Official PDF , 84 pages - World bank documents

6 downloads 69867 Views 3MB Size Report
Pleasc contact Raquel Luz, room N 1- ... Pliodud cz at tlie PPR DisseminaLion Center ..... Inertial inflation, or more precise'p ?ersistence, could arise from.
Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized

Policy, Planning, and Research

WORKING

PAPERS

MacroeconomicAdjustment and Growth Country Economics Department The World Bank December 1989 WIPS323

The Old and the New in Heterodox Stabilization Programs Lessonsfrom the 1960sand the 1980s

MiguelA. Kiguel and NissanLiviatan

Heterodox stabilization programs can bring down inflation quickly without costing much unemployment in the short run. But costs that appear up front in orthodox programs are delayed in heterodox programs. Tight fiscal policy and a strong nominal anchor are critical to their success in the long run.

The Policy, Planning, and Research Complex distributes PPR Working Papers to diserninate the findings of work in progress and to encourage the exchange of ideas among Bank staff and all others interested in development issues. These papers carry the names of the authors, reflect only their views, and should be Lsed and cited accordingly. The findings, interpretations, and conclusions amethe authors' own. They should not be attributed to the World Bank, its Board of Directors,its management, or any of its membercountries.

Plc,Planning,and Research

l

jMacroeconomic Adjustment

andGrowth

H-eerodox stabilization programs combine tight monetary anld fiscal policies, characteristic of the orthodox appi-oach, with wage and price controls. Hcicrodox programs arc usually adopted as an altcrnative to orthodox prograins to minimize the costs of bringing down inflation. The hlcterodox approach can bring inflation down rapidly without largc costs in thc short run (the clsx: p.art of th:e program). The costs arc bome latei wlhctn the autlhorities have to concentrate on sustainirln low infationi. Thc magnitude of these cosIS is not yCetclear.

WVhether Lhe stabilization program should adopt a gradual or slhock strategy depends on the rate ol inflation bcfetre stabilization. The ortlhodox shock is most effcctive for stopping h\per;nflation. In economies with chronic, high intlalion, a. hcterodox snock strategy is generally m ore a;i-,; op atc. 1Ih. quick initild success of hietcrodox piograrnis increases support for thc program, opeinin the possibility for adding fiscal mcasLrcs to dccpe_n the stabilization cffort. l iiis opportunity is usually lorgonie, however. I.nracl and Mexico introduced furthler fiscal aidjustr-ment during this phase. In telC Austral and Cruzado plans (in Argentina ard Brazil), thc initial sulcccss was takenI as evidence that morc fiscal adjustment was not required (or could be

postponed), lapsed.

and the stabilization

A gradual approach in Argentina and Brazil in appropriate for economies wAhere the case for income weaker.

effort col-

such as the onc used the 1960s - is more with low inflation, policies is much

Tight fiscal policy is critical to the success of a heterodox program - particularly during the flexibilization period in which price and wage controls are removed. In the failures of the 1960s and the 1980s, the budget situation had alrcady deteriorated when controls were removed. A relaxed fiscal stance during the freeze led to a sharp acceleration of inflation later. It is also important to maintain a strong nominal anchor, before and during flexibilization. In Israel, for example, using the exchange rate as the main nominal anchor during flexibilization heiped prevent a rekindling of inflation (at thc cost of overvaluing the currency). The flexibilization period is the most critical timC in a heterodox program, but policymakers do not perceive this. The strategy for getting out of the freez,e is perhaps the most important and difficult part of a hetcrodox program. Heterodox programs fail not because income policies arc poorly designed but because the fiscal effort does nct persist.

I'ld.s papcr is a product of the Mlacrocconomic Adjustment and Growth Division, Countiy Economlics Departmnct. Copies arc available frce from the World Bank, 1818 fl Strect NW, Waslhington DC 20433. Pleasc contact Raquel Luz, room N 1057, extension 61,588 (79 pages with figures and tables).

1T. ITR 'ol.lirg Paper Series disseminates the findings of vork under way in the Bank's Policy, Planning, and Rcsearch ( .^ n.l. k An (hiective of Lhe scrics is to get these findings out quickly, even if presentations are less than fully polished. Thc filulaugs, icrtprcttion.s. and concltusionis in these papers do not necessarily reprcsent official policy of the Bank. Pliodudcz at tlie PPR DisseminaLion Center

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II. Difficulties with the Orthodox Approach . . . . . . . . A. Empirical Evidence .... . . . . . . . . . . . . . B. Inertia, Inflation rigidities and Income Policies a. Reasons for Inertia . . . . . . . . . . . . b. Implications for stabilization programs . .

. . . . .

. . . . .

. . . . .

. . . .. . . . . . . . . .

.

1

. . .

5 6 7 7 12

. . . . . . . . . . . . . .

16 17 18 20 20 24 26 26 28

. . . . . . .

30 31 31 34 36 39 40

V. Long Term Aspects of the programs . . . . . . . . . . . . . . . . . . A. The costs of disinflation in the successful heterodox programs . . .43 B. The costs of failure ....... . . . . C. Sustaining price stability in the long run . . . . . . . . . .

43

VI.

. . . . . . . . . . . . . . . . . .

50

......

54

III. Design and Implementation of Heterodox Stabilization Programs . A. Gradualism Versus Shock ... . . . . . . . . . . . . . . . . B. Stages in Heterodox Stabilization Programs . . . . . . . . . C. Prices, Wages and Exchange Rate Policies . . . . . . . . . . a. Income Policies in the First Stage . . . . . . . . . . b. Flexibilization and Decontrolling in the Second Stage D. Fiscal and Monetary Policies ... . . . . . ...... . . . a. Fiscal Policy ... . . . . . ..... . . . . . . . . b. Monetary Policy ... . . . . ..... . . . . . . . . IV. Outcomes of the Programs ... . . . . . . . . ....... A. Inflation .... . . . . . . . . . . ....... a. The Eighties . . . . . . . . . . . . . . . . b. The Sixties . . . . . . . . . . . . . . . . B. Output and Unemployment .... . . . . . . ..... C. The Trade Balance, Consumption, and Investment . . D. Real Exchange Rate and Real Wages . . . . . . . .

Final Remarks and Policy Issues

VI. References

....

. . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . .

. . . . . . . . . . .

. . . .

.

This paper is a major revision of "The Old and the New in Heterodox Stabilization Programs: A Study of the Sixties with the Perspective of the Eighties" and is part of Research Project (RPO 674-24) Stopp:.ngHigh Inflation. We are grateful to Bela Balassa, Adrienne Cheasty, Luis Giorgio, Leonardo Leiderman, Steve Webb and participants at the workshops at CEMA, Tel Aviv University and The World Bank for useful comments on an earlier draft. We thank Jariya Charoenwattana for her patient and effective research assistance.

44 47

I. Introduction

The heterodox stabilizationprogramsof the eightieshave revived the old controversyabout the use of income policies for stopping inflation. Income policies are a central component of heterodox stabilizationprograms which combine the adjustmentin fundamentalspredicatedby the orthodox approachwith price and wage controls. Tbe inclusionof controlshas been motivated by the difficultiesencounteredin orthodox programs (based primarily on tight monetary and fiscal policies) for stopping inflationin chronic inflationcountries,where inflationtends to display downward rigidity. The objectiveof this paper is to gain new insightsabout the strengths and weaknesses of the heterodox approach for stopping inflationbased on the experienceof programs implementedin some Latin American countries and Israel in the sixties and eighties. The most important income policies supported stabilizationprograms in recent years were those launchedin Argentina in June 1985 (the Austral plan), in Israel in July 1985, in Brazil in February 1986 (the Cruzado plan) and most recentlyby Mexico (the Pact for Economic Solidarity)in 1987-88.1 Income policies are not a new feature in the big living laboratoryof stabilizationprograms in Latin America. Indeed, they played a major role in three important stabilizationprogramsof the sixties; those of Brazil in 1964, Argentina in 1967 and Uruguay in 1968. These programs shared some basic common featureswith those of the eighties. Specifically,they were based on wage and price controlsand made use of the

1 It is somewhatmisleadingto include the Cruzado Plan in a study of heterodox programsbecause the fiscal imbalanceswere not addressed in that program. Wa neverthelessdecided to include it in our sample becausemost of the recent literatureon heterodox programshas included it as well.

2 exchange rate as the main nominal anchor. These policieswere combinedwith reductionsin the budget deficits. There is an extensive literaturecoveringeach of the programs of the two eras, while there are a number of works comparingthe programswithin each period.2

This paper will provide a broader perspectiveon heterodox

programsby analyzingwithin one frameworkthe programs of the two eras and by incorporatingthe more recent developmentsin the successesand failures of the eighties. Income policieswere an effectivetool for bringing down inflation in the short run in both eras. As can be seen in figure 1-a and 1-b, the initial resultswere very impressivein the four programs of the eighties and in Uruguay in the sixtieswhere inflationcame down very quickly to much reduced levels. The reductionswere slower, but still significant,in the initialphase of che Brazilianand Argentineprograms of the sixties. The initial success proved difficultto sustain in many programs. In the success stories,Brazil in the sixties and Israel and Mexico so far in the eighties, there was persistenceon the fiscal side and an effort to maintain a strong nominal anchor throughoutthe stabilizationprogram. Lack of persistenceon the fiscal side was perhaps the most important reason for

2 The Brazilianprogram of the sixties is discussed in Fishlow (1973),Cardoso and Fishlow (1988),Foxley (1980) and (1983),Kafka (1968) and Simonsen (1974. among others; de Pablo (1972) and (1974),and Krieger Vasena (1987) analyzed the Argentine 1967 program;and Finch (1979) and Viana (1988) discuss in some detail the 1968 program of Uruguay. Pazos (1969) undertakesa comparativestudy covering the Brazilianprogram and the :nitial stages of the Argentine and Uruguayanprograms. For the programs of the eighties see for example Cardoso and Dornbusch (1987), Heyman (1987) and Fischer (1987),El TrimestreEconomico (SpecialIssue 1987) EstudiosEconomicos (SpecialIssue 1987), Blejer and Cheasty (1988), Bruno et. al (1988),Liviatan (1988),Kiguel (1989) and Solimano (1989).

3 the eventual failuresof the Uruguayan,and the two Argentine programs and of the Cruzado plan. Income policieshave a clear and specifictarget in stabilization programss control of what is commonly, and many times loosely termed the inertialcomponentof inflation. This component,resultingfor instance from indexationor lack of credibility,is usually used to describe that part of inflationthat can not be explainedby fundamentalfactors --namely seigniorage. The use of controlscan be justifiedon a temporarybasis only to deal with the inertialaspects of the inflationprocess. If controlsare maintainedbeyond this stage the resultingdistortionsmight well outweigh their usefulnessfor stabilizationpurposes. The use of income policies is urderstandablyviewed with skepticism. One importantproblem is that price controlscan be and have been used many 3 times as a substitutefor the necessary correctionin the fundamentals. In

addition,experienceshows that once controlsare introducedthey are very difficult to remove. This is a criticalproblem because controls are not the appropriatepolicy to maintain low inflationin the medium and long run. If controls are maintainedbeyond the short term they will have damagingeffects on resourceallocationand induce the emergenceof black markets. The overall economicconditionswere very different in the two eras. Macroeconomicimbalancesprior to the programs,particularlythe size of the budget deficit,were much smaller in the sixties than in the eighties,while

3 This is speciallythe case in the populistprogramswhere the fiscal aspects of inflationare entirelyignored. These type of programs have been popular, in Latin America, as evidencedby Chilean and Argentine program of the early seventies,the Peruvianprogram of 1985-86 and possibly the Cruzado Plan. For a discussionof this topic see Dornbusch and Edwards (1989).

4 the overall performanceof the economy as measuredby growth, investment, etc. was also better in the early period.

The behaviorof inflationwas

also very different in the period preceding the programs.

Inflationwas

much higher ir.cne eightiesthan in the sixties for the countries included in this study. In addition, it had by then become a chronic problem in the eightiesafter many years of very high .nflationand many failed stabilizationattempts. This divergencein the overall economicconditionshad important implicationsfor the strategyand the resultsof the stabilizationprograms. The higher rates of inflation in the eightiescalled for a strategybased on a shock treatmentincludinga sharp reductionin the budget deficit and a freeze of the exchange rate, prices and wages. A more gradual approachwas possible in the sixtiesbecause the level of inflationwas lower thus providing enough time to phase in the various policy measures. The ex:ternal conditionswere also less favorablein the eighties. In contrast to the sixties,the eightieswere characterizedby low growth in the world economy,high real interestrates, and deterioratingterms of trade for the Latin American countries. More importantly,the Latin Americar countries in the eightieswere confrontedwith a sudden halt in their access to internationalcredit since the beginningof the debt crisis in August 1982. These conditionsalso contributedto an easier adjustmentin the sixties as the countries could use external financingto increasepublic and private investmentand sustain growth. The remainderof the paper will be organizedas follows. In section II we illustratewith some examples the difficultiesencounteredwith the orthodoxapproach. We argue that downward rigidityin inflation,sometimes

5

referredto as inflation inertia,observed in the chronic inflationcountries is the major obstacle for the success of orthodoxprograms. We examine some of the common explanationsfor inertia includingbackward indexationand expectational(forwardlooking)reasons r

iltingfrom lack of credibility

and coordinationproblems. We concludethis sectionwith an evaluationof whether price controlsare indeed the right responsein each of these cases. In section III sielook at issues in the design of heterodoz programs. We first argue that the use of the shock approach in the eighties, as opposed to gradualismin the sixties,was motivatedby the higher inflation rates -Ali'igin the more recent era. We also discuss the interrelationbetween fiscal policy and the strategyfor managing the various nominal anchors and the way in which policies -

to be adjustedas the program evolves. We

argue that the strategy fL,.

ting out of the freeze is perhaps the most

importantand difficultpart of a

- odox program. The outcomes of the

programsare discussed in section V. Here

. analyze the strengthsand

weaknessesof each program, and the mLaJr fL-cors behind their success or failure. An importantmessage is that heterodox programscan bring down inflationvery quickly in the short run without much costs in terms of unemployment,the most difficult and costly part usually appears at a later stage when trying to sustain low inflation. In section V we deal with some long term issues such as the costs of failure and of sustainingprice stability. We conclude in section VI with some final reflectionson heterodox programs.

II. Difficultieswith the OrthodoxApproach

6 A. Empirical Evidence The difficultiesencounteredwith the implementationof the orthodox approach in chronic inflationcountriesprovidedthe main motivationfor the inclusionof income policies in stabilizationprograms. The existence of downward rigiditiesin inflation,sometimesreferredto as inflationinertia, limits the effectivenessof orthodoxpolicies to achieve rapid disinflation. These difficultieswere evident during the eighties in Brazil and Mexico (see figure 2).4 These two countries implementedorthodoxprograms during 198S-84 in conjunctionwith IMF programsaimed at the double objective of reducing inflationand improvingthe balance of payments. The adjustment was baged on a significantreductionin the budget deficit (exceeding6 percent of GDP in the operationalbalance in Brazil and 8 percent in Mexico), a tighteningin domesticcredit, and large devaluations. While the programs were effective in restoringexternalbalance, they resulted,paradoxically, in an increase in the rate of inflation. In Mexico, there was ar.important accelerationin inflationin 1987, a year in which the governmentwas having 5 surplusesin its primary and operationalbalances.

Similar frustrationwas

observed in Argentina,a country that abandonedvery quickly an IMF stand by agreement signed in late 1984, once it became apparent that a temporary tighteningin fiscal and credit policieswas having negative repercussionson economic activitybut little impact on inflation.

4 For a short discussionof these programssee Kiguel and Liviatan (1988). 5 The primariy deficit is calculatedby substractingall interest payments from the total deficit. The operationaldeficit is calculatedby substratingthe inflationarycomponentof interestpaymentson domestic debt from the total deficit.

7 A similar

pattern

is observed in the sixtleswhere the Brazilianand

Uruguayan prograumsstarted along orthodox lines. The Brazilian stabilization program started in March 1964, under CastelloBranco. It was based on a significantreductionof the overall budget deficit (from 4.8 of GDP in 1962 to 1.6 in 1965), tight domesticcredit, large realignmentsin relativeprices (particularlyof public sector and other prices subjectedto controls)and frequentdevaluationsof the exchange rate. Income policieswere limited to the control of nominal wages, an element found in other orthodoxprograms of the time. The program achieved rapid success in the external front, (see table 1) ith.le the resultswere more modest on inflation. Inflationremainedhigh during 1964, and the first half of 1965, as can be noticed from figure 1-b, a strikingfact given the cut in the budget deficit. Uruguay initiateda stabilizationeffort in late 1967 supportedby an IMF program. The adjustmentwas initiatedwith a 100 percent devaluationof the peso in November 1967 and was supportedby a correctionin the fiscal imbalance in late 1967 (from 3 percent of GDP in 1967 to 1.7 in 1968) and a tighteningin domesticcredit (see Table 2). Despite the adjustmentin the basic policy variables,inflation remainedat very high levels (at an annual rate of 165 percent) during the first two quartersof the year (see figure 1-b). Additionaldevaluationsin early 1968, added to the inflationaryspiral. On the other hand, there was an improvementin the current account and balance of payments. Overall the outcome was like that of other programs. The orthodoxapproachwas effective to correct externalpoliciesbut had an adverse impact on inflation. B. Inertia,Inflationrigiditiesand Income Policies

8 a. Reasons for Inertia The similaritiesin the results of the various programs discussedabove points to common elements ir.the inflationprocess in these countries. In fact, the orthodox approach fails to deal effectivelywith the inertial aspects of inflationpresent in chronic inflationcountries,and faces major difficultiesin those cases in which inflationdisplays strong downward rigidity. These difficultiesoffer a motivationfor the inclusionof inccie policies in stabilizationprograms. Indeed, the use of wage and price controls in heterodox programs, sometimesin the form of a freeze, is primarily justifiedby ohe existenceof rigiditiesin the inflacionprocess (e.g. Dornbuschand Simonsen (1987),Heyman (1987),and Meller (1987) among others). The term inflationinertia is difficultto define precisely,partly because the existingliteratureuses it in a loose way. This term is usually used to describea situationin which there is persistencein inflation,and where the prevailingrate of inflationremains above the level that can be explainedby fundamentals(namelyseigniorage). Inflation inertia is primarily used in connectionwith countriesthat have experiencedprolonged inflationarypro^esses,the so called chronic inflationcountries,where inflationto a large extent appears to have a life of its own, in many respects independentof the evolutionin the fundamentals. Inertialinflation,or more precise'p?ersistence,could arise from backward looking elements,such as wag: *ndexation,or from what essentially are forward looking factors,such as expectationsabout future policies. Much of the existing literatureon persistenceemphasizesthe importanceof staggered long term contracts (as in Fischer (1977) and Taylor (1979)),and

9 backward indexation(Fischer(1983) and Modiano (1988) among others), althougha number of recent works giva more weight to expectationsand credibilityissues (e.g. Perssosn and van Wijnbergen (1989),Dorr.busch ard Simonsen (1987)). It is useful to illustratethe type of persistencethat emerges from some of the existingmodels. Fischer (1977) end Tayler (1979) both consider the case in which labor contractsare staggered,and where the nominal wage is set for two periods. In Fischerismodel, wage contractsare entirely forward looking and primarily depend on expectationsabout monetary growth. In this set-up there is no persistencebeyond the expirationof the most recently signed contract;because once a new policy is announcednew contractswill be drawn on the basis of the announcedmonetary targets. More persistenceis generated by a Taylor type model, where contractshave forward and backward looking elements. The 'atter arise because one of the factors that enters into the determinationof the marginalwage is the prevailing averagewage, which includesthe wage set in the previous period. There is no backward indexationin this model. The forward looking elements are similar to those in Fischer (1977). In Taylor'smodel the price level can be expressedas a distributedlag of past prices, hence implyingthat past rates of inflationfeed back on current inflationeven after all the contractsthat were negotiatedprior to the new policy announcementexpire. While persistencein these models is mainly driven by backward looking factors, this result rests on the assumptionthat policy announcementsare credible. This, however, is not usually the case as lack of credibility could arise regardingthe ability of the governmentto reduce the budget deficit in a sustainablemannei:,or about its determinationto maintain the

10

nominal anchors,e.g. the exchange rate or the money supply. If inflationary expectationsare a weighted average of say two alternativeoutcomes (i.e. one of high inflationand one of low inflation),it will lead to persistenceas 6 these expectationsare incorporatedin current contracts.

The policy games literature,as developed in the works of Barro (1983), Barro and Gordon (1983.a)and (1983.b),Cukiermanand Liviatan (1989) and Persson and van Wijnbergen (1989) among others, providesnew insights regardingrigiditiesin inflation. If a governmentcannot abide by rules and uses surpriseinflation in a discretionaryway to achieve short term goals, then inflationcan remain persistentlyabove the minimum level determinedby the fundamentals. A fiscal adjustmentin a discretionaryregime is likely to fail to achieve large reductionsin inflation. Paradoxically,as shown in Kiguel and Liviatan (1989.a),it could lead to an increasein inflation. Problemsof coordinationamong private firms, examined in more detail in Dornbuschand Simonsen (1987),providesanother explanationfor persistence. In their frameworklack of credibilitydoes not result from lack of confidencein the government'sability to persistwith sound policies. The problem ar'ses due to the i.abilityof price setters to coordinatea movement to a lower inflationstep. A firm would be willing to make smaller increasesin its price only if other firms follow a similar behavior. Lack of coordinationcould sustain inflationat ongoing levels even if there are changes in the fundamentals. Since each firm sets its price independently,the economy lacks a mechanism to synchronizethe "Jump"

6 This result is consistentwith Calvo (1983)where expectationsof devaluationare taken into account in the determinationof prices in new contracts.

11

to the low inflation step. In Dornbuschand Simonsen'swords 'Everybody would like to jump to a low inflationrate together,but nobody will jump unless the others do also." (p.5) Inflationwill thus continue at the old rates even if a fiscal adjustmentis implemented. Wage indexationis probablythe most frequentlymentioned reason for the existenceof inertia. The effect of indexationon persistenceis, however,not always clear. On the one hand, as is usually argued, backward indexationcreates persistencein the economy as past rates of inflation feed back on currentwages and through costs on current inflation. Thus whatever rate of inflationprevailed in previous periods tends to remain in the system and hence increasespersistence (as in Fischer (1983) and Modiano (1988)). On the other hand, wage indexationcan facilitatea reduction in inflation since once the disinflationprocess starts, it subsequentlyleads to smaller wage increasesand hence reducingpersistence(as argved in Friedman (1974). The former argumentis probablymore importantin models where prices are determinedby a mark-up equationwhile the latter is more relevant for models that use market.clearingprices (e.g. based on the quantity theory). The above discussionillustratesthat there is a component in the inflationprocess that is independentof the evolutionof the fundamentals, at least in the medium run. While seignioragein general will determine a lower bound to the rate of inflation,there is also a non-fiscalcomponent which can move inflationabove this lower bound. The non-fiscalpart is probablyvery importantin the chronic inflationcountrieswhere the available evidencedoes not indicatethat a close relationshipexists between

12 7 seigniorageand inflation.

Income policiesare potentiallymore useful in

those cases in which the non-fiscalcomponentis at the heart of the inflationprocess. b. Implicationsfor stabilizationprograms An implicationof the precedingdiscussionis that when there is persistence,regardlessof whether it results from staggeredcontracts, backward indexation,lack of credibilityor coordinationproblems, the orthodoxapproach does not lead to significantreductionsin inflationin the short run. The heterodox approach is intendedto deal specificallywith this problem,with income policies complementingthe adjustmentin fundamentalsto achieve rapid disinflation. It turns out that the effectivenessof heterodox programs largely depends on the prevailingrate of inflation. As a general rule one could argue that in chronic inflationcountriesa heterodox shock, of the type used in the eighties,is more appropriatefor countriesexperiencinghigh inflation,while a gradual approachwith less relianceon income policiesand where controlsare used in a flexiblemanner, is better suited for low inflationcountries. High inflationcountries ace those where as a result of increasesin the rate of inflationthere is significantsynchronizationof prices and wages. Staggeringis less importantin these countries,and hence persistenceis mainly explainedby credibilityand coordinationissues. i. The Shock Approach A heterodox shock, includinga freeze of prices, wages and the exchange rate can be effectiveto break persistencein this case.

Suppose that there

7 This point is discussed in more detail in Kiguel and Liviatan (1988).

13 is persistencemainly because the private sector dees not believe on the sustainabilityof the fiscal adjustmentor on the willingnessof the monetary authority to maintain the fixed exchangerate, while the governmentis determinedto implementand sustain a serious stabilizationprogram. In this setup expected inflationwill in general exceed the "officialforecast". This is likely to result in rigiditiesin actual inflationas current prices (or wages) are determinedbased on forward looking expectations. A price freeze could be effective in this situationas a mechanism to buy time while the authoritiestake the appropriatefiscal measures and demonstratecomprehensiveness and persistencein the stabilizationeffort. If controlsare not used and inflationstays at rates close to those prevailingprior to stabilization,credibilityproblems could exacerbateas the fiscal adjustmentproves ineffectiveto bring down inflation. Controls could be removed once major doubts about the seriousnessof the program disappear and agents have reasonableconfidencein the success of the program and in the sustainabilityof the situation. Likewise,price controls (in the form of a freeze) can play an important role when there is a problem regardingcoordinationof price decisionsof the type discussedin Dornbuschand Simonsen. A comprehensive price freeze provides the individualfirm with informationthat the market mechanism does not furnish. It tells each firm that all others will set their prices under the same rule thus removingan importantobstacle in moving the economy to a low inflationstep. This argument,however, implies that once the coordinatedmovement to low inflationis completed,usually a short process, the authoritiesshould start to remove controls. Although it seems plausible that a shock treatment is appropriatewhen

14 prices are highly synchronized,why does it need to include income policies? Why can't an orthodox shock of the type used to stop hyperinflationsbe effective in chronic inflationcountries? The answer lies primarily on sustainabilityissues, and on the perceived relationshipbetween deficitsand inflation in the two cases. A hyperinflationprocess is inherentlyunstable and in this respect unsustainable. It is almost unthinkablethat a country can function for prolongedperiods under hyperinflation. Thus a serious orthodox program is likely to be credible in this case. Chronic inflation, on the other hand, can be maintained. In most cases the public and private sectors develop mechanisms (such as indexationof wages, taxes, etc.) that allows them to live with inflation. Stabilizationis thus postponableand hence less credible. The second differenceis related to the nature of inflationprocesses in the two cases. In hyperinflationsthe causes of inflationare clear, excessive seigniorageto finance a large budget deficit. In chronic inflationcountries,on the other hand, fiscal and non-fiscalfactors are equally important to explain inflationin the short run. As a result,while the cure for inflation is clear in hyperinflation,it is less so in chronic inflationcountries. An orthodox shock is effective to stop hyperinflations, but a heterodox shock is more suitableto deal with high, chronic inflation. Is the shock approachalso well suited to deal with backward indexation in high inflationeconomies? The answer is positivebecause contractsare highly synchronizedin these economies,althoughin this case the shock approachdoes not necessarilycall for a prolonged freeze. If the authoritieshave enough informationon the indexationarrangementand about the equilibriumreal wage, then they can determinethe correct initialwage

15 increasewhich in conjunctionwith the eliminationof indexationfor upcoming contracts should be sufficientfor solvingpersistencearising from this source. The impositionof a freeze on wages after the initial adjustmentis not in principlenecessary to deal with backward looking elementsafter the initial adjustmentis done. Income policies,however, c8a still be useful in practice since workers are not likely to give up indexationin the absence of an agreement in which firms accede not to increaseprices. A social pact of this type, car serve as a coordinationmechanism to facilitatethe eliminationof indexation. ii. Gradualism Low inflationeconomiesare those where staggeredlong term contracts are still prevalent. Low inflationis used in a broad sense here, as it will includecountrieswith annual rates of inflationup to 40 percent where wage adjustmentstake place on an annual basis and even countrieswhirh would be consideredhigh inflation for other definitions,with inflation rates in the range of 100 percent per year, but where wages are usually adjustedsemiannually. The importantfeature in our definitionis that due to the prolonged durationof contractsand the asynchronizationin price and wage decisions (resultingfrom staggering),there is significantpersistencein the economies. In other words, the system has a long memory. The case for income policies is weaker in these cases. If the authoritieswant to stop inflationthrough income policies they would need very detailed informationregardingthe time at which each contractwas signed, its duration, the time of the last wage adjustment,the relative weight of forward and backward looking elements in each agreement,etc. They would then use this informationto design a complex set of rules providing

16 differentwage increasesfor each contract. In practice this is likely to be very difficult thus weakening the case for a shock approach relying on income policies in low inflationcountries. While in low inflationeconomies long term staggeredcontractsare likely to be an importantsource of persistence,the scope for dealing with these forces is limited. This is also due to the fact that the non-fiscal factors are usually less importantto explain the on-going rate of inflation. Thus, tha orthodoxcompcnent is the central part of a stabilizationprogram in these economies,although some use of income policies in the form of guidelinesor voluntary agreementscould be includedto deal with expectations. iii. A Caveat Although income policiescan be useful in certain cases to support a stabilizationeffort, they are not the appropriatetool to maintain low inflationon a long term basis. The removal of controls is an importantpart of a heterodox program,but one that is not usuallywell planned. A central objective during the flexibilizationperiod is to avoid an increasein inflationthat would bring it back to the levels prevailingprior to the program. If lack of credibilityabout the sustainabilityof the program or problems of coordinationare still prevalent (as they usually are), the removal of controlsshould proceed gradually. As controls are being removed in the various sectorswithout producingan increasein inflation,this would send a signal to agents that inflationis not going to return to previous levels.

III. Design and Implementationof Heterodox StabilizationPrograms

17

The heterodox stabilizationprogramshave as their main common element the use of price and wage controls. Nonetheless,there are importantdifferencesacross programs regardingthe comprehensiveness, enforcementand durationof controls,and on the supportingpolicies that were utilizedto enhance the effectivenessof the programs. In this sectionwe will look at the way in which the income policiesbased stabilizationprogramswere implementedin the two eras, and highlight the main differencesand similaritiesin the use of policy instruments. A. GradualismVersus Shock We argued in the previous section that the strategyshould depend on the underlyingrates of inflation. The lower rates of inflationprevailing in the sixtiesmade possiblea more gradual approachfor stopping inflation. This was not possiblein the eighties,where the high rates of inflationand the chronic nature of the inflationprocess called for a shock treatment. In Brazil and Argentinaduring the sixties,for example, price and wage controlswere phased in slowly, on a voluntary basis and in the form of ceilingson price increases. In most of the programs of the eighties,on the other hand, a shock treatmentwas followedwith controls encompassingmost sectors of the economy, they were compulsoryand in the form of a freeze. The main exceptionsin each era were Uruguay, a country that facing high inflationadopted a shock treatmentin the sixties,and Mexico, who having the lowest inflation in the eightiesopted for a more gradual approach. The heterodox shocks, used in Argentina,Brazil, and Israel in the eightiesand in Uruguay in the sixties,usually includeda freeze of

18 prices,wages and the exchange rate. In these episodes,particularlyin the eighties, the heterodox shocks were implementedwhen inflationwas at very high levels and at a time when there was a widespreadperceptionthat the non-fiscalcomponentof the inflatioiuary process was relatively important. In our view, the main justificationfor the *iseof controlswas to deal with issues of credibilityand coordination. Thir however,was not necessarilythe reason providedby those who designed and conducted the programs,where backward indexationof wages was mostly blamed for the persistenceof inflation. While backward indexationwas widespread in these countries,the persistencethat could result from this factor was less importantbecause contractsbecame shorter and more synchronizedas a result of increasesin inflation. Since non-fiscalfactorswere an importantunderlyingelement in the inflationprocess a heterodox shock could pulled down inflationvery quickly. This approachwas speciallyappealingbecause these programshad one main objective:stopping inflation. A simultaneousmove on prices, wages and the exchange rate was needed to avert dispersionin relative prices and large changes in income distributionthat could destabilizethe programs. If the freeze had been restrictedto a limitednumber of markets, say only to wages and the exchangerate, prices would have continued rising due to inertialforces deterioratingreal wages and leading to an overvaluationof the currency. The adoptior. of a more gradual approach in Brazil and Argentina during the sixties and in Mexico more recentlywas possible because the rate of inflationwas much lower. The lower rate of inflationallowed more flexibilityin the use of controls in the sense that delays in controlling

19

or decontrollingspecificprices were not likely to create large distortionsor to threaten the viabilityof the program. In addition,the likelihoodof achievinga significantreductionin inflationthrough a heterodox shock when inflationwas already low, was much smaller. B. Stages in Heterodox StabilizationPrograms The analysisof a heterodoxprogram can be simplifiedif we divide its evolution in three stages. We can distinguisha first stage in which the authoritiesconcentrateall their efforts ir.stopping inflation. The priorityhere is to achieve drastic and rapid reductionsin inflation rates. This stage is easy to recognizein the eighties (for example from figure 1-a) and took the form of a shock treatmentincludinga large fiscal adjustmentand a freeze of prices,wages and the exchange rate. The exact timing of the first stage is more difficultto identifyin the sixties, especiallyin Brazil and Argentinabecause controlswere introduced graduallyand in a more flexibleway. In Uruguay, on the other hand, this stage is clearly identifiableand coincideswith the impositionof a freeze similar to that of the eighties. This stage is usually the easiest in a heterodox program. The second stage, by far more difficult,comprisesthe peried during which the stabilizationprocess needs to be consolidatedand prices are flexibilized. Few countrieshave been able to completethis stage successfully. It is in this stage when the problemsand difficultiesthat were not apparentduring the first stage in the less comprehensiveand less persistentprograms start to emerge. The typical problems are an overvaluationof the domesticcurrencyand increasingimbalancesin the external or budgetary accounts.

20 The key aspect of the second stage is the design of a mechanism for decontrollingthe economy and the choice of a strategy for monetary, fiscal and exchange rate policies to supportprice stabilityon a longer term basis. Especiallyimportantis the choice of an exchangerate rule which balances the need for greater flexibilityin managing the nominal exchange rate with the maintenanceof a strong nominal anchor. The second stage is usually a difficultperiod in the programsbecause little is known about the mechanismsand timing for de-controllingprices, the choice of nominal anchor and about the selectionof appropriatefiscal, monetary and exchange rate policies to support the process. The completionof these two stages can be regardedas the end of the heterodox phase of the program (for both successesand failures). There are issues,however, that arise thereafterwhich are still related to the originalprogram. We can thus include a third stage in our analysis,which refers to the evolutionof macroeconomicvariables and policy issues beyond the heterodox phase. This includesissues such as what are the costs of failure,problems of overvaluationobserved in successesand failuresand what are the options regardingthe choice of nominal anchor after the initial heterodox phase. C. Prices, Wages and Exchange Rate Policies The policies for prices, wages and exchangerates varied across programs and were modified as the programsmoved through the different stages of the stabilizationprocess. Controlswere usually more strict during the first stage when the main objectivewas to bring down inflation. Policies for prices, wages and the exchange rate were generallysimilar across programs in this stage. There is more variation regardingthe way

21

in which the various programs removed the controls and the manner in which they confrontedthe difficultiesthat arose at later stages. a. Income Policies in the First Stage i. Price Controls As discussed in the previous sectiona shock approachwas adopted in those countries facing high inflation,while gradualismwas followedin the low inflationcountries.There is little variation in the way in which controlswere applied in the various shock programs, controlswere mandatory and quite comprehensive. None of the programsmade a large effort to enforce controls, in the shock programsthe mere announcementof controlswas enough to achieve rapid disinflation. The use of controlswas even more flexiblein the gradual programs. In the Brazilianprogram of the sixties price controlswere first introducedin 1965, one year after the initialprogram was launched. Price controlswere mandatory for agriculturalfood products,public servicesand other basic goods; while they were voluntary for most industrialfirms in 8 the private sector.

Similarly,price cantrolswere phased in gradually in

Argentina during 1967, also on a voluntarybasis.9

In the Mexican program

8 Nonetheless,the governmentattemptedto influencecompliance through the introductionof tax and credit incentivesfor those firms that restrainttheir increasesin prices. Firms that would agree to restraint their price increasesto 7Z until the end of the year would receive a reductionin income taxes, an expansionin their access to (cheap) lines of credit from Banco do Brazil and an increasein their quota for imports of raw materials. For a more detailed discussionof the structureof controls see Foxley (1980, p. 898). 9 Influencedby the Brazilianstrategy,a mechanismwas introducedin mid 1967 by which those firms that would agree to fix their prices for a pre-determinedperiod receivedfiscal benefits and access to special lines of credits in exchange. The governmentwas successfulwith this strategy as the majority of the leading industrialcompaniesjoined the agreement.

22 there was also flexibilityin this respectas prices for most industrial productswere not directlyunder governmentcontrol. The impositionof controlsdid not result in generalizedshortages, at least in the first stage. There were shortagesin a limitednumber of products,but they tended to be of a temporarynature and in many cases in goods that had seasonalityin either productionnr consumption. Shortages and black markets only became a serious problem at a later stage in those programs,such as Uruguay and the Cruzado,which attempted to maintain the freeze in spite of economywde

excess demand pressures. The absence of

economy-wideshortagesindicatesthat controlswere used judiciouslyin mo' programs. Their role was to serve as guidelinesand in helping to coordinatethe move to low inflation. This stands in oppositionwith the populistprogramswhere controlsare used to repress inflation. ii. Wage policy Wage policy was similar in most programs and usually consistedof a mandatory freeze of nominal wages at the existing levels. In most cases the durationof the freeze was not made explicit. Two areas that deserve special attentionare the initialadjustmentin relativewages, and the approach to wage indexation. The Krieger Vasena program and the Cruzado plan were particularly sophisticatedin realigningrelativewages at the outset. In Argentina, prior to 1967, wage contractswere staggeredand had a duration of one year. The program includedan initial adjustmentin wages, the size of the adjustmentdifferedacross sectorsand dependedon the expirationof

23 10 ongoingwage contracts.

This strategywas followedin order to limit

the size of disparitiesin real wages that would arise as a result of the freeze. The Cruzado program also includeda mechanism to minimize the impact of the wage freeze on relativewages. New wages were determinedso as to restore the purchasingpower of the previous six monthc. ir;some cases includinga reductionin nominalwages. It is difficultto assess whether the initial realignmentin relative wages had a significantrole in later developments. It was probablya strength in the Argentine program of the sixties. In the Cruzado plan it does not appear to have been useful, but mainly because the resulting real wage was much too high to maintain externalbalance. On the other hand, the three programs that achievedgreater success (Brazilin the sixties, and Israel and Mexico in the eighties)did noc deal directlywith this issue in the design and implementationof the policy packages. The approach to wage indexationis a second differenceacross programs. The programs of the eightiesattempted to limit the use of indexationand frequencyof indexation,although this in practicewas very difficultto do. Wage indexationwas prevalent in Israel and Brazil during the eightiesprior to the stabilizationprograms. Indexationwas not eradicatedin either case as wage adjustmentsbecame contingenton inflationexceeding some trigger point. In the Cruzado plan the trigger point was 20 percent while in Israel, indexationwas initiallysuspended

10 The larger increases (24 percent)were for contracts that were about to expire (i.e. the older contracts)with smaller increases (up to 8 percent) for those contracts that were signed just before the plan. Public sector wages were increasedby 15 percent,which roughly correspondsto the average raise in the private sector.

24 for three months with a triggerpoint of just 4 percent, later increased to 7 percent. A differentstory is observed in the Brazilianprograms of the sixti&j,where indexationwas introducedby the program. The formula includedbackward and forward looking elements,with the latter consistentlyunderestimatedby the authoritiesto maintain control on real wages. Nevertheless,given today's perceptionabonitthe rigidities introducedby backward indexation,it looks surprisingthat the authorities introducedrather than eliminatedindexationin the midst of a 11 stabilizationprogram.

iii. Exchange rate policy The exchange rate appears to be the preferrednominal anchor in heterodox stabilizationprograms. In all programsthe exchange rate was frozen at the outset. In some of them, as in KriegerVasena and to some extent in Mexico, this was preceded by a maxi-devaluation which was intendedto provide a cushion to avoid (or at least postpone) an overvaluationof the domesticcurrency. In all cases, the initial value of the exchangerate was changed at a later stage. The main differences (whichwe will examine shortly)were on the timing of the devaluationand on the rule adopted for adjustingthe exchange rate after the first devaluation. b. Flexibilizationand Decontrollingin the Second Stage

ll Indexation,however,was not seen during those days as an obstacle for reducinginflation,just the opposite,people like Milton Friedmanwere advocatingthe use of indexationto minimize the costs of reducing inflation.In Brazil, however, it was apparentlyadopted in order to provide a mechanism to allow the economy to adapt to inflation.

25 Only two of the programs of the eightieshave extendedtheir success beyond the initial stage, Israel and Mexico. The experiencesof the sixties are more difficultto evaluatebecause the beginningof the second stage

is harder to identifywhile the policiesadopted in some cases

implied greater relianceon controls. The flexibilizationperiod in Israel started approximatelysix months after the beginningof the plan. Price controlswere removed on a small number of goods startingin January 1986, and the number of products under control slowly decreasedfrom 80 percent at the beginning of the program to approximately25 percent in January 1987.12 The exchange rate remained fixed against the US dollar throughoutthe flexibilization period, thus continuingto fulfill the task of main nominal anchor. The first deval".ation in fact took place much later, in January 1987. The approachwas different in Mexico where controlsfor most private sector prices were never in effect. After the initial freeze the situation remainedbasicallyunchangeduntil December1988, when the new administrationcame to power. The first substantivechanges in the program were announced at that time and consistedof an annual increase in minimum wages and public sector prices and thc adoptionof a preannounced devaluationschedule. This was followedby an extensionof the Pacto until March 1990. In Mexico the authoritieshave so far been able to maintain control over the nominal anchors during most of the flexibilizationstage, though prices have not been fully liberalizedand in this sense this stage

12 There had always been price controlsin the Israeli economy for public utilities,some foodstuffsand for those industrieswhere firms had monopoly power. By January 1987 the nature and comprehensiveness of control returnedto the pre-stabilization period.

26 can not be consideredas completed. The design of this stage in Israel and Mexico stands in sharp contrastwith the two failuresof the eighties. Neither the Austral or Cruzado plans had a strong nominal anchor during the flexibilizationstage. In both programs all the initialanchors were abandonedat that time, as wages, controlledprices and the exchangewere essentiallyadjusted accordingto past inflation. The outcome is less clear cut in the programsof Uruguay and Argentinaof the sixties. Although it is difficult to trace the beginning of flexibilizationto a specificdate, it probably started a few months after the programswere launchedcoincidingwith increases in wages and some controlledprices. The increasedflexibilitywas undertakenwithout the eliminationof price controls. The end of the flexibilizationperiod in Uruguay could be linked to the devaluationof March 1972, while it is more difficultto determine it in Argentina,where the exchange rate remained fixed ance controlson prices and wages had been in practice removed. The Brazilianprogram of the sixties is a unique case which deserves separatetreatment. Price controlswere not removed during the flexibilizationstage; in fact their use was intensifiedin 1967 as they 13 became mandatory for most of the large industrialsectors.

Flexibilization,insteadwas initiatedwith the adoptionof a crawlingpeg rule for the exchange rate and with the introductionof a new indexation

13 Indeed two institutesthe CONEP (NationalCommissionof Price Stabilization),and the CIP (Interministerial Price Commission)were created at that time to supervisecontrols (see Simonsen (1974)).

27 rule for wages. D. Fiscal and Monetary Policies a. Fiscal Policy An adequate and sustainablereductionin the budget deficit is a necessary condition for maintainingprice stability,particularlyin the medium term. In almost all the programs includedin this study there was an initial reductionin the budget deficit. There is less homogeneity regardingthe sustainabilityof the fiscal effort. In the Brazilian program of the sixties,and in the Israeli and Mexican programs of the eighties,the reduction in the budget deficitwas maintainedthroughout, while in the two Argentine programs,the Cruzado and the Uruguayanprogram, the reductionin the deficit did not survivethe initial stages. The role of fiscal policy is particularlyimportant in the second stage of a heterodox stabilizationprogram. A tight fiscal starnceat this stage is crucial to ensure that demand pressures remain under control at the time in which prices are flexibilized. A significantdifferenceexists between the size of the fiscal adjustmentin the two eras. The reductionsin the budget deficit were more impressivein the eightiesmainly because the fiscal imbalanceswere also larger. In Argentina the deficit was cut by over 8 percentagepoints of GDP between the first and third quarter of 1985; likewise,a reductionof over 7 percentagepoints was effected ir.Israel over the same period. In the sixties,on the other hand, the fiscal adjustmentin Brazil only require4 a cut in the budget deficit of approximately3 percentagepoints of GDP, while in Uruguay and Argentina the reductionswere even smaller. The large adjustmentof the eightieswas facilitatedby the drastic fall in

28 inflation which partly offset the erosion in governmentrevenuesresulting from the Olivera-Tanzieffect. In the successfulcases of both eras the fiscal adjustmentaechieved roughly a balance in the fiscal accounts,and in two instances (Israelin 1986 and Mexico in 1987) there was a small surplus in the operational balance of the public sector. Large reductionsin the deficit which fall short of the non-inflationaryfinance target are bound to fail, as it happened for example in the Austral plan where a cut in the deficit of about 7 percentagepoints of GDP was not enough to sustain stability. Also importantis the nature of the reductionin the budget deficit. In the success stories (Brazilin the sixties,Israel and Mexico) the fiscal adjustmentwas effectedin a balancedway through reductionsin expendituresand increasesin tax revenues (see tables 1, 6, and 7). Reductionsin governmentexpendituresare particularlyimportantto increasethe credibilityof the program and the sustainabilityof the fiscal adjustment.In the failed experiments,on the other hand, most of the reductionin the deficit was effected through increasesin tax revenues while very little was done on the expenditureside. b. Monetary Policy A casual look at the behaviorof monetary aggregatesin the heterodox programsof the two eras suggeststhat monetary policy did not play a central role in most of them. Money supply was primarilyaccommodatingin the programs of the sixties,despite some tighteningin the supply of domesticcredit to the public sector. Tables 1 through 3 show the annual rates of growth of Ml and H2, which during those years were proportional to, and in most years exceeded,the increasesin the price level. We only

29 find some evidenceof tight money in Brazil in 1966, when M2 expandedby only 22 percentwhile inflationremainedat around 40 percent per year and in Uruguay in 1968. A similarpattern is observedin the eighties,where although tight domesticcredit was used in some instances (especiallyin Israel and in Mexico during the first months),monetary variablesdid not perform the role of nominal anchor. In the Austral plan monetary policy appears to have been accommodating,while the growth in monetary aggregatesexceeded inflation,this was not necessarilyinflationarybecause the demand for money was expected to ittcrerse in the new low inflationenvironment. The Cruzado plan adopted the op,osite,unsound approachas the authorities opted for a large expansionin the money supply at the outset in order to bring down interestrates quickly. High real interestrates is a feature observedin most of the heterodoxprograms of the eighties. As can be seen in figure 3, specially in Israel and Mexico during the first stage, the initial rapid fall in inflationwas not accompaniedby a proportionalreductionin nominal interest thus yieldinghigh real ex-post interest rates. Lack of reliable evidencemakes difficult to examinewhether a similarphenomenonoccurred in the sixties. Interest rates in the officialmarket (for which data exists) were under governmentcontrolwhile data on curb market rates is difficult to obtain. The existenceof high interestrates was used as evidencethat monetary policy was tight during this period.1 4

14 See for example Dornbusch (1986).

These high real rates,

30

however, could insteadhave reflectedlack of credibilityon the sustainabilityof the program,particularlyon the ability of the central bank to peg the exchange rate. This alternativeexplanationwould argue that while real interestrates were high in an ex-post sense, ex-ante real interest rates were not necessarilyhigh. This second view perhaps providesa more appropriateexplanationfor the persistenceof high real interest rates, speciallybecause in economieswith open capital markets, as is the case in Mexico, the monetary authoritiescannot control the money 15 supply over extendedperiods.

,7.

Outcomes of the Programs

The heterodoxprograms of the two eras were successfulin bringing down inflationduring the first stage. The use of income policieswas an importantasset to achieve rapid reductionsin inflation. In many programs however,price stabilitywas difficultto maintainbeyond the first stage. The failuresof the Argentine and Uruguayanprograms in the sixties,and of the Austral and Cruzado plans in the eightiesare evidenceof this fact. The medium term success storieswere few: Brazil was able to maintain low inflationfor over eight years in the sixties,Israel has been successful for over four years, and Mexico has maintainedits program in course for almost two years. While there are many similaritiesin the evolutionof the successful programs, they share some common featureswith the less successfulones,

15 For empiricalevidenceon Mexico see Cumby and Obstfeld (1983).

31

particularlyin the first stage. The Israeli and Argentine programs of the eighties,discussed in more detail in Blejer and Liviatan (1987),yielded similar results during the first few months. Likewisethere are many similaritiesbetween the outcomesof the Brazilianprogram of the sixties and the Krieger Vasena program. Developmentswere also analogous in some of the failed experiments,as for example between the Uruguayan and the Cruzado plans. In the remainderof this sectionwe will compare the evolutionof the various programsduring their various stages. We will attempt to highlight those aspects that are unique to each program and that are importantto understandtheir success or failure. A. Inflation a. The Eighties The four programsof the eightieswere successfulin sharply reducing inflationduring the first stage. The effectivenessin this respectwas common to all programs, regardlessof whether or not they undertook a large fiscal adjustment;the short run reductionin inflationwas similar in the Cruzado plan where the budgetary situationwas ignored and considered irrelevantas it was in Israel,where the deficit was cut substantially. The easinesswith which inflationcan be brought down in the short run is a strikingcharacteristicof the heterodoxapproach; a shock treatmentbased on price and wage controlsis usually very effective in pulling down inflationvery quickly,in chronic,high inflationcountries. The resultswere mixed during the flexibilizationstage. This is usually the criticalperiod of a heterodox stabilizationprogram since it is at this stage when the overall design and implementationof the policy package becomes an importantelement for its success. Two aspects of the

32 program are part"nularlyimportantto sustain the stabilizationeffort during this stage. First, there is a clear need to maintain the 'fundamentals*at levels consistentwith price stability. The fiscal accounts should be roughlybalanced,while in some cases an over-adjustment might be needed to support a policy of generalizedexcess supplieswhile prices are flexibilized,and to provide an unambiguoussignal about the authorities'commitmentto maintain the stabilizationprogram. Second, it is necessary for the system to have a strong nominal anchor in order to avoid increases in inflationwhich are disproportionate with the underlying situationin the fundamentalsduring flexibilization. The failuresof the Cruzado and Austral plans surfacedduring this stage (as can be noticed from figure 1-a). In Brazil, a policy of unsustainablebudget deficits,high real wages and loose monetary policy resultedin a situationof overall excess demandswhich led to a sharp accelerationin inflation startingin November 1986. Flexibilizationin the Austral plan started in April 1986. Inflationincreasedduring this period, although in a much more moderateway than in the Cruzado,mainly because the fiscal imbalancewas much smaller. This increase in inflation reflectedboth a lack of fiscal ar.dmonetarydiscipline,and the absence of a strong nominal anchor. As can be seen in figure 4-a, the fiscal adjustmenteffected in the second semesterof 1985 started to be relaxed prior to the flexibilizationperiod. Since there had also been an expansion in expendituresby the private sector, this meant that flexibilizationtook place in an expandingeconomy. As importantwas the lack of a strong nominal anchor; the ones used in the first stage were not maintainedduring the flexibilizationstage. New rules were announced for

33 wages, public sector prices and the exchangerate which in essence amounted to backward indexationof nominal variables. Israel and Mexico, on the other hand, successfullycompleted the first stage and were more effectivein undertakingflexibilization. In both cases a tight fiscal stance was maintainedthroughout;Israel overadjusted on the fiscal side and had a budget surplus in 1986; Mexico started from a surplus in 1987 and while it had a small deficit in its operationalbalance in 1988 (causedmainly by a deteriorationin the price of oil in world markets and a substantialincreasein domestic real interestrates), it sustaineda large surplus in its primary balance (figure4-a). Both countries also maintainedstrong nominal anchors through the flexibilizationstage, and attributedto the exchange rate a central role in this phase. The flexibilizationperiod started in Israel in January 1986, six months after the implementationof the program. There was a gradual lift of controls of private sector prices, followedby greater flexibilityin wages. The exchangerate remainedessentiallyfixed to the US dollar throughoutthis period, except for the fact that startingin August 1986 the shekel was pegged to a basket of currencies. In Israel the fixed exchange rate was somewhat easier to maintain than in other countriesthe Shekel depreciatedby 17 percent in real terms against the basket of currenciesas a result of the fall of the Dollar in the world markets. In Mexico, the first adjustmentin nominal variables took place in December 1988, coincidingwith a decisionto extend the Pacto. It was announceda rule for the devaluationof the Peso (1 peso per day) until June 1989, amounting to approximately8 percent for the whole period. The extension

34 includedincreasesfor minimum wages and controlledprices, while most prices in the private sector were allowed to be freely determinedas before. Tight fiscal policy in conjunctionwith the maintenanceof a strong nominal anchor were effectiveto prevent an increase in inflationduring the flexibilizationperiod. In both countries,however, annual inflation remainedin the 20 percent step, and it has proved extremely difficultto bring it down below that level. In fact, this appears to be a broader phenomenonobservednot only in successfulheterodox programsbut also in orthodoxprograms such as those of Chile and Bolivia. b. The Sixties The three programsof the sixtieswere also successfulin bringing down inflation. The beginningof the income-policiesbased stabilization in Brazil could be traced to 1965 coincidingwith the introductionof voluntary price controls. As can be seen in figure 1-b there was a sharp decelerationin inflationin the third quarter of that year, followedby an almost continuousfall until it basically stabilizedin the low twenty percent range by 1967. Likewise,the fall in inflationwas also slow but relentlessin the Krieger Vasena program followingthe introductionof controls in the second half of 1967. Inflationfell from 25 percent in 1967 to 9.8 in 1968 and to 6.8 percent in 1969. More dramaticwas the fall in inflationin Uruguay after June 1968; inflationwas brought to a halt in just one month, and the price level remainedstable for the remainderof the year.1 6

This supportsthe findingsof the eightiesthat a heterodox

16 Inflationfell from 13.5 in June, to just over half a percentage point in July and the price level actuallydropped by 1.6 in August.

35

shock treatment in cases where inflationis high and acceleratingcan produce spectacularresults in the short run. The three programsevolved along differentpaths after the initial reduction in inflation. Brazil was the most successfulcase, where inflationremained low for a number of years while it experiencedvery high rates of growth. As can be seen in table 1, inflationfell even further during the early seventiesreachingjust 12 percent in 1973. This spectaculareconomicperformancefor over six years in which low inflation was accompaniedby robust growth is now widely known as the Brazilian Miracle. Tlhefall of inflationwas less persistentin Argentina and Uruguay, but still lasted for over two years. Persistenceon the fiscal side was a crucial element for the success of the Brazilianprogram. This was supportedby a nominal system in which price controlswere the main nominal anchor. The relianceon income policies for the purpose of controllinginflationon a long term basis was a unique feature of the Brazilianprogram. This un-orthodoxapproachwas necessary in order to provide a strong nominal anchor to the system, since the other two originalanchors (wages and the exchange rate) were by then being adjustedin accordanceto past inflation,while monetarypolicy had been accommodatingfrom the outset. While it is difficultto establishthe beginningof the flexibilizationstage in Argentina and Uruguay, the evidenceindicates that some flexibilizationtook place in both programswithout igniting significantinflationarypressures. In the Krieger Vasena program, for example, the two first wage increases (in January 1968 and 1969) did not lead to any noticeableincreasein inflation. Likewise, in Uruguay the

36 increase in wages in the private and public sectorseffected in December 1968 and in December 1969 did not appear to have had an effect on 17 inflation.

In both countries this was possiblebecause the economy

maintainedtwo strong nominal anchors,the fixed exchange rate and price controls. The eventualfailure of the programs of the sixties in Argentina and Uruguay can be tied to the lack of persistencein maintaining fiscal discipline. This can be seen from figure 4-b, which shows the behaviorof inflation,devaluationand budget deficitsover the period. In both cases the budget deficitsdisplay a U shape. There was an importantfiscal adjustmentat the beginningwhich lasted for approximatelytwo years which was critical for the initial positive resultsof the anti-inflationeffort. This was followed,however,by a relaxationat a later stage in the fiscal effort which led to a resurgenceof inflation,although this was partly postponedby delayingexchange rate devaluations. In the two programs, the deteriorationin the fiscal balance preceded the resumptionof inflation. The fiscal cycle is a feature observed in many stabilizationprograms. In Argentina,for example, all major stabilizationefforts during the last three decades startedwith significantreductionsin the budget deficit, which in all cases were later abandoned (see Kiguel and Liviatan 1988). B. Output and Unemployment The heterodoxapproach is primarilyused as a way to limit the costs of reducing inflationusually encounteredin orthodoxprograms. The availableevidence indicatesthat in most cases the programshave been

17 For details on these increasessee data in Viana (1988).

37

successfulin at least limitingthe unemploymentcost of disinflation during the first stage. Figures 5-a to 5-b show the behaviorof these variables in the programs under consideration. None of the programs led to large increasesin uraemployment or to a prolonged recessionaryperiod. Just the opposite,in both the programs that succeed and in those that failed there was an expansion in the economy that started during the first stage. In the Cruzado plan, for example,the program resultedin a sharp increasein industrialproductionand a corlsequent reductionin unemployment. The economicboom lasted until November,when the authoritiesmade an adjustmentin policies includinga devaluationand the removal of the freeze. Although in Brazil the expansionwas partly caused by an ir.tial set of policies favoringexcess demandswhich made the Cruzado 'incredible",a similar pattern for output and unemploymentis observed in Israel and in the Austral plan. Even in the Israeli program which adopted and persistedon a tight fiscal stance, the economy started to experiencean expansionduring the first stage that would last for over two years. After a small increasein unemploymentin the third quarter of 1985, unemploymentfell and remainedlow until the end of 1987, while industrialproductionremainedstrong. Likewise, industrial productionstarted to expand in Argentina in the first months of the program, at a time when the fiscal situationremainedunder control and the program could be thought to have at least some chance of success. While Mexico was also able to cut inflationwithout noticeablereductionsin economic activity,we do not observe the expansionshared by the other .programs. The programs of the sixtieswere even more effective in terms of

38 growth. As can be seen from figures 5-b the period of price stabilitywas accompaniedby sustainedhigh rates of growth. In Brazil, for example, the annual rate of growth between 1i67 and 1973 exceeded11 percent,while Uruguay experiencedthe highest growth rate of the decade (averaging5 percent per year) in 1969-70. These systematicincreasesin output are difficultto explain using traditionalaggregatedemand based type of arguments. The tighteningin fiscal policy observed in most programs, at least in the early stages, and the initial fall in real wages would a priori lead to a fall in aggregate demand and output. The recent work emphasizingthe intertemporalaspects of consumptiondecisions is perhaps more vueful to understandthis phenomenon. An increasein demand could take place if the exchange rate based stabilizationprogram is not credible and hence is perceived as temporaryby private agents (as in Calvo 1987). In that case, if agents anticipatethat there will be a balance of payments crisis in the near future resulting in a devaluationand in the introductionof import restrictions,consumptionof tradeablesand hence importswill increase initially. Consumptionof domestic goods will also increase if domestic and foreign goods are complements. Lack of credibilityon the sustainabilityof the stabilizationeffort was an important endogenousforce behind the increasein output in all the programs. In the programs of the sixties this was reinforcedby government policiesaimed at increasinginvestmentwith the objectiveof restoring long term growth. In particulartV- announcementof the programs included as objectivesthe need to maintain the availabilityof credit to the private sector (part cv which was providedat subsidizedinterest rates),

39

while in the three programs there was an increasein the public sector investment. The increasein output, however,was not sustainedin most programs and the costs of stabilizationwere faced at a later stage. In Israel, a successfulprogram, the initialexpansionwas reversedat a later stage and the economy entered a recessionaryperiod in 1988 which had even turned worse recently. The cost was also paid later in the failures;in Uruguay, the Austral and the Cruzado there was a reductionin economicactivity in the aftermathof the failures. C. The Trade Balance,Consumption,and InvestLient The increase in output was accompaniedby a deteriorationin the trade balance. This was most dramatic in Uruguay and in the Cruzado plan, althoughas can be seen from figures6-a and 6-b this phenomenon is observed in almost all the programs. This behavior of the external accounts is just opposite to the one observedin the orthodoxprograms where the initial impact was a sharp improvementin the externalaccounts (see Kiguel and Liviatan1988). In the programsof the sixties the increasein demand was driven by a combinationof increasesin private consumptionand total investment (privateand public). The expansionin investmentwas partly due to special tax and financial incentiveswhich were implementedduring the programs. In the eighties,on the other hand, while there was also an expansionin consumption,investmentremainedrather depressed. As can be seen from tables 4 to 6 there were reductionsin investmentin Brazil and Israel,and to a lesser extent in Argentina. Two factorsmight be importantto explain the dissimilarbehavior of

40 investmentin the two eras. First, by and large the stabilizationprograms of the eightieswere launched in a more unstablemacroeconomicenvironment in which the fiscal problemsand the debt crisis made long term investment decisions especiallydifficult. Thi was not the case in the sixties, where a) the size of the fiscal disequilibriawere much smaller (e.g. in Argentina the budget deficitwas 3 percent of GDP in 1966, comparedwith 12.6Z in 1984), and b) the countries (bothpublic and prlvate sectors)had access to internstionalfinancialmarkets. Second, the high real interest rates that were observed in the eightiescreated an additionalbias against investment. The programsof the sixties,on the other hand, did not encounter this difficultymainly because they were implementedin financiallyrepressedeconomieswith controlledinterestrates. In addition, there was a deliberateeffort by the authoritiesat that time to increaseinvestment,particularlyin the public sector, since the plans were designedas comprehensivestrategiesaimed at the double objectiveof stoppinginflationand restoringgrowth. The better fiscal situationof the sixtiesmade this increasein public investmentfeasible. Private investmentwas stimulatedthroughadditionallines of credit with favorable terms and tax benefits. D. Real Exchange Rate and Real Wages There were also similaritiesregardingthe behavior of real wages and the real exchangerate (see figure 8-a and 8-b). In most programs the initial maxi-devaluation was followedby an exchange rate freeze. Since in general inflationtends to remain above internationallevels during the exchange rate freeze, this policy results in a continuousappreciationof the real exchangerate. Wages were also frozen at the beginning in most

41 cases, although they vere graduallyflexibilizedover time. Real wages portray a U shape, they deterioratedduring the inLtial freeze but recoveredprevious levels over tise. The reaeon for thiL pattern is related to the design of the programs. As discussed ln Blejer and LLviatan (1987),the deslgn of an lncomes policiesbased stabllizatlonprogram calls for an initLal reductionln real wages accompaniedby a deprecLationln the real exchange rate. As the exchange rate usually is the main nominal anchor, a central objectLve is to keep it fixed for as long as posmible. In order to maintain the exchange rate fixed while inflationcontinues (probablyat much lower levels) as a result of "inertial"elements, the initialdevaluationshould produca an overshootingof the real exchangerate. Real wages, on the other hand, should be reduced initiallyto support a policy of excess suppliesand be allowed to increase over time as the overall situationimproves. Nevertheless,even this type of strategycannot guarantee that the real exchange rate will not be overvaluedat a later stage. In this respect, the three programs of the sixties and the programs of the eightiesdisplayeda similarpattern. As can be seen from figures 8-a and 8-b, there were nominal devaluationsat the beginningwhich coupled with the impositionof price controlswere effective in achieving an initial real depreciationof the currency. This period was followedin most cases by a continuousappreciation,as prices continued rising while the exchange rate remainedfixed to provide a nominal archor to the system. In some cases, such as the Cruzado, the appreciationoccurredvery rapidly, while in others, namely in Israel and in the Krieger Vasena program, this was a slower process.

42 In all cases, regardlessof their success in controllinginflation, the authorities faced great difficultieswhen trying to determinewhether and when a devaluationwas warranted. There is usually a trade off between avoidingan overvaluationof the currencyand a deteriorationon the externalbalance on the one hand, and the fears that the devaluationmight destabilizethe anti-inflationprocess on the other. The available evidence suggeststhat the authoritiesare usually reluctantto undertake a devaluationonce a program of this type is in place. Uruguay in the sixties and the Cruzado plan in the eightiesare clear examplesof cases where as a result of aa unwillingnessto devalue the currencybecame grossly overvalued. As can be seen from figure 9-a and 9-b, in both cases the depreciationof the exchangerate in the black market started to take place well before the officialdevaluationsin the officialexchange rate were finallyeffected, resultingin large premiums on the parallel exchange rate. The size of these premia made the devaluationall but unavoidable. The situationwas more dramaticbecause, as can be seen from figures 7-a and 7-b, in both episodes the period of appreciationwas accompaniedby significantdeteriorationsin the trade balance. The devaluationscame too late and were followedby a rapid accelerationin inflation,highlightingo.aeof the biggest dangers of this type of policy. Even in the more successfulprogramsthere were periods in which the domestic currencybecame overvalued. The Cruzeiro (in the Brazilian program of the sixties)and the Shekel in Israel during the eighties experiencedovervaluationduring the fixed exchange rate period. The Mexican evidence is less clear in this respect,althoughthere is some

43 indicationthat the currencywas overvaluedby the end of 1988. The approachadopted to overcomethis difficultyvaried across countries. Israel and Argentinaunder Krieger Vasena opted for step devaluations,while Brazil in the sixties and Mexico more recentlymoved to a crawling peg. There are trade-offsbetween these two approaches. A step devaluationhas the advantageof providing a nominal anchor to the system and in this respect it can send a strongersignal regardingthe determinationof the governmentnot to accommodateongoing inflation. The main risk with this strategyis that the devaluationcan be delayed,and hence that a larger devaluationwill be needed in the future with its potentialdestabilizingeffect on inflation. The main advantageof the crawlingpeg is that, through very frequent adjustmentsin the exchangerate, it greatly reduces the risks of gross overvaluationof the domesticcurrency. In additionit avoids the cycles in the real exchange that are bound to occur under step devaluations. The main disadvantageis that the exchange rate may lose its role as nominal anchor when it takes a passive role accommodatinginflation.

V. Long Term Aspects of the programs

A. The costs of disinflationin the successfulheterodoxprograms A central reason for followingthe heterodoxapproach is to dampen the costs of disinflation. While the heterodox approachappears to be effective in this respect during the first stage, the available evidence suggests that difficultiesmust be faced at some stage (eitherbefore or kfter the heterodoxphase), and that at least in the successfulprograms

44 the costs of disinflationhad to be born at some point. A full evaluation in this respect thus requiresthat the programsbe examined from a longer term perspective. In Israel, for example,the costs in terms of unemploymentwere very small during the first two and a half years of the program. There was a very short recessionaryperiod (less than a year) at the beginning followed by a strong expansionin 1986 and 1987. The recessionaryphase came much later, but it has extendedfor over a year with unemploymentrates exceeding9

percent in 1989. To what extent was the ensuing recession

related to a program that started over two years earlier? It essentially was, accordingto a line of analysisdiscussedin more detail in Liviatan (1989);once controlswere removed the program became in effect an exchange rate based orthodox stabilizationprogram and hence the resultingcosts were similar to other programs of that type. In the other two success stories,Brazil of the sixties and Mexico, the orthodoxphase preceded the heterodoxperiod. In Brazil, the worse part of the recessionoccurredbetween 1963-65,with some recoveryduring the first stage of the heterodoxperiod (1966 and 1967), and strong growth thereafter. Likewise in Mexico, economicgrowth was dismal during the orthodoxphase, but unlike other heterodoxprograms it did not recover during the heterodoxphase. This evidence indicatesthat the heterodoxapproachcannot be expectedto avoid altogetherthe costs of disinflation. A recessionary period can be consideredas a test for the authoritiesregardingtheir determinationto maintain low inflation,and in this way provide an importantsignal to the private sector. This means that at some point the

45 authoritieswill have to refrain themselvesfrom accommodatinginflation (via monetary or exchange rate policies) in order to sustain the stabilizationeffort and eventuallysucceed in controllinginflation. B. The costs of failure A common feature in heterodox programs is the reluctancegenerally to the fixed displayedby the authoritiesto devalue. The over-comr.itment exchange rate is speciallydangerouswhen there are large underlying imbalancesresultingfrom substantialbudget deficits. In most of the failuresthese imbalancessurfacedprior to the devaluationin the form of large trade deficits, increasesin the premium on the black market exchange rate, an overvaluationof the domesticcurrencyor high real interest rates. An extreme example of this phenomenonis Uruguay in the sixties. In that case there were clear indicationsof macroeconomicimbalancesby 1970. In that year, as can be noticed in table 3, the current account and the balance of paymentswere in deficit,the exchangerate started to show signs of being overvaluedand there was a significantrise in the premium on the parallel exchange rate (see figure 9-b). The macroeconomic situationunderwent further deteriorationin 1971, as the current account deficit increasedaccompaniedby a further real appreciationof the domestic currencyand a drastic increasein the premium. Throughoutthis time the exchange rate remainedfixed. A devaluationwas undertakenin late 1971, but it was followedby a significantincreasein inflationand a devaluationinflationcycle (see figure 1-b). The imbalanceswere not as pronouncedin the Krieger Vasena program where some signs of macroeconomicproblemsemerged in 1969. During that

46 year the exchange rate displayed some indicationsof overvaluationwhile there was a large deficit in the current account (see table 3). The imbalances,however,were much smaller than in Uruguay, partly because the fiscal situationhad not yet deteriorated. The responsewas a 25 percent devaluationin June 1970, which was effectivein restoringexternal balance. Despite some increasein inflation,it was not as dramaticas in the case of Uruguay. This differencein the timing of the adjustmentin both countriesis importantto understandthe ensuing developmentsin inflation. In Uruguay the adjustmentwas long overdue by the time of the devaluation. Postponing the adjustmentwas a costly strategybecause a massive devaluationwas required and the result was a significantincreasein inflation. In Argentina,on the other hand, the imbalanceswere smaller and the adjustmentwas more gradual. A similar differenceis observedin the speed of events after the initial failuresof the Austral and Cruzado plans in the eighties. The exchange rate freeze was overextendedin the Cruzado plan and the result, as can be noticed from figure 1-a, was an explosivepath of inflation immediatelyafter the devaluationwas undertaken. During the Austral plan, on the other hand, the freeze was abandonedwhile imbalanceswere not as large and, at least initially,the result was a more moderate increase in inflation. The overall performanceof the economy deterioratedin the aftermath of the programs in all cases. In the two failuresof the sixties the countriesexperiencedsustainedgroath during the period of price stability. Annual GDP growth in Argentina averaged6 percent between 1968

47

and 1970, and 5.4 percent in Uruguay in 1969 and 1970. This trend was reverseOduring the inflationaryperiod, with the rate of growth in Argen .. ta averaging 2.3 percent during 1971 and 1976 while it was slightly negative in Uruguay between 1971 and 1973. A similarpattern can be observedregarding investment(see tables 2 and 3), which in both countries increasedstronglyduring the period of macro-stabilitybut experienceda sharp fall as inflation rekindled. The evidenceis less clear on the external side, where there are no marked differencesbetween the two periods,as can be seen from the evolutionof the trade and current account balances. It should be pointed out, however, that the good external performanceduring the high inflationyears was partly due to an improvementin the terms of trade, and that in Uruguay this was supported by an adjustmentin the fiscal imbalance. The performancein the eighties in Brazil and Argentina after the failuresof the Austral and Cruzado plans has been dismal. By and large both countries experienceda period of high uncertaintyabout macroeconomic performancecharacterizedby low and erratic growth, recurrentbalance of paymentscrises, high real interestrates and pronouncedinflationstabilizationcycles. These cycles (shown in figure 1-a)

are a novel

feature and are a direct consequenceof the failure of the heterodox approach. As discussed in Kiguel and Liviatan (1989.b)both countries entered a new regime in which inflationcan acceleratevery quickly and even reached hyperinflationlevels, only to be stopped through the reimpositionof controls. This regime is self defeating,as inflation accelerateseach time controlsare removed awaitingthe new round of controls.

48 C.

Sustainingprice stabilityin the long run It is difficultto pinpointthe moment in which a stabilization

program can be consideredas completedand a new era starts in which the problems become more akin to those of low inflationeconomies. When price stabilityis achievedin economiesthat traditionallybelonged to the group of chronic inflationcountries,the low inflationstage is likely to differ from that of the traditionallow inflationeconomies. For example, if the countriesdeveloped indexationmechanismsduring the precedinghigh inflationperiod those mechanismswill be difficultto eradicate in the new equilibrium. This happened in Israel where, paradoxically,the trigger rates of inflationunder which wages would be adjustedwere smaller in the low inflationera than in the previousone. In these episodes,the indexationrule and other institutionalarrangementsadopted after the stabilizationprogram were a reflectionof the the inflationhistory of the economy and of the characteristicsof the stabilizationstrategy. For example, indexationis difficultto eradicatein countrieswhere failed stabilizationis the rule rather than the exception. It is useful to compare the policiesand strategiesadopted to maintain low inflation in the only two cases for which we have a long enough perspective,Brazil in the sixties and Israel in the eighties. An importantdifferencebetween these two successeswas the philosophytowards indexation. In Brazil, the arrangementsin the labor markets, namely wage indexation,and the exchange rate rule were maintained throughoutthe period of price stability. Institutionalarrangementswere not changed during this period, althoughthey are certainly less justified in an economy with low inflation. Israel has taken the opposite approach

49 in dealingwith this institutionalinertia. The authoritiesare attempting to remove indexationfrom the economy as a way to reduce the destabilizing effects that these arrangementsmight have if the economy faces an external shock. The Brazilianprogram of the sixties providesthe most interesting insightsfrom a long term perspective. The long period of low inflation was disrupted in the mid-seventies,coincidingwith the first oil shock. Although the increase in inflationat that time was a world-wide phenomenon,inflationnever came down from the higher levels. Just the opposite, inflationreached a new plateau (of around 45 percent)and stayed there until 1979, the time of the second oil shock. The increasein inflationin 1973-75 was caused by a combinationof factors,some internal,other external. According to estimatesby Malan and Bonelli (1977),Brazil reached full capacityduring 1972. Traditional arguments suggest that the economy was becomingoverheatedand that in the absence of restrictivepoliciesit would enter an inflationaryperiod. On the external side, the oil shock had importantcost effects,particularly because it led to large devaluations(19 percent in 1974 and 22 percent in 1975) to deal with the large imbalancesin the trade and current account balances. Why did inflation remain in the new higher plateau once the economy adjustedto the shocks? The answer to this questionhas been partially addressed in section IV. The system developedduring the "miracle"years was not strong enough to confrontthe nominal shocks of the mid-seventies. Monetary and exchange rate policieswere basically accommodating,while wages were adjusted accordingto an indexationrule. Price controlsbecame

50 the main nominal anchor of the system,and was in effect the only available mechanism to avoid accelerationsin inflation. Israel adopted a differentstrategyto tackle this problem. The initial relianceon price controlsand the fixed exchange rate was slowly being replacedby one in which the exchangerate and high interest rates plaved more importantroles as a stabilizationinstrument. There has also been an effort to limit the extent of backward indexationo. ages from the economy in order to ensure greater flexibilityin relativeprices. Although this seems to be an appropriatepath, more time will be needed to evaluatewhether this strategywill lead to sustained low inflationand to assess what are the costs of a continuouscommitmentto using the the exchange rate as the main nominal anchor in the process of bringing down inflation.

VI. Final Remarks and Policy Issues

Heterodoxprograms are usually adopted as an alternativeto the orthodoxapproach in an attempt to minimize the costs of bringingdown inflation. The evidenceobtainedin this study indicatesthat the heterodox approachcan indeed bring down inflationvery rapidlywithout large costs in the short run; this stage, however, is the easy part of a heterodox programs. The costs have to be born at a later stage when the authoritieshave to concentrateall their efforts to sustain low inflation. This second stage is the difficultand costly part of heterodox programs. The magnitudeof these costs and how do they comparewith those of an orthodox program are yet not clear, more evidencewill be needed in this

51

respect. What is clear is that while these costs appear up-front in orthodoxprograms they are delayed in the case of heterodox plans. Whether the heterodox program should be based on a gradual or shock strategylargely depends on the rate of inflationprior to stabilization. In chronic,high inflationeconomiesa shock strategyis generallymore appropriate. The main advantageof the heterodox strategyis in facilitatinga rapid transitionto low i.lflation.When the orthodox approachwas followedin chronic inflationcountriesthe reductionsin inflationwere typicallyslow and in this sense they underminedthe support for the programs. The heterodox approachhas an advantage in this respect, since the initial success in reducinginflationincreasesthe overall support to the program thus opening the possibilityfor the introductionof additionalmeasures on the fiscal side to deepen the stabilizationeffort. This opportunity,however, is usually foregone,only Israel and to some extent Mexico used this success to introducefurther fiscal adjustment. In most cases, as for example in the Austral and Cruzado plans, the initial success is taken as evidence that more fiscal adjustmentis not required leading to a completecollapseof the stabilizationeffort. A gradual approachis more appropriatefor low inflationcountries. This is in fact the strategyfollowedin Argentina and Brazil in the sixties. Income policies probablycontributedto a faster reductionin inflation in both countries,but their role was not as criticaland their effects not as dramaticas in the shock programs. The case for income policies is much weaker in these cases. Tight fiscal policy and a strong nominal anchor are two critical componentsfor the success of a heterodoxprogram. The need to maintain a

52 strong fiscal positioncertainlygoes beyond the first stage and is particularlyimportantduring the flexibilization period. In the failures of both eras, the budgetary situationhad already deterioratedby the time controlswere removed. This is speciallyclear in the Cruzado plan in the eightiesand in Uruguay in the sixtieswhere a relaxationin the fiscal stance during the period of the freeze led to a sharp accelerationin inflation later on. There is also a need to maintain a strong nominal anchor throughoutthieprogram. Price and wage controlsand the exchange rate freeze provide enough strong nominal support at the beginning. Countries found it more difficultto maintain the nominal anchors during the flexibilizationstage. In Israel, for example, the exchange rate had the role of main nominal anchor during the flexibilizationperiod and while it helped to prevcnt a rekindlingof inflationit was at the cost of overvaluationof the currency. A central message emergingfrom this paper is that the flexibilizationperiod is the most critical time in a heterodox stabilizationprogram. The availableevidence indicatesthat policy makers do not perceive that this is the case. In the Austral plan, for example, the initial freeze was preceded by carefulplanning includinga sophisticatedmechanism to avoid wealth redistributionresultingfrom the sharp and unexpectedreductionin inflation (the desagio). Likewise, the Cruzado plan includedan elaboratearrangementto avoid large disparities in real wages across sectors. There was no planning,on the other hand, on policiesto support stabilityor on a strategyto get out of the freeze. An additionalmessage is that there is not much point in putting effort in designinga sophisticatedprogram if there is no determinationto support

53 it with sound fundamentalpolicies. The failuresin both eras did not result from poor design on the income policiespart of the package. They failed because of lack of persistenceon the fiscal side. In a previouspaper we argued that orthodoxstabilizationprograms representa costly and difficult approachfor stoppinginflationin chronic inflationcountries. Are heterodoxprograms the alternative? On the basis of this paper we are inclinedto favor their use, althoughwe now understandbetter that the costs of disinflationunder this approach are larger than we originallythought. We also learned that the misuse of heterodox programs is very common and that the costs of failure are likely to be greater than under orthodoxprograms. While there is no easy way to achieve price stability,the heterodox approachcan be effective if fiscal disciplineis maintainedand there is persistencein sustainingthe stabilizationeffort. In fact, there are reasons to believe that in high, chronic inflationcountries the heterodoxapproachhas an edge over orthcdox. Heterodoxprogramsusually have smaller costs initially,while the costs that they face in the later stages associatedwith overvaluation of the currer.cy are very similar to those observedin exchange rate based orthodoxprograms.

54 VI. References

Barro, Robert J. (1986). 'Reputationin a Model of Monetary Policy with Incompleteinformation,'Journal of MonetaryEconomics,Vol 17, January, pp. 3-20. Barro, Robert J. and David Gordon (1983.a). 'A PositiveTheory of Monetary Policy in a Natural Rate Model," Journal of PoliticalEconomy,Vol 19, August, pp. 589-610. (1983.b).

'Rules, Discretion

and

Reputationin a Model of Monetary Policy,'Journal of MonetaryEconomics, Vol 12, pp. 101-21. Blejer,M. and LW,iatan,N. (1987),"Stabilization Strategiesin Argentina and Israel, 1985-86', IMP Staff Papers, Vol. 34, no3, pp.403-38. Blejer,Mario I. and Adrienne Cheasty. (1988) 'High Inflation, Heterodox Stabilizationand Fiscal Policy.' World Development,Vol. 16, No.8, August. pp.867-881. Bruno, M., G. Di Tella, R. Dornbusch and S. Fischer (1988). InflationStabilization,Cambridge:MIT Press. Calvo, Guillermo (1983). 'StaggeredContractsand Exchange Rate Policy,' in ExchangeRates and InternationalMacroeconomics,Jacob Frenkel editor, Universityof Chicago Press, pp. 235-55. Calvo, Guillermo (1987). 'On the Costs of TemporaryPolicies,' Journal of DevelopmentEconomics,Vol 27, pp. 245-62. Cardoso,Eliana and Albert Fishlow (1988). 'The Macroeconomicsof BrazilianExternal Debt,' mimeo. Cardoso,Eliana and Rudiger Dornbusch.(1987) 'Brazil'sTropical Plan," American EconomicReview Papers and Proceedings,May. Cukierman,Alex and Nissan Liviatan (1989). 'OptimalAccommodation by Strong Policy Makers Under IncompleteInformation,"The Foerder Institute for Economic Research,Working Paper No. 13-89, May. Cumby, Robert and Maurice Obstfeld (1983). 'CapitalMobility and the Scope for SterilizationsMexico in the 19709,' in Financial Policies in the World Capital Markets: The Problem of Latin America, P. Aspe, R. Dornbusch and M. Obstfeld editors,Chicago:Universityof Chicago Press, pp. 245-76. De Pablo, Juan Carlos (1972). PoliticaAnti-inflacionariaen la Argentina 1967-1970. Buenos Aires: Amorrortu Editores. De Pablo, Juan Carlos (1974). 'RelativePrices, Income Distribution

55 and StabilizationPlan.t The ArgentineExperience,1967-1970,"Journal od DevelopmentEconomics,Vol.1, pp.167-89. Dornbusch,Rudgier (1986). 'Money,InterestRates and Stabilization," Review of EconomicConditionsin Italy, Vol.3, pp. 439-53. Dornbusch,R. and H. Simonsen (1987), "InflationStabilizationwith Incomes Policy Support,A Review of the Experienceof Argentina,Brazil and Israel",Group of Thirty, New York. Dornbusch,Rudiger and SebastianEdwards (1989). 'EconomicCrisis and the Macroeconomicsof Populism in Latin America: Lessons from Chile and Perd," mimeo. El TrimestreEconomico, (1987) "PlanesAntinflacionariosRecientesen la America Latina" Mexico: Fondo de Cultura Economica. Estudios Economicos (1987). "ProgramasHeterodoxosde Estabilizacion," Mexico: El Colegio de Mexico. Finch, M.H.J. (1979). 'StabilizationPolicy in Uruguay since the 1950s," in Inflationand Stabilizationin Latin America, R. Thorp and L. WhiteheadEds., New York: Holmes & Meier Publishers,pp.144-180. Fischer,Stanley (1977). 'Long Term Contracts,Expectationsand the Optimal Money Supply,"Journal of PoliticalEconomy,Vol 85, No.1 (Feb), pp. 191-205. , (1983). "Indexing and Inflation,"Journal of MonetaryEconomics,Vol 12, ;ovember,pp. 519-42. , (1987). "TheIsraeli StabilizationProgram 198586," American EconomicReview Papers and Proceedings,May. Fishlow,Albert, (1973). 'Some Reflectionsin Post-1964Brazilian Economic Policy," in A. Stepan, ed. AuthoritarianBrazil, New Heaven: Yale UniversityPress. Foxley, Alejandro, (1980). "StabilizationPolicies and Stagflation: The Cases of Brazil and Chile," World Development,Vol 8, pp. 887-912. Foxley, Alejandro, (1983). Latin American Experimentsin NeoConservativeEconomics,Berkely:Universityof CaliforniaPress. Friedman,Milton, (1974). "HonetaryCorrection,"in H. Giersch et al., eds. Essays on Inflationand Indexation,WashingtonDC: The American EnterpriseInstitute. Helpman,Elhanan and LeonardoLeiderman (1988). "StabilizationIn High InflationCountriestAnalyticalFoundationsand Recent Experience," CarnegieRochester ConferenceSeries on Public Policy, No. 28, pp. 9-84.

56 Heyman, Daniel. (1987), *The Austral Plan", American EconomicReview, Vol.77, No 2 pp. 284-87.

Kafka, Alexandre. (1968) 'The Brazilian StabilizationProgram 196466,' Journal of PoliticalEconomy,Vol. No. pp.596-634. Kiguel, Miguel A. (1989). 'Inflationin Argentina:Stop and Go Since the Austral Plan,' PPR Working Paper, No. The World Bank. Kiguel, Miguel A. and Nissan Liviatan (1988). 'Inflationary Rigiditiesand Orthodox StabilizationPolicies:Lessons From Latin America," The World Bank EconomicReview, Vol. 2, No.3 September,pp.273298. Kiguel, Miguel A. and Nissan Liviatan (1989.a). 'Implicationsof Policy Games for High InflationEconomies,'Mimeo, The World Bank.

(1989.b) 'TheInflation StabilizationCycle After the Austral and Cruzado Plans,' mimeo, The World Bank. Krieger Vasena, Adalbert (1987). El ProgramaEconomicoArgentino 1967/69:Un Atague Global y Simultaneoa la Inflacion,Buenos Aires: AcademiaNacional de Ciencias Economicas. Liviatan,N.and S.Piterman(1986),"Accelerating Inflationand balance of PaymentsCrises: Israel 1973-1984',in The Israeli Economy, Yoram Ben-PorahEditor. Cambridge:Harvard UniversityPress, pp.32C-46. Liviatan, Nissan (1988), 'Israel'sStabilizationProgram:A Three year Perspective,' PPR Working Papers, The World Bank, No. 91, September. Malan, Pedro S. and Regis Bonelli (1977), 'The BrazilianEconomy in the Seventies:Old and New Developments,'World Development,Vol. 5 Nos. 1/2, pp.19-45. Meller, Patricio (1987). 'ApreciacionesGlobalesy Especificasen torno del Plan Cruzado,"El TrimestreEcon6mico,Vol 54. September,275-92. Modiano, Eduardo M. (1988). 'The Cruzado First Attempt: The BrazilianStabilizationProgram of February1986,' in Inflation Stabilization,Bruno et. al. editors,pp. 215-58. Pazos, Felipe (1969). Medidas para Detener la Inflaci6ncr6nica en America Latina, Mexico: CEMLA Persson, Torsten and Sweder Van Wijnbergen (1988). wSignalling,Wage Controlsand Monetary DisinflationPolicy,'Mimeo, The World Bank. Simonsen,Mario Henrique (1974), 'The Anti-InflationPolicy," in The

57

New BrazilianEconomic, Mario HenriqueSimonsenand Roberto de Oliveira Campos editors. Rio de Janeiro, Brazil: Crown Editores Internacionais. Solimano,Andr4s (1989). "Inflationand The Costs of Stabilization: Country Experiences,ConceptualIssues, and Policy Lessons,"The World Bank, PPR Working Paper No. 226, July. Taylor. John. (1979) "StaggeredWage Setting in a Macro Model," American Economic Review Papers and Proceedings,vol. 69, pp. 108-13 Viana, Luis (1988). 'The StabilizationPlan of 1968," mimeo, Banco Central del Uruguay.

Table I 8 ..

c0 Un

l

MIc,oeconoam.c Indicators

1962-1975

1062

1963

1Q64

1065

1Q66

1967

1Q68

1969

1970

1971

1Q72

1073

1974

1075

1. Inflation and Offic.-l D-ealuation (Percent in annual ter") a. CPI Inflat.on b. Official De.alustion c. CPI Inflati(Dec to Dec.) d. Official Decaluat,o. (Dec to Doc

51.1 43.6 55.6 35.9

70 8 48 3 80.2 28.3

Oi.4 117.8 86.6 98 2

65.9 51.6 45.5 19.5

41.3 16.9 41.1 0 0

30 4 20.0 24 5 18 6

22.0 27 5 24.0 32 8

22 7 20 0 24.2 12.8

22.3 12.7 20.0 13.1

20 2 15 1 18 1 13.0

16 5 12.2 14 0 9 8

12 7 3 2 13.7 0.1

27 6 10 e 33.8 17 9

29 19 31 20

2. Monetary Statistics a. Ml- Percont of CDP b. F2- Percent of CDP c. SeignlorageMonetary Bass d. Ml- Percentage Change a. F2- Percentage Change f. Monetary B.XJ- Percentage Change g. Monetary System no.atic Credit h. The Central Bank Oceestic Credit

17.5 18.5 4.8 70.0 67.0 61.7 61.3 62.5

16.1 16.8 5.2 64.7 63.6 71.1 60.5 63.4

15.0 15.6 5 2 82.1 82.8 83.7 76.7 78.9

17.5 18.0 5.1 76.5 75.5 68.6 49.4 27.0

16.0 17.2 2.3 16.7 22.6 26.5 27.0 22.9

16.1 17 8 2.0 41.9 45.6 25.5 58.2 37.4

16 3 18.3 3.2 43.0 44.6 45.7 55.6 55.2

16.2 18 4 2.3 28.6 32.5 29.2 34.7 26.6

16.3 18.6 1.5 2S.6 28.0 19.0 29.3 16.2

15.9 iB 4 2.2 21.0 10.6 30.9 39 3 39.8

15.7 18.1 I.S 40.5 44.4 21.7 45.3 50.0

15.1 16.6 3.1 47.5 44 6 51.5 50.5 55.6

14.0 15.1 2.1 34.5 33.0 33.5 50.7 67.9

13.6 14.7 1.7 44.4 44.8 29.6 57.3 70.2

3. Aggr"atoe Dend (Percent a. Private Consueption b. Public Consueption c. lnvestment d. Eaporta

prices) 60.0 12.1 20.3 4.0

56.9 12.6 16.9 8.9

58.6 11.9 19.0 6.5

57.1 10.9 20.0 7.7

58.2 11.2 19.9 6.8

62.0 11.6 17.0 S.9

60.8 11.1 18.9 6.0

56.3 10.8 22.4 6.6

68.6 11.3 10.5 7.0

69 6 11.0 21.1 6.4

69.7 10.8 21.2 7.3

67.5 10.4 23.2 8.3

70.7 9.7 25.4 8.0

66.5 10.6 26.8 7.5

9.7 3.2 7.7 40.0 19.4

8.9 15.2 6.7 -0.2 6.2

11.3 13.6 8.9 14.1 5.5

12.0 11.4 8.S 16.5 24.2

14.0 12.1 12.3 24.6 14.3

9.0 12.5 0.2 16.5 2.3

5.2 -0.9 15.2 10.1 11.6

100.0 100.0 100.0

103.2 102.2 98.0

107.4 112.6 100.3

112.6 125.9 101.1

-341.0 -1307.0 -2.7 1696.2

-244.0 -1489.0 -2.5 4132.7

7.0 -1688.0 -2.1 6359.9

of IP

at current

4. Economic Activity (Percent a. MP Growth b. Private Consumption c. Public Coneumption d. Investment a. Eaporte

Change in Real Tere) 5.2 6.6 2.0 4.6 -7.4

0.9 0.7 7.2 -9.5 28.0

3.5 3.3 -2.5 8.7 -13.5

3.1 0.7 -6.3 10.3 3.6

4.2 4.3 4.7 6.6 11.6

4.9 9.6 9.6 -S.9 -3.1

11.4 10.2 5.8 19.5 16.1

S. Relative Prices (1970B100) a. Rea Wage b. Real Exchange Rate c. T-ree of Trade

110.5 121.6 86.5

112.6 110.0 86.5

106.4 130.2 100.9

103.2 118.6 95.3

96.6 103.7 88.5

93.7 95.6 85.1

96.6 102.6 88.5

112.0 -114.0 -0.5 69.0

344.0 40.0 0.2 154.0

655.0 368.0 1.7 421.0

438.0 54.0 0.2 380.1

213.0 -237.0 -0.8 154.5

26.0 -508.0 -1.5 212.3

318.0 -281.0 -0.8 611.3

232.0 -562.0 -1.3 1141.7

11.2 9.5 1.7

I1.5 10.3 1.2

11.1 10.5 0.6

21.4 11.0 0.4

6. Eaternal Sector (Millionof US. Oollars) a. Trade lalence -89.0 b. Current Account Balance -369.0 c. Current Account- Percent of CDP -2.0 d. International Reserves 60.0 7. Public Sector (Percent a. Total Expenditure b. Total Revenue a. Operational DOficit

Notes: Sources:

Seignorage

98.09 104.5 95.3

112.6 126.0 74.5

-4690.0 -7122.0 -6.6 5215.6

n.o. 120.5 74.5

-3540 0 -6700.0 -5.4 3980.4

of CDP)

. M8(t)-l(t-1)|CDP(t),

12.9 8.6 4.3

where

13.0 8.8 4.2

MB is monetery

12 4 9.2 3 2

14.9 13 3 1.6

12 1 11.0 1.1

baa.

International Financial Statistics axcept: S(-) from L. Bachs end L. Teylor. eBrazilian Income Distribution in th IgQ60a. S(b) Cottani, eTrond in Real Exchange Rate Behaviour in Solected ODeeloping Countries 5(c) from the World Bank 7(a.b.c) from A. Leegruber, 'Inflation in Brazil', published in *World-ide Inflation: edited by L. Krause end V. Solent.

Theory and Recent Experience,

11 8 11 5 0.3

12.7 12.5 0.2

13 5 13.6 0 1

12.8 13.5 0.7

0 7 2 1

n.s n.*. n.s.

Table Uug..sy

1965-1973

1965

1966

1967

1968

1969

19O0

1971

1972

1973

56 4 63 9

73.6 112.9

89 2 58 2

125 3 130 1

20 3 6 2

17 0 0 0

24.0 2 0

76 5 110 2

97 0 61 5

Peto,y Stetistics *. 11- Percent of ODP *. l2- Percent of CDP Seignoreg.ltt h. Sei5noregemonetary

20 0 35.2 n.s. na

14.8 24.0 3 8 3.4

18 2 27 5 8.6 5.?

12 7 19.1 4.68 4.2

15 2 21 3 r 4 4

14 6 21 6 I4 16

It 7 27.3 5.8 4.6

l 25 S 5

0 5 S 3

13 9 21 1 S 6 S5

U.7 81.9 11.4

69.3 62.3 12.0

70 7 84 a 14 1

74.0 87.0 10 I

73 1 67 9 I II

74 5 89.9 11S

72.3 68.6 12.6

75.5 67 8 11 8

72 0 86 2 12 6

1.0

3.4

-4 1

1.6

4.7

-1.1

-35

3 3

106. 5 9'.1 114.7

116.0 116.4 114.7

112.1 97.4 103.6

100.0 100.0

1IS 7 71 4 104.0

959 99 4 130 6

94 3 104 2 184 5

S. Extornal Sector (Ifilliene af US. Dollars) a. Trred lblane b. Cwrernt Account blance c. Curront Account- Percent of CDP d. Internetional Reservee

7t.2 73.2 4.2 23.7

58.1 58.1 3.6 30.0

13 4 -4.2 -0.3 22.3

7. Poblic Sector (Perent S. Total E.penditure b. Total Rteven

15.3 13.2

14.2 14.4

2.1

-0.2

2

3.

Os"

Percent of MP a. Privets Conauwtion T b. otal Consuation c. inveatoent

4. Economic Activity a. CDP0rortb 5. Relative Prices (1966.100) a. Rtel Wge b. Rtel Exchange Ret. c. Ters of Trad.e

Un

2

c Ind-ct.o,

Inflation and Official Osealustion (Percent Change in Annual Teom) a. CPZ Inflation b. Official Devsluation

I

O8

Na*co.c--o.l

c. Primry

Deficit

1

6.1

111.5 90.0 7

110 0 80 6 O03.S

33.6

29.2 -19.1 -0.9 19.2

21.0 -45.1 -1.9 13.6

-6.2 -63.5 -2.2 20.1

102.0 56.t 2.5 69.0

79.0 37.2 1.3 100.6

14.5 12.2

13.4 12.9

13.4 12.2

13.0a 13.6

6.6 14.3

15.0 13.5

14 5 14.5

2.3

0.5

1 2

0.0

4.3

1.5

0.1

"97

43.4 23.4 1.S

of CDP)

Wot.~ I. Se;g"rsGott * tI (t)-WI(t-1)/t(t) 2. Signworags- tbnotsry bee -*9(t)-U(t-l)QDP(t) Sources: Internoti.'al Financial Statistic *scept: T 3(.b,c). 4(t) and 7(m.b.c) froo L. Vi na. ' he Stabiligation Plan of 1948 S(s) Finch, Stobiliation Policy in Uruguay since heh 1960*'. published in Itnflation edited by R. Thorp and L. hitehead. S(b) Cotteni;. Trends in Real fechange Rate Sahavior in Selected DO.eloping Countries-. 5(c) The World Bank

and Stsbilieation

in Latin America'.

Table 3 Argen Ln.

1064

1965

1. Inflation and Off.ical Oevaluat on (Percent in Annual Terms) a. CPI Inflation b Official Doealu&tion c. CPI Inflation (Dec -Dec.) d. Official Devaluation (Dec.-D.c.)

22.5 0.0 17.1 0.0

28.0 100.0 33.5 50.0

2. Monetary Statistics a. IlPercent of CDP b. M2- Percent of CDP c. SeignioreagMl d. Ml- Percent.. 0 Change . 42- Percentage Change f. Monetary Base 0. Coneolidated Monetary Syat-'m Do_oetic b. The Central ank's Domestic Credit

12.8 17.8 n.e. na. na. 40.7 30.9 36.4

3. Aegrete De nd (Percent 8. Private Consumption b. Public Conaumption e Investment C. Exports 4. Ecoemic Activity (Percent a. CDP Crowth b. Unemployment c. Private Conaumption d. Public Consumption *. Inveetmwnt f. Exports

1. 2. 3. 4.

Sources:

Seigniorage Unem loyment The domestic Th- Central

1966

1Q67

1Q68

31 0 27 0

9 0 8 0

29 4 50 0 25.2 33 3

16.1 0.0 9 6 0.0

12.0 16.9 2.8 29.8 31.1 39.3 26.7 36.7

12.3 17.1 3 9 28.2 26 6 27 2 28.4 12.2

12 2 16 8 3.3 32 3 31 3 41 4 24.6 15.2

72.s 7.4 19.7 10.6

72.3 6.9 19.4 10.7

72.0 7.3 17 9 11 7

10.3 6.5 II1S -1.1 26.0 -6.4

9.2 1.4 6.9 1.8 9.6

0.6 6.0 0.3 6.6 -7.2 9.6

83.3 65.6 105.5

92.1 106.0 111.2

95.6 83.9 106.8

1064-1973

1Q70

1971

1Q72

1Q73

6 0 0 0

13.5 33.3 21 1 25.0

34.9 2S.0 34 2 20.0

58.4 0.0 49 3 0.0

61 0 34 0

13.8 19.3 3.4 30.0 31.6 13.0 28.9 26.4

14.2 20 3 1S 20 1 24.0 0 9 1338 9.0

14 8 21 6 2.8 13.3 15.5 19.7 18.8 8.2

13.0 19.5 4.3 25.2 28.1 28.4 41.9 48.1

11.0 16.9 4.2 39.4 43 8 43.0 5903 57.3

11 3 17 7 7 7 76 9 78 4 143 1 94 7 56.0

72.0 7.2 16.2 11.2

71.8 7.1 19.3 10.6

70.3 6.8 21.6 13.4

69.5 6.5 22.0 11.6

69.9 6.5 23.5 10.1

69.6 6.3 23.1 10.0

69 5 6 5 21 4 10.8

2.6 6.5 2.6 1.9 4.5 -1.2

4.3 5.5 4.0 2.9 10.6 -1.4

8.5 4.7 6.4 2.7 21.4 16.1

5.4 5.3 4.1 0.6 7.4 7.2

3.7 6.2 4.4 5.0 10.8 -9.7

1.9 6.9 1.5 -1.5 0.4 1.5

3 4 5.8 3.3 5 7 -4.5 11.8

92.1 64.5 98.6

96.4 81 0 101.3

100.0 100 0 100 0

103.4 100 4 109 0

217.0 1612.0 1395.0 -230.0 -0 9 402.8

1969

7 0 6 0

2 0 2 0

Change in Real Terms)

6. External Sector (Million of U15) a. Trade Belnce (i) Exporte (ii) Import b. Current Account Balance c. Current Account- Percent of CDP 6. International Reservee

Note;

Indicatora.

of CDP at 1960 prices)

5. Relative Pr;es> (1970m100) a. Real Wege b. Reel Eachenge Rate c. Term of Trade

7. Public Sector (Percent a. Total Ezpenditure b. Total Revenue c. Operational Deficit

Credit

Macroeconoe.c

457.0 1410.0 953.0 457.0 1.6 82.0

7.3

95.0 98.6 101.3

431.0 1493.0 1062.0 222.0 1.2 170.0

596.0 1593.0 995 0 259.0 1.2 132.0

494.0 1464.0 970.0 130.0 0.6 643.0

333.0 1368.0 1035.0 -53.0 -0.2 650.6

32.7 29.8 2.9

35.3 31 6 3.7

37 9 36.0 1.8

37.1 35.4 1.7

274.0 1773.0 1499.0 -163.0 -0.7 532.6

87.3 1740.0 1652.9 -390.2 -1.6 192.4

93.8 70 6 107 2

106 1 48.8 121.5

2S6.0 1941.0 1685.0 -226.9 -O.S 313.3

1268 3 3266 0 1977 7 710.5 I 0 1148 7

36 0 31 9 6.1

41 5 32.9 6.6

of CDP) 35.6 30.2 5.6

* M1(t)-Ml(t-l)/00P(t) is the rate of unemployment in urban area. credit of conaolidated monetary systae is line 32 (IFS). ankh' domestic crelit is total claims on the goernment and the private sector

36.1 34.9 1.3

lea

International Financial Statistic eacept: 2(a.b.d.e) from FIeL 4(b) C. Sanchez and 0. Ciiordano. 'Exchange Rate Policiea and the Structure of the Labor lMarket(3). 4(c.e4.9f). and S(a) from Eatudio-. Fundacion Moditarranes 5(b) from Cottani. OTrends in Real Eachange Rate Behavior in Selected Developing Countries. S(c) from the World Bank 2(f) and 7(a,b.c) from Oe Pablo. Nacroeconomic Policy Cries and Long-Term Crowth,

govornment

39.5 37.5 2.0

deposit.

38 7 34 1 4.6

Table Argentina:

1Q80

4 Indicators.

Hacroeconom;c

1981

1982

1983

1980-1988

1984

1985

19t8

1987

1988

1. Inflation. Doealuatiron and Int.eoat (Percent in annual terms) a. CPI Inflation b. Official Devaluation c. Interest Rate on Loans d. CPI Inflation (Dec. to Dec.) e. Offic.oi 0evaluation (Dec. to DOc.)

100.8 39.5 98.6 87.6 80.5

104 5 139.7 207.1 131.3 296.2

164 8 488.8 213.8 209.7 484.8

343.8 306.2 682.4 433.7 364.6

626 7 542 4 748.0 688.0 652.3

672.1 78906 512.3 385.4 398.0

90.1 56.7 117 8 81.9 51.4

131.3 227.6 227.8 174.8 190.4

343.0 308.0 407.8 387.7 271.3

2. Monotary Statistics (Percentage Chango) a. Ml- Pere nt of CDP b. N4- Percent of CDP c. SeigniorageMI d. Rate of Change in I *. Rate of Change in It4 f. ConsolidateAd Monetary Syate Dometic g. The Central Sank Doestic Credit

8.2 24.7 4.8 119.5 123.3 111.3 371.8

4.9 25.4 3.5 51.6 95.3 209.3 268.4

4.7 20.7 7.8 165.9 124.7 224.0 557.8

3.8 13.5 8.6 281.4 216.0 400.8 400.8

3.4 13.2 7.1 550.6 603.2 538.0 501.2

3.6 12.7 6.5 683.7 603.5 n.e. n.a.

4.9 15.4 3.4 158.3 132.0 n.e. n.e.

5.2 17.2 4.0 107.1 131.3 n.a. n.a.

3.8 15.4 7.1 327.8 452.3 n.s. n.s.

63.2 23.7 11.4

85.4 19.4 12.9

80.3 16.4 14.1

81.0 14.3 14.7

83.7 12.4 14.3

82.1 10.3 16.8

83.8 11.6 14.8

83.1 13.2 14.4

80.6 12.5 16.6

2.4 n.e. 7.7 9.2 -5.1

-6.8 4.8 -4.3 -23.4 5.6

-4.6 5.3 -10.7 -19.7 3.5

2.8 4.7 3.8 -10.7 7.8

2.6 4.6 5.9 -11.1 -0.7

-4.5 6.1 -6.2 -20.0 12.5

5.7 5.2 7.7 19.2 -7.0

4.1 5.9 2.0 16.1 -6.4

-5.2 6.3 -6.8 -8.2 12.1

100.0 100.0 100.0

112.6 110.5 106.7

115.9 217.6 89.1

175.4 219.8 86.6

180.7 187.8 93.4

140.0 216.1 81.2

113.0 197.4 69.0

143.1 211.2 65.2

136.2 450.1 na

-2519.2 8021.4 10540.6 -4768.0 -3.1 -2493.0

-267.0 9143.0 9430.0 -4714.0 -3.8 -3559.0

2285.9 7623.1 5337.2 -2358.0 -4.1 -6186.O

3522.7 8107.4 4584.7 -2391.0 -3.2 -1744.0

4551.9 8396.1 3844.2 -953.0 -1.5 -556.0

2127.9 6851.? 4723.8 -2859.0 -3.6 -2075.0

556.0 6356.0 5800.0 -4238.0 -2.5 -4100.0

3629.0 8952.0 5323.0 -1615.0 -2.2 -P381.0

26.0 18.0 -8.0 8.6 -16.6

25.4 21.0 -4.4 4.7 -9.1

23.5 21.5 -2.0 3.2 -5.2

24.9 20.6 -4.3 5.0 -9.3

23.5 18.6 -4.9 5.0 -9.9

government

deposit.

Cre.4t

3. A4regate D_nand (Percent of CDP at constant a. Total Conaumption b. Inveetment e. Eaporte 4. Economic Activity (Percent a. CDP Growth b. Un_eloym_nt c. Total Conu_ption d. Inw t_mnt *. Esports

Change in Real Terso,

5. Relative PriceO (1960.100) a. Real Wa 4 b. Rel Exchange Rat .0 c. Torso of Trade 6. External Sector (Hilliona of US. Dollars) a. Trade Balance (i) Exports (;;) *_orte 6. Current Account Balance c. Current Account- Percent of GOP d. International ReAerves Change ?

Public Sector (on Cash Basis, S. Total Eapnditure b. Total Revenue c. Operational Deficit d. The Central Bank Deficit a. Total Deficit

Note:

prices)

Prcent

of CDP) na na na no no

na na na no na

1. Seigniorago. F l(t)-l(t-1)/00(t) 2. Real exchange rate in 1986-87 is obtained by interpolation. S. Coneolidated monetary systm dometic credit i line 32 in IFS. Th. Central ak'e Domestic Credit is total clai; on gornment 4. Operational Deficit a Revenue - Eaponditure Total Deficit a Operational Deficit * Central Sank Deficit.

Sources:

3331.8 7838.0 4504.1 -2461.0 -3.6 -2572.0

na no no na na

29.3 20.5 -8.8 -11.0 -19.6

and prioate

aector

1(wb). 2(f.g). 3(c.d). end 6(d) from IFS 2(a.b.d4.). 4(a.b). and 5(a) from De Pablo. e"Mcroeonomic Policy Cris 1(c). (3). 4(c.d.o). 5(c). 6(*-d) end 7(a-d) from The Central Sank of Argentina 4(b) from Fundacion Nditerrones 5(b) from Hyde (1989). OR&*l E.change Rate Trends. 1960-1987'.

l-ea

Table 5 89az,1

1980 1. Inflation. Devaluet,on. and Intrest (Percent in annual terms) a. CPI Inflation b. Official Oe.slustion C. Interest Rat. on .pos.t. d Inflation (Doc. to Dec.) .. Official Devluation (Dec. to Doc ) 2. Monetary Statistics a. 141- Percent of CDP b. M4- Percent of CDP c. Seigniorege- Monetary B... d. MH- Rate of Change *. M. Rate of Change f. Monetary Base- Rat, of Change g. Consolidated Monetary System's Dometic f. lhe Central Bankas Domestic Credit 3. Agregte Dmnd (Percent a. Private Consumption b. Public Coensmption c. Investment d. Esporta 4. Ecommic Activity (P*ercqt a. COP Growth b. Ul-employmeet c. Toutl Coumption d. [sweetmet o. Exporte

C D

Change in real

6. Exteral Sector (llo 1 ;rn of U38) . Trad Balance (it Ezporte (ii) Imports b. Current Account aelance c. Current Account- Percent of CDP d. International Resfeves 7. Public Sector (Percent of CDP) a. Operational Oeficit b. Dometic Public Debt c. Foreign Public Debt

Sources:

1981

1982

1980-1988

1983

1984

1985

1986

1987

1988

82.8 05.6 n.&. 86.3 43.6

1OS 6 t6.7 58.4 100.6 66.9

97 92 65 101 69

8 8 9 8 6

142.1 221.4 82.2 177.9 126.7

197.0 220.3 96.9 208.7 119.0

226 9 235.5 108.4 248 6 117.8

145.2 120 2 40.0 63 5 32.3

224 8 187.9 115.9 432 3 385.1

684 6 568 7 n a 1037 6 889 4

10.8 29.3 2.0 70.2 72.3 56.9 80.7 78.6

10.3 35.6 2.0 87.2 120.1 62.8 95A. 53.9

9.7 38.6 1 6 81.7 100.9 63.6 108 6 97.7

7.8 39.3 1.3 97.4 170.2 79.8 179.4 177.0

7.1 47.0 2.4 201.8 250.1 264.1 178.6 69.1

8.0 52.0 2.3 304.3 269.8 257.3 301.9 417.4

11.9 37.5 3.5 306.7 124.9 293.4 n.a. n.

8.1 S0 8 2.5 127.5 276.9 181.5 n.a. n..

7 3 76.8 3.3 571.7 1045.7 622.3 n.a. n a.

70.8 9.0 22.4 8.9

69.0 9.2 27.3 9.3

70.0 10.5 20.2 8.0

72.3 9.6 IS.4 11.3

69.7 8.2 16.5 13.5

67.2 9.7 18.0 12.0

66.2 10.2 20.5 8.8

65.1 12.2 19.5 9.8

n.a. n.e. 17.0 11.2

9.1 7.2 5.9 6.5 22.6

-3.3 7.2 -4.3 -13.3 21.3

0.9 5.5 3.7 -5.4 -9.2

-2.5 6.8 -1.6 -17.0 14.3

5.7 6.8 1.0 4.1 22.1

8.3 5.0 7.1 11.6 -2.0

8.2 3.3 11.6 12.0 -23.6

3.6 3.8 3.4 -3.1 14.7

-0.3 3.9 3.8 -11.2 13.0

100.0 100.0 100.0

115.0 87.2 88.5

122.9 81.7 85.5

116.4 102.4 79.0

118.1 114.3 85.1

135.0 117.4 83.5

154.6 107.1 105.6

143.4 95.8 96.9

n.s. 112.2 n.s.

-2823.0 20132.0 22955.0 -12807.0 -5.3 5769.3

1202.0 23293.0 22091.0 -11734.0 -4.4 6603.5

817.0 20213.0 19396.0 -16311.0 -6.1 3927.9

6472.0 21900.0 15428.0 -6837.0 -3.3 4355.1

13068.0 27005.0 13937.0 45.0 0.0 11507.9

13450.0 25639.0 12189.0 -241.5 -0.1 10604.6

8304.0 22349.0 14045.0 -4476.0 -1 7 5803.0

11184.0 26225.0 15061.0 -1458.0 -0.5 6299.0

19070.0 33782.0 14712.0 n.e. n.e. n.e.

3.6 13.1 18.7

5.6 17,4 10.0

8 3 25.2 23.3

4.8 20.2 30.0

2.7 19.5 31.6

3 6 21 1 30 4

5.5 32.6 24.8

4.3 n.e. n.a

ters)

1. Seigniorase * M8(t)-8(t-1)/00P(t). heer He is monetary basa. 2. Interest rate on deposit is discount rate on 91-day notse. 1(-.b.c). 5(b) from 3. 4(c.d.-). 8(c) from 7(a) free

tldicato.s.

of CDP)

5. Relative Prices (1S100) * IAbl WaW at b. Real lEhnge Rate c. Terso of Trade

Notes:

Credit

Mscroeo.o.,c

2(a.hbc), 4(ae.). S(a) 6(a.b.c) from Conjuntura Economics flde (1989). Real Echange Rate Trends. 1960-88. S(c). 7(b.c) from The Central Sank of Brazil International Financial Statistic DW R"ort

4.3 21.1 29 3

Table 1,.s.

1nflt.o.. DOe.aluation and Intorest (Percent .n A.n. a[ T*re. * CPI Inflation b Official Ds.sluation (A,*,*$*. Shekels/$) c Real 1t-aret Rat on Loans d Real Intrest Rat on Doeos-t * CPI Inflation (Dec. to Dec ) f. Official De.vluation (DOc to Doc Sha &ls/I) *. Official Devaluat,o0 (Dec to DOc aSSketof C.urrni--) 2. a. b. C. d ..

4etaryStatistics Mt- Percent of GDP 12- Percent of GOP le- Percent of (lP Segnioreg- Percent ml- Percent Chmne

of CDP

3. Agaregete Depand (Percent of GDP) a. Pri.ste Cona-_tion b. Public Consumtion C. In.eeteent 4. Ipot Surplu (e.cludeaefence

I c1 '.0

imorts)

4. Eco ;ic Act;ity (Porcent Chatng n n Real Ters>) a. GDCroeth b. Private Coneuegti.o C. P"li;c Cou_tion 4. Investment a. E. orte f. Total Dometic Us 8- Reel D ;isoebl; Ineome (includes transfors froo abroad) %. Total Productivity (1usineas sector) i. Laployant (Percent) S. Ralast;i- Pric Indices (1980-100) a. 1Rel Wage (i) Privatend Piblic Sector (ii) Indust-ial wage Per Unit of Output in tr_a b. Real E.chane Rote (;) Industrial Prices (;;) Eoort Prie("ii) imort Price C. Terms af Trde S. Eaterotl Sector (mi41 ion U53) a. Trade Salanc (i) Eaporte (;;) Imports b. Current ccount alance C. Current Account- Porcent of MP 4. Intornational Racros7

Public Sector (Percent of (DP) a. Total Eaenditre b. Total Peenur C. Oerational Deficit d. Dometic Public Debt e. Foreign Public Debt

not":

of Eporte

is

tim

deposits

Ca

6 1n. .tato's

1980-1988

tq80

1a81

1182

1083

1984

1S85

600 66 7 '2 -237 133 0 113 5 n ..

116.8 120 0 16 9 -6 2 101 5 108 0 n

120 4 118 2 3.6 -24 0 131 5

145.6 133 3 6 2 32 * 190.7 219 9 206.0

3738 423 2 289 -67 444 9 492 5 462 0

49 6S 212 21 442

3 8 5 9 19 2 2.0 93 2

3.4 6.8 19.6 1.7 100.4

2.9 6.6 24.4 2 1 113.5

51 1 302 21t 32

53 0 30.2 20.4 3 5

54.5 30.0 22.2 6 7

3.1 -1 3 3.5 -15 2 6 3 -4 5 1 7 -1.9 46

4.3 12.6 2.0 -5.2 4 3 6.5 14.2 3 I

100 0 100 0

1986

198?

1Q88

3046 302 4 16 4 -7 3 185 2 134 8 154 0

48 26 8 -2 19

199 7 2

6 0

1 3 6 1 3 5 13.9

16 3 3 0 13 8 -4 0 16 4 9.5 0 8

2 3 6 2 268 3 2 2251

2 3 9 5 22 2 6 9 3515

3.7 14 2 23 4 1 a 1694

5.I 17 8 23 7 2 4 60.3

5 4 17.8 23.3 -1 5 32.3

55 7 29 4 22.2 7.3

521 300 209 3 0

55.4 27.0 17 7 0.1

S8 4 24 9 16 5 1

59 7 25 0 18 4 3 0

59 3 25 6 17.2 2.0

1.0 7.7 2.9 14 6 -3.2 7 2 -2 6 -1 S.0

26 8.3 2.2 10.0 2.1 7 0 2 8 -0.2 .5

23 -68 1 3 -71 135 -47 8 7 -0 3 59

39 0.5 -0 6 -106 7 6 -2 1 -10 7 4 2 6 7

36 14 2 -2 9 10 4 4 9 8 7 3 4 3 4 7 1

5.2 6 4 2.6 3 3 t0 7 6 1 7 1 3.5 6 1

16 3.0 2.8 -2.1 -2.3 2 1 5.4 -0 I 6.4

109 6 109.5

111 0 121 1

109 9 133.4

1163 2211

101 1 108.6

112 9 132.3

123.2 138.7

127 8 152 0

100 0 100 0 100 0 100.0

97 7 97.6 96.3 102.0

93 1 94 1 90 0 107.0

07 6 00.4 84.0 111.0

91 91 86 107

104 4 91 5 93.1 107 0

97 5 60 4 81 3 110.0

94 1 75.9 78.9 107.0

84 8 74 2 73 7 112.0

-2324 0 5215.2 7539.2 -907.0 -4.3 3526.0

-2204.1 S23 .5 7522 6 -636 0 -4.0 3614.0

-2742 * 4907. 7649 6 -2003 0 -8.3 4317.0

-3271 1 4822 9 8094 0 -2145 0 -7.9 a780 4

-2204 566 7770 -1620 -6 3255

-2030 9 6875.8 8906 7 906 0 2 9 4867 5

-2903.1 8137 3 11040 4 -235.9 -0.8 5961 7

-2497.2 9361 1 11860.3 -105.0 -0.5 4755.7

76 3 60.1 16.1 120 8 36 4

74.5 59.0 15.S 120.9 37.3

72 5 59 2 13 3 123.6 37.5

67 4 59 3 8a 118.0 40.8

65.5 67 4 -1 121 1 39.8

62 7 61 4 I 3 2 113 5 31.4

62 6 58.2 4 6 110 7 29.0

1. Se;s-ior,ee =- M t)-M8(t-1)/1lP(t). eher. Me ia bas of money 2. Real intereot rate is nominal interest rate deflated by CPI inflation l(c) is coerage short-trm bank loano. 1(d)

M.tr.c-om.,on-

(Dec

116.0 a.

5 1 0 0

4

22 6 0 8 2

73 3 545 s89 123.6 49 3

-1047 2 600 1 7649 3 10946.0 4 8 3793 6 71 2 68 6 2 6 133.2 51.0

1 2 3 0 6

-0 0

to Dec.)

etc.

3. Sb(i) ia b*d on th ratio of induatrial ,rice of troding partners to thosa of Israel aItiDl,.s by official oechangm rate Sb(ii) i based on th ret: of e.port rices to pricee of doestic usee aultiDl i.e by offic.al e-chone rlate. Sb(iii) is boed on the tio of import prices to *ricos of doestic u-e- aultupl.. by off,c,al e.chanae rate 4 7(c) dometic plus fore,gn S. 19'0-84 and 185-68 are based on simple acerages. Source:

Boni. of Israel

IS9

Tab le 7 M.---acaeo~o.

1n,agoe

1980-1986

I1960

1981

1982

196.3

3964

1985

CPI Inflation Official D~a.eaation Into,.at Steo0. loans CPU WInfticri(Dec. to Dec )29 Official Deesl..at.on(l to Dec)

26 3 0 9 20 7 6 2 2

28 0 65 2886 26 7 12 6

so 9 130 2 40 4 98 9 268 3

101 9 112 9 568 80a6 49 1

65 5 39 7 531 59 2 33 8

57 53 52 63 93

2. 101nstary Statistics (Perce,,t) *. RI- Percent of CDP b. MG- Pe,cnt of CP C. Seig.;,6oragM 4 li- Rate of Cha.0sa a 140- Rate of Cha.n* f Monetaryas"e * Consolidated Htintary Svst.aDoseel.,c k.. Tha Cent.ral Sa.,t Dom.st,c C,ed.t

ti 2 30 1 27 3364 45 6 41 1 39 0 37 0

10 a 33 6 2 7 33 3 53 7 45 1 488 43 S

t0 9 37 4 42 54 1 78 6 904 00oo9 157 6

6 4 332 24 424 61l4 so1 564 41 7

8a1 337 30 623 70 4 543 532 31 1

65s.1 10 0 24 8 10 7

64 4 10 7 26 4 10 4

61 6 toS 22 9 15 3

609 68a 375 19

831 92 i60 I7?4

83 4S 75 95 14.9 61

79 4 2 74A 10.3 16 2 11 6

-056 431 -2 5 20 -16 S 21 8

-5 3 6.7 -5 4 7.7 -2831 13.6

S. b. c d *.

3. Aftregat.i Demand (Porcent a P'i.ats Cona..ption b. Pa&IicCCnsumption C. 10.0at.en 4. Eaport.. 4. Ecenemic Acti.ity (P-rcnt a. MP rowth b. Unmeloymmot C. Pri.ste Cona..wtia d. PabIic Consoption o. 1-offitmont I . E.Port.a

Cradt

of CDP at. c."ra,t,

1986

1987

1986

88 2 137.9 80so9 205 7 14685

131 8 123 5 86 7 359 2 139 3

114 2 64 7 676 51 7 21 S

7 6 3122 2 5 53 8 46 7 17 5 69 2 49 0

7 3 35 9 2 9 72 1 100.2 47 7 1020 63 3

6 5 3789 3 5 1297 159 4 70 3 1042 13 7

5 6 31 21 soI 69 0 423 na

6. 5 92 19 2 Is 4

6862 91 19 4 17 2

65 6 87 18 9 19 7

268 48a 3 3 09 82 -4 0

20 4 3 -2 8 2.1 -12.0 15

14 3.9 -3 4 -3 0 -0 7 2222

68 8 98 5 96 3

63 0 3451 63 8

59 1 157S aa

330 3

7 2 7 0

n

Pr,c.s) na na

20 4 a

Chsnoe in Reel Terde) 37 60 33 68 64 5.7

II1 3 6

n

a

~*

S. Relati.e Prices Indices (1960-100) '..C0 a RealeWave b. Rel Echs.ge Rate c. Tomis of Trade 6. Eaternel Sector (04.0iion of RS3) a T,edo Sala-co (i) E.porte (;i) Iworte lo. Cujrrent Account Batlance c. CurentAccountP*rc,,t of CP 4. Internation,al Roe"",. Chen6 . 7 Public Sector (Percent a. Total E.apnditur6. E..penditure s.clud-g

3018 64 2 1065

93 5 3152 103 2

747 3237 965

-230W 0 16066 0 18896 0 -8362 0 -4 # L151 0

-4099 0 19936 0 24037 0 -13899 0 -56 1012 0

6795 0 2'12300 14435 0 6218 0 -3 7 3188 0

13762 0 22312 0 6550 0 5419 0 38 -3117 0

12941 0 24196 0 13255 0 4238 ! 25 -3384 0

8451 0 21563 0 33212 0 1236 7 07 2423 0

4599 0 160330 11432 0 .1672 7 -1 3 -132 0

8433 3 20656 2 12222 9 3966 5 27 6929 4

1754 3 20657 6 18903 4 -2901 2 -16 -7127 0

33 5 30 0

39 7 34 7

44 5 36 3

41 0 28 6

39 3 27 4

30 2 27 7

44 8 26 3

45 0 24 7

39 0 22 4

26 9 6 6 5 6 3 1 8 6 168a

2671 13 0 13 7 8 0 10 7 20 0

26 9 15 6 83 74 32 6 38 9

329 63 3 -43 16 5 430

322 7I 3 06 -48 18s0 388

3122 a60 3 -3 5 36 6 37 0

30 3 14 5 5 7 -2 0 16.2 Ss39

30 6 14 4 03 -5 9 20 0 48 7

2'98 92

of CD)

C. TotalRe.e.en 4 TotloDefic.... e. Operational Deficit f. P'iee'y Deficit 6. Doeptic Public Debt *. Foreign Pu.blicDebt NDt,e

736 1025 972

1000 t000 1000

Inleret

I 2. 3. 4.

S.igiorege * M31(t)-r(t-1)/00P(t) *eal .ae0 S the p.,Ch.s-n Poesr in relat-on to the. oveal onaae price .. da. Interest rtet on Ioa-. Is the 0c0re00 cost of (.ndo c.oepoud an-ually Consolidated eonsetrY $yst.. do"set,. credit a9 In-. 32 *n IFS The Central Sen1 domestic cradit a total ee.-, the u.,--eet ndpvt.ecola.oeren S. P,iaery Deficit = 7(b)-7(c)

Sources: 9enco de "e.;co *.Cetit 4(b) CEPAL 4(c.d.e.f) from Th. World Bsen. Netio,nal Accout Data Ba."7fg 5(a) "*.Cio Nat.onal mi.-v Sala,, Coais,a.on 61(a) IFS 7(*) The World Sn

depoe't

5(c) the World Sank 2(..f) end 6(d) IFS Inforee A~u* 1987. Banco.de K-co

n

-4 n

Figure 1-A CPI Inflationand OfficialDevaluation

AIIGR1 INA 5w1W

NN

BRAZIL

?bMd

50

.

40

IWn

5. U|......

.

N... TtbMStn.7

40

30

30

20

zO

*_ut

>

.

. so

1987

'

'

I~~~~~~~~~~~ * c

5'

£94

OS

7WI

16

IS9£

ISS

USEJULI Can?& 109A OOWIUSC*COWJPOTi,Uw~A

5

0m

,*,

S~~~~ __ 19S51

35

*

se

16

mAWImUI'

Sr^"lu

S"

cLAIunn"ozr

~~

5

5

.

,

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~20

97

1

.

1918

@

30 25

I

20 S~70

is01

to~~~~~~~~~~~~~~~~~~~~~~~~I OffVALUAYION

-----

---- 5

--

ome

.~

ProuncT

- IF*

*'...S

S-i\

s9S

7

5

ISS

159

087* an~~~~~~~~~~~~~~~~~~~uSiSW iSAWn,1w*6MIIPC

368

96

14 0

0

* -

IA 0

.

0 0

N 0

*

:..

-A

0

a C

.. --

0

-

0

A

*

a

.4

U

S -

a

-

-, 3

4

a g a IA

a

----

.4

2

U

a

-

U

-

.4

a

'-a ,.a.

£

It

.4

-

a

*---.

3 -'

3

a M

9.

3

I.-

I 4 5

a a

p

--

0

3

0

*

U S

.4

5

-

0

* C

U

U

0

-

a

a

0

a

-

-

--

-

z

C U

z -

I

3

.9

---

-A

I

at

.4

N

0 p

- 67 -

Figure 2 Fiscal Deficit, CPI Inflationand OfficialDevaluation

RRAZII. 5

.

9

8

_ 50

low8

At 1982

foal

19831 WEI

71~

O.

X 1904 i I'

loss

~~~~~~~~~~so

lo

100

.,~

as

......

12.5 Jol"tRcus.

low*

flo

IS0 Fl

*NI IW Itmsom

100

[a25

1900

1981

1982

196:l

1984

t15

1to1s

19we

Figure Real Interest

3

Rates on Loans and Deposits

40

211

1~Ie. LUi*Nu-01 4 re*

SO

10~~~~~~~~~~~~~~~~~~~ .10~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~, .:

I

e"tet

m-oannt

Lal

~~~~~o

t"

^to@ e ""zrt cog@es rgt-

t"1*^. l|."

1

"ti tv OMA

OWAL maIWOI RAIa

10~~~~~~~~~~~~~~~~~~~~~~I

#rn0;

I

ISRAEL :b;;I;L

-:3

to nuu WYn

I* 19101111

1 53

som

[email protected]" 61guZ

AtUg

S

N *MINAle. lHtY _ 1' - r1-6 AT *N 4?41)@tt

10i0m

---

am. ^ SNW.SIlt

no

tE~~~~~~RAL DEPOIT RATE

to

I

590!. 33*3 IlmiitUkv

1901

le&te a 011014,16%.al4ti

19197 Iwo$ huNc- 14* )

lOS

*m: 11P.

-1

Tn

mUIco

-" -3 1*71 a NuNl

_

03 I1lo|t

_ ,rn.7,.su

*i4a6

OWNUIll

bUOSII

N:

III 14lplI "Al' 1 a INUIVW*l.1101*II3..4q- hnIIA FOIIi 13

RATE

- 69

-

Figure4-B FiscalDeficit,CPI Inflationand OfficialDevaluation AIlM:INlt NA 7

_

;

6

S~~~~~~~~~~~~~~'

6"if

4

*"IIWSM

t41i

1W ISEt

67

tl

69

70

A.tw *

0

71

s

'

tttUnst

6RAZT.I U?WIfc? I4

CarGIWI

it to A"NNal. ?8IIUN

i

5~~~~~~~~~~~~~~~25

' 1i>hI (lAYtON :5

too 75

a

,,~

01

G:

6:4

SCIII(lt

64

65

66

W1

.~~~~~~~~~~~~ C* ourl IANEItIll A

66

Iil.

69

I

70

71

7;:

7a

lA 141111?

125

/elt

^1z-llzs ' *I_ /I*tt

tl_00

3tlitttPt.*l

.1:'e1".1

/

0 el tEIL-itIW

74

1.

NOe (17 SiLNA41%Ml

us 1ISA itATttl.

88 AtIIIIA

70

71 CIP S

7.2

7:1

Figure

S-A

GDP Growth and Unemployment

~~~~~~~~~~~~~~~~~~BRAZII.

AR{;|NTI"^

*S@T El

IUbOIIISCtUNFWUttfU(

IINSOYUUN4¶

UktI?

-1@

@bts

h5

*3A211.~~~~~~~~~~~~~~NI.NNI

AEE;INrINA

4

2

,l~~~~~~~~~~~~~~~0

0

l0 I_~~~~~~~t O

l"14z

~

~~

I9U

go"

S Sf""1-

'

"|@"tel"~~~~~~~~~*

Sb

-.

7.5

1


8'

'

1 1;' SOuC co &m t

87 1 IIItCC

4

I

..

3

01 4

- 75

-

Figure Real Rates

7-B

of Growth of GDP, Private A R(i ¶% N

Conaumption

and Investment

INtA

TQMWIi

20 i\\

IN1MUNt

10

l0

1,1 ^-.s"ne

(is

66

*I.N,Art'lt-

67

(A

70

an

71

72

Ii,lIIIANtFtA^

BRAZIL

40

!

A,.4.t