three generations of crises, three generations of crisis models

6
Journal of International Money and Finance 22 (2003) 1089–1094 www.elsevier.com/locate/econbase Three generations of crises, three generations of crisis models Barry Eichengreen Economics Department, University of California at Berkeley, Berkeley, CA 94720-3880, USA Abstract This paper considers the three devaluation of the pound sterling in 1931, 1949 and 1967 as a window onto the recent theoretical literature of currency crises. # 2003 Elsevier Ltd. All rights reserved. Keywords: Currency crises; Devaluation It is a pleasure to give the Key Note remarks. 1 My first book, Sterling in Decline (Cairncross and Eichengreen, 1983), was on three currency crises: the devaluations of 1931, 1949 and 1967. Alec Cairncross and I began that study in 1979, a year better remembered as when Paul Krugman published his theory of balance pay- ments crises (Krugman, 1979). I recently went back to look at those three crises and found that they provide a very nice illustration of how the subsequent litera- ture on the topic has evolved. Taking the 1931, 1949 and 1967 sterling crises as the subject of my talk is self- indulgent, to be sure. But it is important to recall that before the debt crisis of the 1980s and the emerging market crises of the 1990s, the three balance-of-payments crises experienced by the United Kingdom were probably the three most important examples we possessed of the phenomenon with which this conference is con- cerned, namely, disruptive attacks on pegged exchange rates. So let me take you back time, starting with the 1931 sterling crisis. Recall that this was a period when sterling was pegged to gold and therefore to the dollar, and the world was spiraling downward into the Great Depression. Britain’s exports E-mail address: [email protected] (B. Eichengreen). 1 These remarks were delivered in the Conference on Regional and International Implications of Financial Instability in Latin America, University of California, Santa Cruz, 11 April 2003. 0261-5606/$ - see front matter # 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.jimonfin.2003.09.003

Upload: barry-eichengreen

Post on 19-Oct-2016

218 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Three generations of crises, three generations of crisis models

Journal of International Money and Finance

22 (2003) 1089–1094

www.elsevier.com/locate/econbase

Three generations of crises, three generationsof crisis models

Barry EichengreenEconomics Department, University of California at Berkeley, Berkeley, CA 94720-3880, USA

Abstract

This paper considers the three devaluation of the pound sterling in 1931, 1949 and 1967 asa window onto the recent theoretical literature of currency crises.# 2003 Elsevier Ltd. All rights reserved.

Keywords: Currency crises; Devaluation

It is a pleasure to give the Key Note remarks.1 My first book, Sterling in Decline

(Cairncross and Eichengreen, 1983), was on three currency crises: the devaluations

of 1931, 1949 and 1967. Alec Cairncross and I began that study in 1979, a year

better remembered as when Paul Krugman published his theory of balance pay-

ments crises (Krugman, 1979). I recently went back to look at those three crises

and found that they provide a very nice illustration of how the subsequent litera-

ture on the topic has evolved.Taking the 1931, 1949 and 1967 sterling crises as the subject of my talk is self-

indulgent, to be sure. But it is important to recall that before the debt crisis of the

1980s and the emerging market crises of the 1990s, the three balance-of-payments

crises experienced by the United Kingdom were probably the three most important

examples we possessed of the phenomenon with which this conference is con-

cerned, namely, disruptive attacks on pegged exchange rates.So let me take you back time, starting with the 1931 sterling crisis. Recall that

this was a period when sterling was pegged to gold and therefore to the dollar, and

the world was spiraling downward into the Great Depression. Britain’s exports

E-mail address: [email protected] (B. Eichengreen).1 These remarks were delivered in the Conference on Regional and International Implications of

Financial Instability in Latin America, University of California, Santa Cruz, 11 April 2003.

0261-5606/$ - see front matter# 2003 Elsevier Ltd. All rights reserved.

doi:10.1016/j.jimonfin.2003.09.003

Page 2: Three generations of crises, three generations of crisis models

were tailing off, and its foreign investment income was collapsing. The result, pre-dictably, was a steady loss of international reserves.But attempt to apply the Krugman model to the 1931 crisis, and you immedi-

ately come face to face with an important limitation of that model, namely, that itprovides no explanation for what governments are doing. While Krugman’s cur-rency speculators are smart—they maximize profits, making efficient use of all theavailable information—his governments and central banks are dumb. They followrigid policy rules, mechanically issuing domestic-currency-denominated debt tofinance constant budget deficits while mindlessly intervening to support thecurrency. Because the model makes no attempt to characterize the government’sobjectives, it offers no explanation for why the authorities react as they do.In fact, in 1931 the explanation for the latter was no mystery. The government

was preoccupied not just with the fragility of the financial situation but with thelevel of unemployment. When pressure on the currency was felt, unemploymenthad already reached 20 per cent of the insured labor force. In this context, raisingtaxes or cutting public support for the unemployed threatened to provoke a polit-ical backlash. While the government wished to avoid being tarred with a devalua-tionist brush, it did not relish having to implement economies that would onlyaggravate an already excruciating unemployment problem. In terms of the ‘‘secondgeneration models’’ of balance of payments crises that followed Krugman, whichadded optimizing governments to his framework (Ozkan and Sutherland, 1994;Eichengreen and Jeanne, 2000), Labour was trading off the fixed cost of devaluing,and thereby tarnishing its reputation for being able to govern, against the budget-ary economies needed to reassure the markets.In principle, the Bank of England could have tightened domestic credit more

aggressively to support sterling’s position on the foreign exchange market. Its fail-ure to do so is striking. After raising the rate from 2.5 per cent to 3.5 per cent onJuly 23rd and then to 4.5 per cent on July 30th (rather modest increases by modernstandards), it left monetary policy unchanged until the final surrender to the mar-kets on September 19th.The point is that neither the Government nor the Bank lacked instruments with

which to defend the currency; they were simply reluctant to use them. The politicalcontext is key to understanding this fact. These events took place less than a dec-ade after the Labour Party had formed its first government. They occurred in aperiod when Labour was in the ascendancy and the Liberal Party was in decline.But, given the continued electoral viability of both parties, neither of them couldmarshal a Parliamentary majority. The Labour Government formed in 1929 reliedon the Liberals for Parliamentary support, leaving it incapable of decisive action.And the Bank of England, while nominally independent, was sensitive to politicalconsiderations and in particular to the gravity of the unemployment problem.Unable to move decisively, the Labour Government fell even before the currency

peg collapsed. The peg was quickly abandoned by the newly-formed government ofnational unity, which could assign the blame to its predecessor. With no sterlingparity left to defend, the Bank of England was no longer reluctant to use interest

B. Eichengreen / Journal of International Money and Finance 22 (2003) 1089–10941090

Page 3: Three generations of crises, three generations of crisis models

rates to stimulate employment. Bank rate was cut to 2 per cent, a more appropri-ate stance for the deflationary circumstances of the time.In a sense, then, the 1931 devaluation is a classic illustration of a ‘‘second gener-

ation’’ balance-of-payments crisis model, in which the authorities trade off thecosts of fiscal austerity, which take the form of additional unemployment and adepressed economy, against the benefits of protecting their reputation for financialprobity. In this episode, the deepening slump ultimately sapped the government’sappetite for economies, and currency speculators acted on this fact.If the 1931 devaluation illustrates the relevance of factors highlighted by so-

called second-generation models of balance-of-payments crises, then the 1949devaluation demonstrates the relevance of the third-generation models developedin response to the Asian financial crisis. The problem in Asia in 1997, as you willrecall, was large amounts of short-term foreign borrowing by banks and firms. InBritain in 1949, the analogous problem was that of the ‘‘sterling balances.’’ Sterlingwas pegged again, and in 1947 made convertible, at US insistence. But World WarII had forced the UK to borrow heavily from its allies, from the Commonwealthand Dominions, and from the Sterling Area formed when other countries followedBritain off the gold standard in 1931. In some cases, exchange controls were usedto prevent their holders from using them to purchase goods and services in Britainor from exchanging them for more attractive (typically, US dollar-denominated)assets. Attempts were made to immobilize the rest using moral suasion.A lesson of the Asian crisis is that external liberalization can have disastrous

consequences in the presence of a financial overhang. That is what happened in1947–1949. Current account convertibility created additional scope for other coun-tries to use their blocked sterling balances to purchase imported merchandise andto employ leads and lags to undertake disguised capital-account transactions. Con-trols were still sufficiently pervasive to prevent a crisis from erupting at once; the1949 crisis was more of a slow-motion train wreck. But a wreck it was. Despite theabsence of obvious signs of internal imbalance, devaluation was needed to reducethe financial overhang.Thus, the 1949 crisis was a classic third-generation crisis in which financial fac-

tors dominated. What is striking from this point of view is that the aftermath wasnot disastrous; growth actually stabilized and picked up. The reason is not hard tosee: because their liabilities were denominated in the domestic currency, Britishbanks and firms were not thrust into bankruptcy. Britain emerged from WorldWar II with an unprecedented sovereign debt of nearly 250 per cent of GDP, yetdevaluation did not force the government to default. Sterling’s status as an inter-national currency, in which both residents and foreigners were prepared to borrowand lend, allowed the UK to avoid these dislocations.At first glance the 1967 devaluation is the most difficult of this trio to reconcile

with modern theories of currency crises. You will recall that this was the time whensterling was pegged to the dollar under the Bretton Woods System. It was also aperiod when the current account of the balance of payments was in balance—indeed, it sometimes was in substantial surplus—between 1964 and 1967. Britainwas not suffering from high unemployment. To the contrary, in 1965–1966 the

1091B. Eichengreen / Journal of International Money and Finance 22 (2003) 1089–1094

Page 4: Three generations of crises, three generations of crisis models

percentage of insured workers recorded as unemployed fell to an historicallyunprecedented 1.5 per cent.It is thus necessary to search for deeper vulnerabilities. One was that even faster

growth abroad rendered investment relatively unattractive. Throughout the period,British savers invested abroad, where returns were higher. It was in this contextthat the current account surplus proved inadequate. In addition, wages rose atdouble-digit rates between mid-1964 and mid-1966, reflecting the scarcity of labor.Given the stability of costs abroad, anemic rates of productivity growth were notsufficient to neutralize the impact on Britain’s competitiveness. Expansive monet-ary and fiscal policies stimulated consumption, pushing up prices. The policystance actually grew more expansionary over time. The budget deficit was notinordinately large, and inflation and money growth were restrained by the desire todefend the currency peg. But the fact that the budget was not in substantial surplusand that money was not tight in this period of high employment suggests that bothinternal and external balance were seriously out of whack. If you are reminded ofArgentina in the second half of the 1990s, then you are not alone.The question is why all this stimulus to aggregate demand, resulting in rising

labor costs and deteriorating competitiveness, did not result in a larger currentaccount deficit. In part the answer lies in policy expedients such as import sur-charges and exchange controls. These started with the 15 per cent import surchargeproposed by the Labour Government within days of taking office.What were the officials thinking? The government that came to power in 1964

was committed to the pursuit of growth and full employment, memories of thehigh unemployment over which Labour had presided in the 1920s never being farfrom its consciousness. The postwar economy had already demonstrated an abilityto function at low levels of unemployment. The question was how long this favor-able situation would last. In fact, insofar as part the explanation for low unem-ployment was the wage moderation bequeathed by memories of the 1930s, then asrecollections of that earlier era began to fade and restraint broke down, it waspossible that the age of full employment was already drawing to a close.Here the Labour Government, demonstrating a considerable capacity for wishful

thinking, convinced itself that by avoiding stop–go policies it could encourage evenfaster growth, in turn enhancing the economy’s competitiveness and relaxing theexternal constraint. Stimulus to aggregate demand would encourage investment.And more investment would enhance the competitiveness of British goods. Britainwould be able to utilize the new technologies embodied in the latest capital equip-ment and move into industries characterized by economies of scale. Thus, whereattempts to stimulate aggregate demand would otherwise run up against the bal-ance-of-payments constraint, in these circumstances it was thought that such initia-tives were sustainable. Eminent economists, such as Nicholas Kaldor, advancedthis dubious view.In reality, of course, the government had made two incompatible commitments,

to faster growth and to not devalue the currency, and it was only through this featof intellectual gymnastics that it could reconcile the two. Similar arguments thatpolicy makers were not causing the economy to overheat or courting balance-of-

B. Eichengreen / Journal of International Money and Finance 22 (2003) 1089–10941092

Page 5: Three generations of crises, three generations of crisis models

payments problems but merely allowing the country to exploit its full growthpotential were of course also heard in Asia in the 1990s. Asian policy makers ulti-mately learned that high levels of investment do not solve all balance-of-paymentsproblems. Not all investments are equally productive, and time is required for evenproductive investments to enhance competitiveness. This is what British officialsalso learned, at considerable political cost, in 1967.Thus, while the 1967 devaluation might at first seem difficult to explain in terms

of conventional models of balance-of-payments crises, the explanation is ultimatelyconsistent with the predictions of those models. Monetary and fiscal policies weretoo expansionary to be consistent with the external constraint. The hope thatdemand stimulus might raise the sustainable rate of growth and relax the externalconstraint proved illusory. Yet the government, when confronted with theseuncomfortable facts, was reluctant to accept more restrictive monetary and fiscalpolicies as the price for defending the currency. Joblessness may have fallen to lowlevels, but policies that contemplated even marginally higher rates of unemploy-ment were unacceptable. When foreign assistance proved limited, devaluationbecame inevitable. This, then, is a classic example of a first-generation balance-of-payments crisis.So I would argue that the theory of balance-of-payments crises has come a long

way in the last 20 years. Today, with three generations of balance-of-payments-crisis models in hand, one can write a better history of British monetary andexchange rate policy in the 20th century was possible two decades ago.But if our understanding of the phenomenon has advanced so significantly, why

haven’t we made more progress in preventing it? I would point to three reasons.First, politics continue to lead governments to run policies that create conflicts

between internal and external balance. This was readily evident in my three Britishcrises. And the importance of the point is hard to ignore in the context of LatinAmerica today.Second, the fact that none of these British devaluations was contractionary,

because balance-sheet effects were absent, points up a major constraint on stabilitytoday—namely, the problem of original sin (Eichengreen and Hausmann, 1999),for which we as yet have no solution.Third, there is the progress of securitization. Today, the typical emerging market

bond issue is held by many thousands of different individual and institutionalinvestors. This is very different from the situation in the era of bank finance thatextended from the late 1960s through the early 1980s. The presence of large num-bers of investors encourages herding that makes crises more likely and creates col-lective action problems that render them more difficult to resolve. The difficulty ofcreditor coordination is what recent initiatives like the Bank of France’s ‘‘Code ofConduct’’ and the IMF’s Sovereign Debt Restructuring Mechanism are designedto address. But these initiatives provide at best incomplete solutions to the problemat hand. That the institutional environment is incomplete in important respects is,at some fundamental level, simply what distinguishes international macro-economics from closed economy macroeconomics. And it is what makes crises in

1093B. Eichengreen / Journal of International Money and Finance 22 (2003) 1089–1094

Page 6: Three generations of crises, three generations of crisis models

financially open, financially dependent emerging markets so difficult to prevent andmanage.My conclusion, then, is that more attention should be paid to the political econ-

omy of policy making in emerging markets. More attention should be directed towhy emerging markets are on such an undesirable point on the risk-return trade-off—why they can only borrow in foreign currency. And more attention should bedevoted to understanding the incentives of market participants in the decentralizedsetting of contemporary international financial markets. That is my modestresearch agenda for international macroeconomics going forward.

References

Cairncross, A., Eichengreen, B., 1983. Sterling in Decline: The Devaluations of 1931, 1949 and 1967.

Blackwell, Oxford, (Second edition Palgrave Macmillan, 2003).

Eichengreen, B., Hausmann, R., 1999. Exchange rates and financial fragility, in Federal Reserve Bank of

Kansas City. New Challenges for Monetary Policy. Federal Reserve Bank of Kansas City, Kansas

City, Missouri, (pp. 329–368).

Eichengreen, B., Jeanne, O., 2000. Currency crisis and unemployment: sterling in 1931. In: Krugman, P.

(Ed.), Currency Crises. University of Chicago Press, Chicago, pp. 7–47.

Krugman, P., 1979. A model of balance of payments crises. Journal of Money, Credit and Banking 11

(3), 311–325.

Ozkan, F.G., Sutherland, A., 1994. A Model of the ERM Crisis, CEPR Discussion Paper no. 879.

B. Eichengreen / Journal of International Money and Finance 22 (2003) 1089–10941094