues that the proposal could also help both to limit misalignments and to secure policy consistency. Section 5 explores the relation of the target zone approach to proposals for "objective indicators."

2. T H E T A R G E T Z O N E P R O P O S A L The target zone proposal envisages a limited number of the major countries negotiating a set of mutually consistent targets for their effective exchange rates. The minimum number of countries needed for a meaningful system would be the three biggest: the United States, West Germany, and Japan. Current proposals for policy coordination involve rather more countries: the Group of Five includes the two other countries

with currencies in the SDR, namely France and the United Kingdom, while the Tokyo summit added Canada and Italy, to make the Group of Seven. It appears to be envisaged that the Group of Seven will be supplemented by the Managing Director of the IMF when meeting to agree on targets and policy adjustments called for by deviations from the targets, so as to provide both a neutral referee and a spokesman for the interests of the smaller countries, including the developing countries, in the decisions of the major countries on policy coordination. 1 The exchange rate targets should be selected so as to reconcile on average over the medium term what James Meade (1951) termed internal and external balancefl By internal balance is meant the highest level of activity and employment consistent with the satisfactory control of inflation. External balance may be defined as a current account balance that matches the underlying capital flow that reflects thrift and productivity. Neither concept is unambiguous, but they do offer the sort of technical criteria that are needed as the foundation for any serious attempt to achieve agreement on internationally compatible policy objectives. Nominal exchange rate targets should be adjusted automatically to compensate for differential inflation so as to prevent the targets becoming outdated as a result of inflation. The participating countries would be expected to conduct their macroeconomic policies with a view to limiting deviations of their exchange rates from the agreed targets, and particularly with a view to preventing exchange rates going outside a broad zone (perhaps + 10%) around their targets. If an exchange rate threatened to breach the edge of its zone, the participants would meet in order to review their policies and decide which © 1987 Institute for International Economics, Washington, DC. *The author is indebted to C. Fred Bergsten, Shafiqul Islam, and Stephen Marris for comments on a previous draft.

1437

1438

WORLD DEVELOPMENT

country or countries should take what actions with a view to returning exchange rates toward the agreed targets. The principal instrument that should be used for exchange rate management is monetary policy (for the simple reason that it works), reinforced by jawboning and intervention in the exchange markets. If the resulting monetary policy threatens internal balance, fiscal policy should be adjusted appropriately; one may distrust the feasibility of a return to "'fine tuning'" while still recognizing that broad consistency of fiscal policy is essential to the preservation of domestic stability. One possible reaction to a threatened breach of a target zone would be to change the zone. It would make a mockery of the system if countries were entitled to change their zones just because the market were threatening to breach the zone, but the possibility that the market is conveying information about a change in the fundamentals that calls for a change in the target, or about a past error in interpreting the fundamentals, should always be borne in mind. Where either factor is present, it is appropriate to authorize a change in the target. Where a change in the zone is not authorized, however, one or more countries should be expected to modify their policies with a view to pushing rates back toward their targets. If only one currency is significantly out of line with its target, it is natural to assume that the country whose currency is misaligned would be the one to act. But if two or more currencies are simultaneously close to, or beyond, the edge of their target zones, it would be necessary to decide which of the countries involved should act. Four solutions can be envisaged: (1) a McKinnon-type rule whereby the strongcurrency countries would act to reduce interest rates if aggregate monetary expansion was below some predetermined target rate, and the weak-currency countries would raise interest rates in the converse case (McKinnon, 1984); (2) a commodity-type standard whereby the strong-currency countries would act to reduce interest rates if the price of a basket of primary commodities was falling, and the weak-currency countries would act in the converse case; (3) a regime of discretion, whereby the strongcurrency countries would act if it were judged that deflation posed a more serious global threat than inflation, and the weakcurrency countries would act in the converse case; (4) a rule of automatic sharing, whereby both countries would be expected to act in pro-

portion to the deviation of their rates from the targets. At one time I had a great deal of sympathy with the first proposal, but the increasing evidence of instability in the demand for money (on any and all concepts) has eroded the attractions of this approach. Despite its consistency with the popular criterion of symmetry, I would reject the fourth approach as one which would neglect a worthwhile opportunity of strengthening the defenses against global inflation and deflation. This leaves a choice between the second and third approaches; I doubt whether an automatic commodity-price rule would be wise, but the question as to whether it might provide a useful presumptive guide to a discretionary regime deserves careful thought. Note that I do not even list a s worthy of consideration the old asymmetrical practice of Bretton Woods in which ( n - l ) countries adjusted to whatever the largest country in the system happened to choose to do. A central purpose of international monetary reform is to institutionalize the practice of the United States taking account of the impact of its actions on the rest of the world, and that purpose would be prejudiced from the start by the asymmetrical solution. Two other characteristics of target zones as I envisage them merit mention. The first is that I strongly believe that the targets should be published, both to provide a focus for stabilizing speculation and to ensure that policymaking is subjected to critical public discussion. The second is that I favor "soft buffers" to the target zones, meaning that the authorities should not be subject to an absolute commitment to maintain rates within the agreed zones. There are again two motivations for this. One is to allow the participating governments time to reflect whether a change in zones is called for when market pressures carry rates to the edges of the zones, without forcing them to commit resources in intervention that could recreate the old problem of the one-way bet if the equilibrium rate had in fact moved outside the zone. The other purpose is to remove any danger that the rules of a target zone system would encourage a country with an excessive fiscal deficit (like the United States in recent years) to monetize the deficit and rekindle inflation if the exchange rate hit the top of the zone. Under this circumstance the government in question would be pressured by its peers to change its policy mix, tightening up on fiscal policy while relaxing monetary policy, so as to push the exchange rate back toward its target without jeopardizing internal balance. But if the "quasi-crisis" induced by breaching the zone did not serve to muster the

EXCHANGE RATE FLEXIBILITY political will to restore fiscal probity, the least bad option would be to have the currency appreciate beyond its target zone, as the soft buffer would permit. Maintaining the target zone even with the rate outside it would at least warn speculators that the intent of policy would be to push the rate back to a realistic level as soon as this could be done safely, which should serve to discourage the sort of speculative bubble that by all appearances was responsible for the dollar ascending far above what could be justified by the fundamentals from mid-1984 to September 1985. 3. T H E S O C I A L FUNCTIONS O F FLEXIBILITY A first function that exchange-rate flexibility can perform is to reconcile differential inflation rates. A country with faster inflation can depreciate in order to prevent its tradable goods sectors from becoming uncompetitive and a payments deficit developing, while a country with slower inflation can repel imported inflation. This function is satisfied by the target zone approach, since the nominal targets would be routinely adjusted as new inflation figures were published in order to preserve the same real targets. Critics sometimes charge that this implies that the system would automatically accommodate inflation. It is true that the exchange-rate mechanism would not be used to resist inflation, but countries have other means than pegging the exchange rate to control inflation, notably domestic demand management policy. In fact, a country that wished to use real appreciation as a weapon to counter inflation could still do so within the limits set by the target zone. The one thing that would be ruled out is the pretence that inflation was being controlled while in fact it was being temporarily exported as the exchange rate became progressively more overvalued, a policy course that has often (Britain 1967, Argentina 1982) ended in a disastrous intensification of inflation. A second function of exchange rate flexibility is to facilitate the adjustment process by changing the incentives to export and import when payments adjustment proves necessary. This function can also be fulfilled by a target zone system, by virtue of the ability to change real targets in the light of real shocks to the balance of payments. A third function of exchange rate flexibility is that of liberating monetary policy to pursue targets at variance with those in the rest of the world. If one country is suffering a deeper reces-

1439

sion than its partners, it may legitimately wish to ease monetary policy relative to other countries, and that will be feasible only if its currency can depreciate so as to create an expectation of a subsequent rebound that will compensate investors for the temporarily low interest rates. Conversely, a country with abnormally severe inflation may legitimately seek to raise interest rates temporarily, which will require an appreciation. One of the two functions of the wide zone 3 is to give countries latitude for such divergences in anticyclical monetary policies (McKinnon, 1971). Admittedly the existence of edges to the zone does limit this freedom, though it can also be argued that the existence of credible targets will help focus expectations of future exchange rates and in that way increase the amount of interest rate divergence consistent with a given depreciation or appreciation. The final legitimate function of exchange rate flexibility is to absorb speculative pressures. As explained above, this is one of the two purposes of the "soft buffers" to the zones. At times, a temporary move of the exchange rate beyond its zone may be less damaging than the distortion of monetary policy that would be necessary to prevent it. My list of the legitimate functions of a flexible exchange rate excludes that of letting the market decide what the exchange rate ought to be. I exclude it because of a conviction that markets have proved themselves inadequate to this task: in particular, because the speculative bubble that took the dollar well above what could be justified by rational forward-looking analysis in 1984-85 was broken only when governments resolved to break it. The explanation for the unsatisfactory performance of foreign exchange markets appears to be that almost all operators take a short-term view, many of them based on what is euphemistically known as "technical analysis," which can institutionalize extrapolative expectations. In the absence of official actors prepared to take a long-term view (at least two or three years, beyond the J-curve), the market lacks an anchor to long-term fundamentals. There is, in other words, a persuasive answer to the rhetorical question as to why government economists should be able to do a better job than the market of knowing where the exchange rate ought to be" they can ask the right question. 4 4. M I S A L I G N M E N T S A N D POLICY CONSISTENCY In my judgment the two great failings of floating rates have proved to be very different to any

1440

WORLD DEVELOPMENT

of the disadvantages that were discussed at length before the move to floating. One failing has been the propensity to generate misalignments, by which is meant deviations of exchange rates from their "fundamental equilibrium" levels, the rates that would reconcile internal and external balance in the medium run and that would therefore provide the targets under a target zone approach. The second failing has been the lack of pressure to coordinate macroeconomic policies among countries. Both failings have been damaging to developing countries as well as to the industrial countries directly involved. It is now generally agreed that misalignments have been huge under floating. According to my own estimates, the dollar was overvalued by over 50% at its peak in February 1985, and the pound by almost 50% in January 1981. These figures are at least twice as large as any misalignments that arose (through reluctance to change exchange rates in a timely way) under Bretton Woods. The costs of misalignments in the industrial countries involve adjustment costs as resources are shuttled back and forth between the tradable and nontradable sectors; the unemployment attendant on delays in initiating those adjustments; the loss of productive capacity in the tradable goods sectors in countries with overvalued currencies, excessive investment in those sectors in countries with undervalued currencies, and perhaps inadequate investment in total because of the additional uncertainty regarding the most economic location of investment; ratchet effects on inflation; and a stimulus to protectionism in countries with overvalued currencies (Williamson, 1985). The developing countries suffer directly from induced protection in the industrial countries and indirectly from additional inflation, unemployment, resource waste and any shortfall in investment in those countries, all of which tend to lower industrial country demand and thus the value of developing country exports. In addition, developing countries suffer from the shifting profitability of exporting to different markets as exchange rates change, dictating a need to redeploy their sales effort (e.g., from the United States to Europe and Japan as the dollar depreciates). They also suffer from the additional complications introduced into their own exchange rate policy. They can peg to a single currency, which makes it easy for their traders to engage in international transactions by invoicing and covering in that currency, but this poses a macroeconomic risk that an appreciation or depreciation of their peg currency will lead to a loss of competitiveness or to imported inflation. Alternatively they can peg to a currency basket reflecting their own trade flows, which largely eli-

minates the macroeconomic risks, but at the cost of subjecting their traders to microeconomic risks stemming from the lack of any predictable relationship between their national currency and a vehicle currency (Williamson, 1982). The second great failing of internationally unregulated floating has been the lack of pressure to coordinate policies. Without any exchange rate obligations or any formal mechanism of policy coordination, it was all too easy for the major countries to pursue policies that appeared attractive to themselves with no thought of their international ramifications. This has in fact been a disaster for the world economy, including the major countries themselves. It contributed to the severity of the recessions of 1975 and 1980-83, as well as to the lopsided nature of the subsequent recovery and the cripplingly high real interest rates that were a major factor in generating the debt crisis. It has now left the industrial countries with massive payments imbalances that will take years to correct. How much better it would be for the United States if it had been constrained from allowing the dollar to rise so high in 1982-85! A target zone approach would clearly address the first problem. It would make the avoidance of major misalignments a deliberate and important objective of policy, and would assign the tasks of limiting misalignments to a policy weapon with the clear capability of influencing exchange rates. It is also likely that a target zone approach would provide a mechanism for furthering policy coordination. The very process of agreeing on the exchange rate targets would require the participating countries to achieve a meeting of minds on the level of real activity that was to be considered normal and on the desirable pattern of current account balances. Once having achieved agreement on those matters, it would be strange if the subsequent monitoring of the target zones was to be focused solely on exchange rates. It would seem altogether more probable that, for example, a country that achieved its target exchange rate but that achieved a larger than planned current account surplus by virtue of a lower than assumed growth rate would be called to order. Similarly, a currency that threatened to break out of its target zone because of the need for high interest rates to offset the domestic effect of an over-expansive fiscal policy would advertise the unacceptable nature of that country's policy mix. The quasi-crisis that would be produced as the currency left its target zone and the international community diagnosed the cause as fiscal indiscipline should help the political system to recognize the necessity for a change in policy. Thus a target zone approach would provide a

EXCHANGE RATE FLEXIBILITY mechanism for furthering the policy coordination that is needed if the world economy is to function more satisfactorily in the future than it has in the past 15 years. That is not to say that it is the only approach that could further policy coordination. A possible alternative, the use of objective indicators, aroused interest at the April 1986 meetings of the Group of Ten and Interim Committee and was endorsed at the Tokyo Summit in May. 5. OBJECTIVE I N D I C A T O R S Two crucial questions about "objective indicators" need to be answered before it is possible to evaluate their potential role. The first question concerns what happens when a variable that has been designated as an indicator misses its target value. The second question is which variables are to be chosen for the role of indicators. On the first issue, at least four alternative reactions can be envisioned: initiation of consultations an obligation to adjust policies - - a presumption that policies should be adjusted - - a penalty. The weakest reaction is for the deviation of an indicator from its target value to be treated as a trigger to initiate consultations, the arrangement apparently envisaged by the Tokyo Summit. The need for policy changes would then presumably be analyzed jointly by the group of participating nations, 5 and peer pressure would be relied on to persuade countries to modify their policies in the direction that the analysis suggested to be called for. A much stronger system would involve deviations of the indicators triggering direct, prespecilied policy changes. For example, a common European interpretation of the reserve indicator proposal advanced by the United States during the Committee of 20 negotiations in 1972-74 was that excessive reserve accumulation should automatically trigger a revaluation. This embodies a large element of "automaticity," a property of international monetary arrangements specifically commended by US Secretary of the Treasury James Baker (1986) at the meeting of the Interim Committee in April 1986. An intermediate solution involves using indicators to create a presumption of the need for action. A good example is provided by my earlier discussion of an exchange rate breaking out of its target zone: this would create a presumption of the need for a more expansionary monetary policy (when the currency has appreciated), but this might have to be overridden if the country -

-

-

-

1441

were unable to secure a matching fiscal contraction and hence monetary expansion would threaten a revival of inflation. A quite distinct approach would be to use the indicator to trigger some form of penalty, what used to be referred to as "sanctions" and have more recently been called "pressures," a term revived in the most recent report of the Group of 24 (1985, paras 80-88). This implies that the indicator is being used to define a country's international responsibilities, or as what I have previously termed a "definitional indicator" (Williamson, 1977, Chap. 5), rather than as a "diagnostic indicator" as all three of the preceding approaches imply. How countries choose to react to the deviation of the indicator from its target value is their own decision, but the presumption is that their reactions will be influenced in the direction of international consistency by the recognition that they will incur a penalty of some type if they do not correct the deviation. This was the US concept of how a reserve indicator system would work: countries with excessive reserve accumulation would have recognized that adjustment action was in their own national interest because otherwise they would have incurred interest penalties and forfeited the right to convert currency balances into primary reserve assets. Of these four possibilities, by far the most promising appears to me to be the use of indicators on a presumptive basis. It is difficult to believe that indicators that merely trigger consultations, as seemed to be envisaged by the Tokyo Summit, will have a serious impact on the policies that countries pursue. On the other hand, the world is too irregular and unforseeable a place to persuade countries either to precommit themselves to certain policy reactions or to accept penalties if certain outcomes deviate from those hoped for. The presumptive use of indicators incorporates an element of automaticity without the risk of penalizing countries for not doing what everyone can agree would be silly. The second crucial question regarding a system of objective indicators concerns the set of variables that should be designated as indicators. The communiqu6 from the Tokyo Summit suggested no less than 10 such variables: the growth of GNP, inflation, the rate of interest, unemployment, the fiscal deficit, current account and trade balances, monetary growth rates, reserves, and the exchange rate. As a trigger for consultations there might be no objection to such an extensive list, but such a wide range of indicators would be ill-suited for any stronger role. The reason is that they constitute a vastly overdetermined system.

1~2

WORLD DEVELOPMENT

Suppose, for example, that a country agreed to targets for all 10 of these variables. If fiscal and monetary policy were on track, how could one demand an automatic or even presumptive policy adjustment (or impose a penalty) no matter how much some or all of the other variables deviated from their targets? Conversely, if growth, inflation, the balance of payments and the exchange rate were on track but fiscal and monetary policy were not, would one expect policies to be adjusted? In practice it is easy to anticipate what would happen: different indicators would point in different directions, and each country would then exploit the resulting ambiguity in order to continue doing precisely what it preferred without any concern for the international repercussions of its policies. An overdetermined set of indicators, like restricting the use of indicators to the role of triggers to initiate consultations, is a formula that could result in objective indicators being reduced to the same impotence as the Fund's multilateral surveillance of the past dozen years. The aim of policy coordination is to ensure that the fiscal and monetary policies chosen by different countries add up to a satisfactory global outcome. The strategic choice involved in selecting indicators is whether to attempt to commit countries to desired values of the policy variables themselves, or whether to commit them to desired outcomes. This choice raises all those issues that were debated at length in the Keynesianmonetarist controversy. Those who believe that the best policy is for governments to precommit themselves to fiscal and monetary policies will presumably wish to select indicators for those two variables rather than for desired outcomes, in the spirit (if not always the practice) of Margaret Thatcher's medium-term financial strategy. Those of us who have been disillusioned by monetarism because of the instability of the relationships that were supposed to be constant but proved not to be will prefer to seek indicators for outcomes rather than policy variables. The important objectives are growth (or its dual, unemployment); low inflation; and the current account balance. From the viewpoint of short-run monitoring, however, it may be better to concentrate on intermediate variables that are more readily observable and more quickly influenced by fiscal and monetary policy. It might be natural, for example, to select the growth of nominal income and the exchange rate as suitable intermediate variables in terms of which to monitor policies, while growth, inflation and the current account provided the ultimate goals whose behavior could provide a basis for agreeing to modify the intermediate objectives.

6. CONCLUDING REMARKS A sensibly constructed indicator system might therefore amount to an amplification of the target zone approach. Both approaches could hope to limit misalignments and both would provide a framework for enhancing policy consistency, while both could allow exchange rate flexibility to perform its important social functions of reconciling differential inflation, facilitating payments adjustment, permitting a measure of differentiation in the stance of anticyclical policies, and absorbing speculative pressures. The choice is not so much between target zones and objective indicators as it is between finding an acceptable synthesis of these two approaches and perpetuating the ineffectiveness of multilateral surveillance undisciplined by quantitative targets.

7. POSTSCRIPT This paper was written in May 1986, just before the Tokyo Summit went unexpectedly far in endorsing the indicator approach to international economic policy coordination. One of the two principal claims made above on behalf of the target zone proposal was that it would create pressures for more comprehensive policy coordination: since it seemed that the authorities were prepared to contemplate a comprehensive regime, the natural approach seemed to be to seek to embody the central ideas of the target zone proposal within such a regime. This task was first addressed in Edison, Miller and Williamson (1987) which developed an "extended target zone proposal" that supplemented an exchange rate target by a target for the growth in nominal income. It embodied three "assignment rules." One said that interest rate differences among countries should be governed primarily by the objective of limiting the deviation of exchange rates from their target values, which is very much in the spirit of the target zone proposal laid out above. A second said that the average level of world interest rates should be determined with a view to limiting the deviation of "world" nominal income growth from the aggregate of target rates of growth: this provides an implicit answer to the question "which country should adjust when more than one is at the limit of its target zone?" that was not foreseen in the preceding paper. The third rule suggested assigning fiscal policy in each country to its own nominal income target, thus formalizing the principle on this subject laid out above. The proposal is refined one stage further in

EXCHANGE RATE FLEXIBILITY Williamson and Miller (1987). Instead of a nominal income target, this suggests that the second intermediate target should be the rate of growth of demand. The difference between the growth in nominal income and demand is simply the change in the current account balance. The main purpose of the change is thus to establish that countries need to provide in their demand management policy for the target increase or decrease in their current account position, and to avoid coun-

1443

tries claiming success if, for example, their nominal income were to increase as planned simply because an undesirably large external surplus were to increase even further. Despite these substantial developments in the proposed system, the spirit is still very similar to that laid out above. Most of the analysis thus remains relevant to the debate as it stands in mid1987.

NOTES 1. This proposal is consistent with the realistic recognition of the Group of 24 that any effective surveillance mechanism will have to be run primarily by the major countries involved. "Although the role of the developing countries in influencing such an exchange rate system is necessarily limited - - and therefore the related mechanism of coordination and surveillance will be essentially concerned with the developed countries - - it is important . . . that this coordination should take account of the needs of developing countries" (Group of 24, 1985, para 66). 2. Meade's work was in its essentials anticipated by Keynes's (1923) classic Tract on Monetary Reform.

3. The other function is to make certain that the equilibrium rate does indeed fall within the target zone despite the considerable difficulties in estimating equilibrium rates accurately after a long period of instability in market rates. 4. The main papers to have developed this line of thought are Boughton (1984), Krugman (1985), and Frankel and Froot (1986). 5. This suggests that it will be necessary to call on the services of a sophisticated secretariat like the IMF staff, which provides a reason for seeking to bring this exercise within the framework of the IMF.

REFERENCES Baker, James A., "The functioning of the International Monetary System," Statement to the IMF Interim Committee (Washington, DC: US Treasury, 9-10 April 1986). Boughton, James M., "Exchange rate movements and adjustment in financial markets," IMF Staff Papers (September 1984). Edison, Hali J., Marcus Miller, and John Williamson, "On evaluating and extending the target zone proposal," Journal of Policy Modeling, Vol. 9, No. 1 (1987). Frankel, Jeffrey A., and Kenneth Froot, "'The dollar as a speculative bubble: A tale of fundamentalists and chartists," NBER Working Paper No. 1854 (1986). Group of 24, Report of the Deputies, IMF Survey (September 1985). Interim Committee, Communiqu6, IMF Survey (21 April 1986). Keynes, John Maynard, A Tract on Monetary Reform (London: Macmillan, 1923). Krugman, Paul, "'Is the strong dollar sustainable?" NBER Working Paper No. 1644 (1985).

Meade, James E., The Theory of International Economic Policy: Vol. L The Balance of Payments (London: Oxford University Press, 1951). McKinnon Ronald I., Monetary Theory and Controlled Flexibility in the Foreign Exchanges, Princeton Essays in International Finance No. 84 (1971). McKinnon, Ronald I., An International Standard for Monetary Stabilization (Washington, DC: Institute for International Economics, 1984). Williamson, John, The Failure of World Monetary Reform 1971-74 (New York: New York University Press, 1977). WiUiamson, John, "A survey of the literature on the optimal peg," Journal of Development Economics (1982). Williamson, John, The Exchange Rate System (Washington, DC: Institute for International Economics, revised edition, 1985). Williamson, John, and Marcus Miller, Targets and Indicators: A Blueprint for the International Coordination of Economic Policy (Washington, DC: Institute for International Economics, 1987).