Journal of Development Economics 32 (1990) 247-277.

North-Holland

MACROECONOMIC POPULISM*

Rudiger DORNBUSCH
Massachusetts Institute of Technology, Cambridge, MA 02139, USA
National Bureau of Economic Research, Cambridge MA 02138, USA

Sebastian EDWARDS
University of California, Los Angeles, CA 90024, USA
National Bureau qf EconomicResearch, Cambridge, MA 02138. USA

Macroeconomic populism is an approach to economics that emphasizes growth and income

distribution and deemphasizes the risks of inflation and deficit finance, external constraints and
the reaction of economic agents to aggressive non-market policies. The purpose of our paper is
to show that policy experiences in different countries and periods share common features, from
the initial conditions, the motivation for policies, the argument that the country’s conditions are
different, to the ultimate collapse. Our purpose in setting out these experiences, those of Chile
under Allende and of Peru under Garcia, is not a righteous assertion of conservative economics,
but rather a warning that populist policies do ultimately fail; and when they fail it is always at a
frightening cost to the verv erouDs who were SUDDOSed to be favored. Our central thesis is that
the macroeconomics of various experiences is very much the same, even if the politics ditTered
greatly.

1. Introduction

The purpose of the study is to develop a paradigm of populist macro-

economics.’ Macroeconomic populism is a policy perspective on economic
management that emphasizes economic growth and income redistribution
and deemphasizes the risks of inflation and deficit finance, external con-
straints and the reaction of economic agents to aggressive non-market
policies.
We analyze two populist economic programs, Chile during Allende’s
Unidad Popular (197G-1973) and Peru under Alan Garcia. We show that
policy experiences in different countries and periods share common economic

*This is a much shortened and revised version of a paper presented at the second meeting of
the InterAmerican Seminar on Economics, IASE, Bogota, Colombia, March 30-April 1, 1989.
The authors are indebted to conference participants, Eliana Cardoso, Vittorio Corbo, Al
Fishlow, Javier Iguiniz, Richard Eckaus, Eduardo Engel, Jose de Gregorio, Caterina Nelson,
Eva Paus, Andres Solimano, Jose Pablo Arellano, Miguel Savastano, and Andrew Zimbalist for
helpful suggestions. The research reported here is part of a project supported by the World
Bank. S. Edwards gratefully acknowledges the support of the University of California Pacific
Rim Program.
‘See, also, Sachs (1989) where the same issue is pursued.

03043878/90/$3.50 0 1990, Elsevier Science Publishers B.V. (North-Holland)

248 R. Dornbusch and S. Edwards, Macroeconomic populism

features, from the initial conditions, the motivation for policies, the argument
that the country’s conditions are different, to the ultimate collapse. It is clear
that the Unidad Popular of 1970-73 experience in Chile had political goals
that were very different from the experience in Peru. Even so, we want to
emphasize that the political mobilization strategy had strongly similar
elements and that, overall, there are remarkable similarities between the
Allende experience and that of Alan Garcia’s Peru.
The similarities in approach in Chile and Peru are particularly striking
with respect to the way policymakers viewed the objective conditions of their
economy, how they proposed that strongly expansionary policies should and
could be carried out, and how they rationalized that constraints could be
dealt with. Most impressive is the fact that in the end. foreign exchange
constraints and extreme inflation forced a program of violent real wage cuts
that ended in massive political instability, violence, and in the case of Chile
even in a coup. We do not doubt the sincerity of the policymakers who
embark on these programs, and we share their conviction that income
distribution is unacceptably unequal. The very sincerity of these policy-
makers convinces us of the usefulness, and indeed the necessity of laying out
exactly how and why the programs do go wrong.
External influences (debt crises and economic blockades, among others),
domestic structural policies (socialization of firms, banks nationalization),
and inconsistent fiscal and exchange rate policies bring about an unsustain-
able situation; inflation goes out of control, and the foreign exchange
constraints force realism on policymakers. Accounts of these experiences
often emphasize politics and, especially, external factors as central to the
demise. External destabilization can be an important part of the unravelling
of an economic program. But we emphasize the extreme vulnerability that
makes destabilization possible. By and large this potential results from
unsustainable economic policies. We think that only to the extent that the
mechanics of the macroeconomics of populist programs are fully understood,
will these policies cease to be popular among politicians.
Before embarking on the case studies we emphasize that we do not cover
the political issues which surely are equally if not more important in the
historical developments in the two countries. We omit politics not because
we think they are irrelevant, but because we want to highlight to the clearest
extent possible the economic developments. The view we present is therefore
possibly biased because it omits the political motivation, on occasions, for
economic motives.

2. The populist paradigm

Populism has traditionally been a fuzzy concept. Political scientists have

struggled to provide a meaningful and precise definition. Drake (1982), for
 R. Dornbusch and S. Edwards, Macroeconomic populism 249

example, emphasizes three elements of a tentative definition: populism uses

‘political mobilization, recurrent rhetoric and symbols designed to inspire the
people’, it draws on a heterogeneous coalition aimed primarily at the
working class, but including and led by significant sectors from the middle
and upper strata and, third, populism ‘has connoted a reformist set of
policies tailored to promote development without explosive class conflict’. He
notes [Drake (1982, p. 218)]: ‘[The programs] normally respond to the
problems of underdevelopment by expanding state activism to incorporate
the workers in a process of accelerated industrialization through ameliorative
redistributive measures’.
Conniff (1982, p. 5) has argued that ‘populist programs -frequently over-
lapped with those of socialism’. We emphasize that the redistributive
objectives are a central part of the paradigm. Whether they are motivated by
a strategy of massive social reform is consequential, but is not central to our
own discussion.
Populist economic programs exhibit strong similarities. In this section, we
set out in paradigmatic fashion what we see as the critical common factors.
In later sections we analyze these for the experiences of Chile and Peru.’
The intellectual identification of issues is summarized under the following
headings:

Initial conditions. Dissatisfaction with the country’s growth performance.

Most typically, though not always, the country has experienced moderate
growth, stagnation or outright depression as a result of previous stabilization
attempts. The experience, often - though not necessarily ~ under an IMF
program, has reduced living standards. Serious economic inequality provides
economic and political appeal for a radically different economic program.
Preceding stabilizations have improved the budget and the external balance
sufficiently to provide the room for, though perhaps not the wisdom of, an
expansionary program.
No constraints. Policymakers explicitly reject the conservative paradigm. In
their view idle productive capacity provides the leeway for expansion.
International reserves and the ability to ration foreign exchange give room
for expansion without the risk of running into external constraints. The risks
of deficit finance are portrayed as exaggerated or altogether unfounded. It is
argued that expansion is not inflationary (if there is no devaluation), because
spare capacity and decreasing long-run costs contain cost pressures; in any
event there is room to squeeze profit margins by price controls.
The policy prescription. Populist programs emphasize three elements: reacti-
vation, redistribution of income and restructuring of the economy. The
common thread here is ‘reactivation with redistribution’. The recommended

‘In a larger project we expect to look at a significant number of experiences in Latin America
to get a sharper picture of the phases and ultimate breakdown of programs.
25q R. Dornhusch and S. Edwards, Macroeconomic populism

policy is to redistribute income, typically by large real wage increases that

are not to be passed on into higher prices. Inflation notwithstanding,
devaluation is rejected because of the inflationary impact and because it
reduces living standards. The economy is to be restructured to save on
foreign exchange and support higher levels of real wages and higher growth.

In populist experiments events tend to evolve in four phases or stages:

Phase I. In the first year or so, the policymakers are fully vindicated in
their diagnosis and prescription: growth of output, real wages and employ-
ment are high, and the macrOeconomic policies are nothing short of
successful. Controls assure that inflation is not a problem, and shortages are
alleviated by imports. The run-down of inventories and the availability of
imports (financed by reserve decumulation or suspension of external pay-
ments) accommodate the demand expansion with little impact on inflation.
Phase II. Bottlenecks, partly as a result of a strong expansion in demand
for domestic goods, and partly because of reactions to the scarcity of foreign
exchange, start to appear. Whereas inventory decumulation was an essential
feature of the first phase, the low levels of inventories and inventory building
are now a source of problems. Price realignments and devaluation, exchange
controls, or increased protection become necessary. The government tries to
stabilize, but fails to put a check on wage increases and on the growth of
government expenditure. Inflation increases significantly, but wages keep up.
A massive underground economy emerges. The budget deficit worsens
tremendously as a result of pervasive subsidies on wage goods and foreign
exchange.
Phase 111. Pervasive shortages, extreme acceleration of inflation, and an
obvious foreign exchange gap lead to capital flight and demonetization of the
economy. The budget deficit deteriorates violently because of a steep decline
in tax collection and increasing subsidy costs. The government attempts to
stabilize by cutting subsidies and by undertaking a real depreciation. Real
wages fall massively, and politics become unstable. It becomes clear that the
populist policies have failed.
Phase IV. Orthodox stabilization takes over under a new government. An
IMF program will be enacted; and, when everything is said and done, the
real wage will have declined abruptly, to a level significantly lower than
when the whole episode began! Moreover, that decline will be very persistent,
because the politics and economics of the experience will have depressed
investment, decapitalized the manufacturing sector, and promoted capital
flight. The severity of real wage declines is due to a simple fact: capital is
mobile across borders, but labor is not. Capital can flee from poor policies,
labor is trapped.

After briefly analyzing the propagation mechanisms of high inflation in

 R. Dornbusch and S. Edwards, Macroeconomic populism 251

populist-type experiences, we turn to the episodes of Peru and Chile to

sketch the policymakers’ diagnosis and the actual events. We start with the
Chilean case, because the facts of the entire experience can now be seen. The
results suggest what to look for in the Peruvian experience.

3. The mechanisms of high inflation under populist macroeconomics

In this section we briefly review the relationship between key macro-

economic variables and sketch the propagation mechanisms of high inflation
in populist experiments. The emphasis is put on the relationship between real
wages, the real exchange rate. real interest rate, the fiscal budget and
inflation. We show that policies based on wage increases and deficit
expansions face severe tradeoffs. While higher wages will be achieved,
inflation will erupt, competitiveness will be eroded and an external crisis will
take place. We show in this section that in the short run, while international
reserves last, the more severe aspects of this tradeoff can be kept under
control.
The starting point is the relation between real wages, u’, the real exchange
rate, R, and employment. Consider a world where there are exportables,
importables and nontraded goods. The consumer price index is

P= P”P”P”
x m “3 (1)

where the superindexes x and m are the share of exportables and importables
in consumption. The home goods share thus is n = 1 -x - m. Home goods are
produced using labor and importables:

P, = a WZS, ?2)

where a is the per-unit requirements of labor and imports, W is the nominal

wage, Z represents the profit markup and S the subsidy [i.e., S= l/( 1 +s)
with s the percentage subsidy rate]. As discussed later, the subsidy may
either take the form of direct cash transfers or of negative interest rates on
credit.
Exports prices are given in the world market so that the domestic price of
exportables is equal to the world price P* times the exports exchange rate e:

P, = P*e. (3)

On the imports side prices are determined by the given world price, P*,
the imports exchange rate and the tariff, T:
252 R. Dornbusch and S. Edwards, Macroeconomic populism

P, = c$P*eT, (4)

where cx denotes the terms of trade and 8, if it differs from unity, measures
the differentiation of the exchange rate structure between imports and
exports. Define the real exchange rate, R, as the ratio of wages to traded
goods prices on a weighted basis for export and import competing industries:

R=(W/PJm”(W/P,)“=(W/eP*)Km”, K=a/lT.

Eq. (5) indicates, then, that the real exchange rate is given by the nominal
wage rate measured in terms of exportables with an adjustment for the terms
of trade and for exchange rate or tariff protection.
The real consumption wage, w, is given by the ratio of money wage to the
consumer price index:

w= W/P=d(R, K,ZS). (6)

Eq. (6), substituting from (1) to (5) in W/P, highlights that real wages and
competitiveness are negatively related. Subsidies or reductions in profit
margins play a special role since they increase real wages without a
deterioration in competitiveness. Protection increases competitiveness, but it
reduces real wages.
On the other hand, assume that employment, L, is demand determined by
profitability in the traded goods sector and by the derived demand for labor
in the home goods sector. A fraction K of labor income is spent on home
goods, and G represents government spending on home goods. In addition
there is credit expansion Q which is made available to state enterprises or the
private sector for spending on home goods. We then have

Ld=p(R)+pc(WL+P,G+Q)/P,, p’>O (7)

or

L = p( R)/( 1 - K/ZS) = L( R, ZS, G, K/W). (7’)

According to (7’) employment is positively affected by subsidies or by a

squeeze of profit margins because these raise real wages and hence demand
for nontraded goods. Moreover, according to this equation, an increase in
government spending and a gain in competitiveness (i.e., a decline in R) will
raise the level of employment.
We further assume that there are various sources of acceleration of the rate
of inflation:

AIT= n(Ld-L)+A(w’-w)+GAR/R-+AS/S. (8)

 R. Dornbusch and S. Edwards, Macroeconomic populism 253

It is implicit in this formulation the presence of automatic indexation and

hence inertia. According to eq. (8) inflation stays constant unless one of three
shocks occur:
_ corrective inflation from real depreciation or subsidy removal,
_ demand inflation associated with an excess of labor demand over the level
of full employment,
_ social conflict shocks which arise when the target real wage, w’, exceeds
actual real wages.

The model is closed by the financing equations: the external balance and
the budget. The external deficit depends on competitiveness and the interest
differential between domestic and foreign assets adjusted for the anticipated
depreciation of the exchange rate, v:

b = B( R, v); b, >O, b,>O. (9)

The budget deficit, on the other hand, is assumed to depend on subsidies, on

foreign exchange losses and tariff revenues, and, via tax collection, on the
level of activity and hence on competitiveness:

d = D(S, /W, R). (104

The deficit, in turn, is financed by reserves decumulation and by money

creation. Money creation depends on the budget (b), the external balance (d)
and credit expansion (q) each measured relative to nominal income:

h=(d+b-q)PY (lob)

To derive actual time paths of alternative policies and their impact on the
path of employment, real wages and inflation the model would have to be
simulated. For our purposes, however, it is already instructive in that it
highlights the basic tradeoffs that are faced by a policymaker who seeks to
expand the level of economic activity (i.e., reducing unemployment) via wage
rate and expansionary fiscal policies. The model in fact shows that real wages
can be raised either by a real appreciation (i.e., real exchange rate over-
valuation) or by subsidies. But real appreciation interferes with the external
constraint and subsidies are inflationary. Employment can be expanded by
subsidies or by real depreciation. The former is inflationary and the latter
hurts real wages. The room for policies that raise real wages and employment
254 R. Dornbusch and S. Edwards, Macroeconomic populism

is therefore severely limited. While there is foreign exchange available the

tradeoffs are not apparent, but once reserves run out the tradeoffs will
suddenly, and cruelly, appear: inflation will then rapidly accelerate.
The acceleration of inflation, once it takes place, is magnified by adjust-
ments in the financial sector. 3 Velocity rises in response to increased
inflation and that implies a decline in the base for the inflation tax. But with
a decline in real balances, even higher inflation rates are required to finance
a given budget deficit. The increasing inflation, in turn, interferes with the
efficiency of revenue collection, leading to a widening of the deficit. Once
inflation becomes very high the relation between inflation and the budget
becomes dominant. Moreover, because of the sharp rise in velocity the
deficits that can be financed at a given inflation fall radically. At this point a
vicious circle takes over, and the only way out is by implementing a severe
stabilization program that cuts the deficit, generates a major real
depreciation and reduces real wages. As will become apparent from our
analysis of the Chilean and Peruvian cases in sections 4 and 5, this model
captures some of the most important aspects of the mechanics of macro-
economic crises under populist regimes.

4. Economic policies in Allende’s Chile

In September of 1970, Salvador Allende, the socialist candidate of the

Unidad Popular (UP), was elected President of Chile. A unique political and
economic experience evolved in the following three years.4 The Unidad
Popular was a political coalition of left and center-left parties dominated by
the Socialist and Communist parties which sought to implement deep
institutional, political and economic reforms. Its program called for a
democratic ‘Chilean road to socialism’.5 Both communists and socialists
initially recognized the multiclass nature of the Unidad Popular and
considered the alliance and the politics that sustained it to be a tactical
intermediate step that would help set the basis for the transition to socialism
[see the discussion in Zammit (1973)]. Vuskovic (1973, p. 50) noted:

. . . economic policy is subordinate, in its content, shape and form, to

the political need for increasing the Popular Unity’s support . The
urgent need to achieve rapid recovery of the economy, and to extend the
benefits to the mass of the working population, cannot be undertaken in
isolation from the structural changes; they are all necessarily inter-

%ee Dornbusch and de Pablo (1988) on the inflation dynamics.

%ee Oppenheim (1989) for a recent review of the literature on this experience.
‘See Stallings (1978) and de Vlyder (1974) for a discussion of politics and economics in Chile.
 R. Dornbusch and S. Edwards, Macroeconomic populism 255

dependent. It is not possible to make deeper changes without broaden-

ing the Government’s political support, and economic reactivation and
income redistribution will provide an impulse to these fundamental
changes.

4.1. Initial conditions, diagnosis and short-run program

The Unidad Popular faced a somewhat stagnated economy with a rapidly

increasing rate of inflation. Between 1967 and 1970, Chile’s real GDP per
capita grew only at 1.2% - significantly below the Latin American average.
Inflation had steadily increased during the last few years of the Frei
administration, reaching 35% in 1970. On the positive side, the balance of
payments showed substantial surpluses during all but one of the Frei years
and the central bank had a significant stock of international reserves of
approximately U.S. $400 million or half a year of imports.
The short-run economic objectives of the UP included:
_ initiating, at a rapid speed, a whole range of structural economic transfor-
mations, including a massive nationalization program;
_ raising real wages, especially for the lower classes;
_ reducing inflation;
_ increasing the rate of output growth;
_ increasing consumption, especially among the poorer groups; and
_ reducing the economy’s dependence on the rest of the world.

The nationalization program was to be achieved by a combination of new

legislation, requisitions, and stock purchases from small shareholders. The
other goals - output and consumption growth, with rising salaries and
declining inflation - were to be accomplished by an increase in aggregate
demand, mainly generated by higher government expenditures, accompanied
by income redistribution measures and severe administrative controls over
prices.
This macroeconomic program, in the structuralist tradition, was based on
a number of key assumptions. First, it was believed that there was ample
excess capacity in the manufacturing sector. Second, it was thought that this
low rate of capacity utilization was closely related to the existing pattern of
consumption and income distribution. Third, it was assumed that there was a
dualistic manufacturing sector, where firms producing ‘luxury’ goods had
excessively high capital/labor ratios. Fourth, inflation was considered to be a
reflection of the economic structure, not of financial or monetary pressures.
In what follows, we will analyze these key assumptions in some detail.
Significant excess capacity in the manufacturing sector was at the center of
the macroeconomic program and provided the intellectual base for the belief
256 R. Dornbusch and S. Edwards, Macroeconomic populism

that large fiscal deficits would not necessarily be inflationary. Americo

Zorrilla, Allende’s first Minister of Finance said:

The subutilization of installed capacity is another feature of the

current economic situation . In 1969 it was possible to increase
production, due to subutilization, by more than 30 percent . .
[Alccording to recent studies . unutilized capacity has reached, in the
last few years, 61 percent in the cloth industry, 50 percent in the baking
industry 74 percent in the shoe industry, etc. [As reproduced in
Garcia (1972, p. 72).]

An important corollary of the unutilized capacity hypothesis was that

firms in many sectors faced decreasing average costs so that, under proper
administrative price controls, demand increases would not generate inflation-
ary pressures. Moreover, to the extent that many of the larger firms were
indeed nationalized, as called for by the UP program, it was thought that
output could be greatly increased even with price reductions. In a document
prepared by the Planning O&e (ODEPLAN) in 1971, it was stated that by
‘combining the increase in production derived from a higher use of installed
capacity . the former monopolies will be able to absorb, without problems,
the required wage increases, while maintaining, or even reducing, prices and
still generating the same surplus’.
The lack of ‘full’ utilization was, in turn, attributed to two fundamental
factors: The monopolistic nature of the manufacturing industry and the
structure of income distribution [Vuskovic (1970)]. Based on this diagnosis,
it was thought that if income was redistributed towards the poorer groups
through wage increases and prices were properly controlled, there would be a
significant expansion of demand and output.
Regarding inflation, the UP program emphasized rigidities, bottlenecks,
and the role of monopolistic pricing and played down the role of fiscal
pressures and money creation6 Very little attention was given to the
financial sector. In fact, in his memoirs, the former Allende Minister and Vice
President Clodomiro Almeyda relates how, in the first meeting of the
economic team after the elections, the CEPAL-oriented technocrats expressly,
and convincingly it would seem, argued that monetary and financial
management did not deserve too much attention.’ Alfonso Inostroza, the
President of the Central Bank, stated in early 1971 that the main objective of
the monetary policy was to ‘transform it into a key instrument . . to achieve
the complete mobilization of productive resources, and their allocation to

60f course, this was consistent with the structuralist view on inflation. See Sunkel (1960) and
Baer and Kersternetzky (1964).
‘See, for example, the 1971 CIAP Report reproduced in Panorama Economico, no. 260,
Feb.-Mar. 1971, p. 36.
 R. Dornbusch and S. Edwards, Macroeconomic populism 251

those areas that the government gives priority to . .’ [see Inostroza (1971,
P. 811.
As to Chile’s external vulnerability, the UP basic program stated that,
along with a reduction in import dependence, a priority of the new
government would be to ‘execute a foreign trade policy tending to expand
and diversify our exports . .’ [Unidad Popular (1969, p. 24)]. This objective
was to be achieved without providing any price incentives. Quite the
contrary, the UP economists thought that changes in the exchange rate had
very little, if any, effects on exports or imports. In fact, it was stated that an
important goal of the policy was to ‘avoid the scandalous devaluations of
[the] . . . currency’ (p. 24). Indeed, one of the first measures undertaken by
the Allende administration was to eliminate the system of crawling peg that
had been adopted, with great success, by the Frei administration.
The UP view of the way the economy functioned ignored many of the key
principles of traditional economic theory. This was reflected not only in the
greatly diminished attention given to monetary policies, but also by the
complete disregard for the real exchange rate as a key variable in determin-
ing macroeconomic equilibrium. Moreover, the Unidad Popular failed to
recognize that their policies would be unsustainable in the medium term and
that capacity constraints were going to become an unsurmountable obstacle
to rapid growth. Also, the UP technocrats greatly underestimated the role of
expectations and the capacity of the public to react to severe inflationary
pressures. Bitar (1986, ch. 5) portrays very clearly the government’s inability
to control events, and to shift the emphasis from redistribution to
accumulation.

It turned out to be very difficult to contain the forces unleashed in 1971.

The sequential conception of redistribution followed by accumulation
assumed that basic political and social conduct could be altered and
popular expectations changed virtually instantaneously. In the next few
months [early 19721 it proved impossible to apply this thinking with the
facility that had been hoped for.

4.2. Phase I: Rapid growth with repressed inflation

Armed with the intellectual framework analyzed above, the UP rapidly

began to implement its program in late 1970. In terms of structural reforms,
two basic measures were immediately undertaken: First, the Agrarian Reform
was greatly intensified with a very large number of farms being
expropriated.* Second, a project for a Constitutional Amendment aimed at

*The agrarian reform law passed by Congress during the Frei administration provided the
necessary tool. See Alaluf et al. (1972).
258 R. Dornbusch and S. Edwards, Macroeconomic populism

1
90

80

70
(
60 I 11

1s)67 68 69
I

70
I

71 72
/

73 74 75
I I

76 77
I /

78
1

79 80
I /

81
1

82

YEAR
Fig. I. Chile: The real wage (index 1967-69= 100).

nationalizing the large copper mines - until then jointly owned by large U.S.
firms and the Chilean state - was studied.”
In terms of macroeconomic policy, the government rapidly applied
measures that were consistent with its program. Government expenditures
greatly expanded and in 1971 real salaries and wages in the public sector
increased by 48”;, on average. Salaries in the private sector grew at
approximately the same rate. Fig. 1 shows the gain in real wages during this
year. The short-run package was topped by a generalized scheme of price
controls.
In the first two quarters of 1971, manufacturing output increased 6.2:/, and
10.6”!~; compared to the same periods in the previous year. Manufacturing
sales grew at even faster rates: 12O, during the first quarter and 1I’:6 during
the second quarter. Overall, the behavior of the economy in 1971 seemed to
vindicate the UP economists: real GDP grew at 7.7”/,, average reul wages

90n June I I, 1971, Congress unanimously approved the constitutional reform that nationa-
lized large copper mines. See Geller and Estevez (1972). and Ffrench-Davis and Tironi (1974).
The reforms of the banking system and large manufacturing firms were somewhat more difficult,
because the government lacked the institutional channels for implementing the nationalization
program. Initially. this obstacle was overcome by purchasing blocks of shares especially bank
shares at very high prices. These acquisittons were compiemented by a process of requisition
based on an old, and until then forgotten, decree law promulgated during the short-lived
Socialist Republic of 1932.
 R. Dornbusch and S. Edwards, Macroeconomic populism 259

increased by 17%, aggregate consumption grew at a real rate of 13.2%, and

the rate of unemployment dipped below 4%. Not too surprisingly, given the
behavior of real wages, there was a significant improvement in income
distribution. In 1971 the labor’s share of GDP reached 61.7x, almost 10
points higher than its 1970 level of 52.3%; and this happened without an
acceleration of inflation. What was even more important for the UP was that
the policies rapidly paid off politically. In the municipal elections of 1971, the
UP parties saw their share of the vote climb from the 36% They had received
in the 1970 presidential election to around 50”/;;.”
All of this created a sense of euphoria in the government. The fact that the
fiscal deficit had jumped from less than 3% of GDP in 1970 to almost 1 l?G
in 1971 did not worry the UP technocrats. Nor did they think much on the
fact that the rate of growth of the money supply had surpassed 100% in
annual terms in the fourth quarter of 1971, and that the rate of growth of
domestic credit to the public sector was approaching 300%. All of this, in
fact, was part of the plan [see Lopez (1972)].
Of course, these macro policies were rapidly generating a situation of
repressed inflation. The high growth rate of GDP in 1971 rested heavily on
an almost 40% increase in the imports of intermediate goods. As a result, the
stock of international reserves inherited by the Allende government was
reduced by more than one-half in that year alone. A steep reduction of
inventories was another important factor contributing to the expansion of
consumption. By the end of 1971, the mounting inflationary pressures
became evident (see table 1). The supply problem was aggravated by a series
of labor disputes in many large establishments that resulted in the takeover
of those firms by their workers. In fact, this procedure became the
institutionalized way in which the government seized a large number of
firms.’ ’

4.3. Phase II: Inflation, bottlenecks and.fuiled stuhilization progrums

During 1972, the macroeconomic problems continued to mount. Table 1

shows that inflation reached 217% and the fiscal deficit surpassed 13% of
GDP. Domestic credit to the public sector grew at almost 300%, and
international reserves dipped below $77 million. Consumption growth domi-
nated the expansion in demand (see table 2).
The underground economy grew as more and more activities moved out of
the official economy, and as a result more and more sources of tax revenues

‘OThe strong showing in early municipal or Congressional elections was not uncommon in
Chile. In fact, the Frei government had a similar experience in 1965.
“See Dornbusch and Edwards (1989) for a more detailed discussion of the nationalization
process. See, also, Alaluf et al. (1972).
260 R. Dornbusch and S. Edwards, Macroeconomic populism

Table 1
Chile: Main macroeconomic indicators.

1970 1971 1972 1973

Inflation” 34.9 34.5 216.7 605.9

Growth 2.1 9.0 -1.2 - 5.6
Real wages ( 1970: 3 = 100) 98.4 115.1 103.5 70.3
Government revenueb 23.7 20.4 18.2 20.2
Government spendingb 26.4 31.1 31.2 44.9
Budget delicitb 2.7 10.7 13.0 24.7
Money growth” 52.9 99.3 100.9 264.4
International reserves ($ million) 320 129 95 36
Trade balance ($ million) 246 73 - 161 -73
Black markei premium (“,,) 99 358 898 2349

“Percent, Dec.-Dec.
bPercent GDP.
Sources: Banco Central de Chile; International Monetary Fund; Edwards
(1986); Edwards and Cox-Edwards (1987): Solimano (1988).

Table 2
Chile: The growth rate of real GDP and demand (percent
per year).

1970 1971 1972 1973

GDP 2.1 9.0 - 1.2 -6.6

Consumption
Private -0.5 13.2 7.7 -6.6
Government 5.9 12.4 5.7 1.7
Investment 6.5 - 2.0 -20.1 - 6.0
Exports 2.1 0.8 ~ 15.1 2.8
Imports 0.9 8.5 3.2 -5.4

Sourw: Banco Central de Chile. Indicadores Economicos

)‘ Socictles l960- 198.5.

disappeared. A vicious cycle took over; repressed inflation encouraged the

informal economy, resulting in reduced taxes, higher deficits, and, thus, in
even higher inflation. In 1972, two stabilization programs were implemented,
and boith failed.
When evaluating the difficulties, the dominant view among UP economists
was that the authorities had failed to impose appropriate controls in
implementing the macroeconomic program [Garcia (1972)]. This view guided
the first, rather weak, attempt at stabilizing the economy which was launched
in February of 1972. No serious measures aimed at correcting the major
macroeconomic causes of these problems were undertaken. It simply was not
convenient to reduce government expenditures, the policy of granting salary
 R. Dornbusch and S. Edwards, Macroeconomic populism 261

30

ts
I
4
m I

I’ I I 1 ( I ’ I
1964 1967 19bO 1973 1976 1979 1962?&
YEAR
Fig. 2. Chile: The black market for dollars.

increases that exceeded inflation was maintained, and a significant devalu-

ation was ruled out.12
The most serious problem was the loss of control over wages. Unions in
both nationalized and private firms demanded increasingly higher wage
adjustments. The government faced a dilemma. If it refused to grant the wage
adjustments, it would help the macroeconomy, but would hurt its political
objectives. Under these circumstances, the government chose, every time, to
uphold its revolutionary label.13
By mid-1972, it became apparent that the February stabilization program
was a failure. The underground economy was now generalized, output began
to fall, open inflation reached an annual rate of 70% in the second quarter,
foreign exchange reserves were only $82 million, and the black market rate
was climbing at a very fast pace (see fig. 2). From a political perspective,
parliamentary elections scheduled for March of 1973 made the situation
particularly difficult. An economic crisis would work against the UP in those
elections. In August of that year, and under the political monitoring of the
Communists, a new stabilization program was launched.
Unlike the previous plan, the cornerstone of the August program was a
massive devaluation of the escudo. It was expected that, as a result, the ever-

‘% 1971 a small devaluation was followed by the creation of four different exchange rates.
13Moreover, a number of economists argued that higher real wages could be sustained as long
as the government was able, via increased controls, to extract additional ‘surplus’ from the
private sector. See Bitar (1979).
262 R. Dornbusch and S. Edwards, Macroeconomic populism

mounting pressures on the balance of payments would subside. The program

also called for two basic measures to contain the fiscal pressures. First, price
increases for the nationalized firms were authorized so as to reduce the
financing requirements of the newly formed nationalized sector. Second, the
program called for a massive increase in production as the major way to
close the gap between aggregate supply and aggregate demand. This increase
in output was expected to be a response to political rather than economic
incentives. In fact, at that time, the Communist Party’s main slogan became
‘Let’s win the battle for increased production!‘. The devaluation and a large
number of price increases resulted in a monthly rate of inflation of 22.77; in
August and 22.2”, in September.
In spite of the action taken on the exchange rate front, the program failed
as no change in the wage rates policy was introduced. In the second week of
August, the government announced that it had reached an agreement with
the national federation of workers (Central Unica de Trabajadores, CUT)
with respect to an across-the-board wage adjustment to be granted on
October 1st except for firms subject to private bargaining. The new wage
policy called for an increase in public and private sector wages by a
proportion equal to the accumulated rate of inflation between January and
September. In addition, the new policy called for more frequent wage
adjustments (Que Pmz, no. 70, Aug. 17, 1972, p. 14). In this way, by a stroke
of the pen, the effects of the devaluation were fully offset.

4.4. Phase III: Acceleration qf iqflution and chaos

During the first quarter of 1973 Chile’s economic problems reached

chaotic proportions. Compared to the first quarter of 1972, inflation reached
an annual rate of more than 120”;; industrial output declined by almost 6%;
the real exchange rate was even more overvalued, and foreign exchange held
by the Central Bank was barely above $40 million. The black market by
then covered an ever widening range of transactions in foreign exchange. The
fiscal deficit continued to climb as a result of ever higher expenditures and of
rapidly disappearing sources of taxation. In that year, the fiscal deficit
exceeded 230/ of GDP.
Once more the policymakers faced the options of implementing a major
corrective stabilization program or of furthering the extent of controls. And
once again they opted, in March of 1973, for the latter. The extent of the
economic crisis quickly alienated the middle classes, and the political
confrontation with the opposition became increasingly severe.14

140f course, it must bc remembered that, in the midst of this polarization, the support
commanded by Allende actually grew: whereas in 1970 the UP received the votes of 30% of the
electorate, in the municipal elections of 1971 it reached 50’4 and even in 1973 its share still
stood at 44’5”.
 R. Dornbusch and S. Edwards, Macroeconomic populism 263

In concluding, we must comment on the role of domestic economic

opposition and the foreign economic blockade. The strategic use of economic
disruption by the opposition, foreign enterprises and foreign governments
indeed played a role in the ultimate unravelling of the Allende policies. A
more neutral external environment probably would have allowed the Allende
experiment to continue for some time. But the uncontrolled side effects of the
consumption growth policy (shortages, inflation and related effects) sufh-
ciently weakened the ability to govern.” As a result, destabilization by
domestic opposition forces and foreign companies and governments could be
effective. We advance this view as a hypothesis.
On the 11th of September 1973, the Allende presidency came to a sudden
and shocking end. When the military took over, the country was politically
divided and the economy was in shambles. Inflation was galloping, relative
price distortions - stemming mainly from massive price controls ~ were
generalized; black market activities were rampant; real wages had dropped
drastically; the economic prospects of the middle class had been greatly
damaged; the external sector was facing a serious crisis; production and
investment were falling steeply; and the government finances were completely
out of hand. This was the beginning of Phase IV, a stage where frightful real
wage cutting (see fig. 1 above) took over.

5. Growth with redistribution in Garcia’s Peru

When Alan Garcia assumed the Peruvian Presidency in August 1985, he

captured the world’s imagination: a dynamic, charismatic leader taking
charge of a country desperately in need of social and economic progress. The
fact that he adopted a confrontational attitude on external debt did not hurt
his image, either in Latin America or in progressive circles in Europe and the
United States. On the domestic front, he had an unambiguous message:
growth and redistribution. That policy lasted two years before running
aground in a catastrophic manner.
In early 1988, Alan Garcia’s populist government staged a dramatic turn-
around on the policies that had driven the country to bankruptcy: budget
cutting, real wage cutting and massive exchange depreciation were the
predictable aftermath of three years of reckless mismanagement. But that was
not the end of the story, hyperinflation was to follow and the political
consequences of economic destruction and pauperization are still to come.
Real wage cutting and yet worse poverty, in the Peruvian context, may well

‘5Rosenstein-Rodan [in Orrego Vicuna (1975, pp. 219-220)] notes of the Allende policies,
oriented toward consumption: ‘This part of Allende’s policies was more Populism than
Socialism. Even Fidel Castro is supposed to have observed ‘Marxist socialism is a revolution of
production - this is a revolution of consumption’.
264 R. Dornbusch and S. Edwards, Macroeconomic populism

Table 3
Peru: Growth and inflation (percent per year).

195(1-60 196&70 197&75 1975-80 198g-85

Growth per capita 2.8 2.6 1.6 -1.0 -4.2

Inflation 8.0 9.3 12.6 50.0 102.1

SOWW: Kuczynski (1977) Central Bank and Ministry of Finance

be the opening phase for massive and perhaps violent confrontations.” fin
the rest of this section we analyze in some detail the recent experience with
populist policies in Peru. The analysis shows a remarkable similarity between
this case and the Chilean episode.

5.1. From Beluunde to Gurcia

In the 1950s and 1960s Peru experienced significant growth of real per
capita income and moderate rates of inflation.” Table 3 shows that the
problems of declining per capita income and high inflation date from the
second half of the 1970s.
In the past 25 years, three major stop-go phases can be discerned. The first
was the Belaunde expansion which crashed in 1967-68. The second phase
ran until 1974. During this period the extremely favorable world environ-
ment permitted a massive expansion until 1974, with an average growth rate
of per capita income of 3.7% per year in the 1969-74 period (see fig. 3).
The second Belaunde administration (198tL85) had to cope with extra-
ordinarily adverse conditions. Belaunde inherited an economy with deep
social problems: Per capita income had declined since 1974 and, the external
balance improvement of the late 1970s notwithstanding, the interventionism
of the military government had created pervasive distortions. The possibility
of advancing the economy was drastically limited as a result of a combi-
nation of shocks: the world recession of 198tL82, terms of trade deterio-
ration, the explosion in world interest rates and the resulting rise in debt
service obligations, external credit rationing, and natural disasters. These
shocks combined to choke off any room for expansion. In 1982-83, under an
IMF program, real GDP per capita declined by 16% and inflation nearly
doubled to 112Y,,.

“More so than in any other country of Latin America, economic performance is central to
maintaining the very precarious social peace in Peru. At issue is not only the possible
confrontation between left and right. Far more dangerous is the widening conflict opened by the
Maoist Shining Path guerrilla. The outcome is wide open because of the divisions between rich
and poor, the city and the sierras. white and indios.
“See Thorp (1987). Thorp and Bertram (1978). and Kuczynski (1977) for history and
extensive references.
 R. Dornbusch and S. Edwards, Macroeconomic populism 265

110

105

100
1

95

90

85 I

80

7oy , I I , / I I I I I / 1 I I
1960 62 64 66 68 70 72 74 76 78 80 82 84 86 88

YEAR
Fig. 3. Peru’s real per capita income (index 1980= 100).

This disastrous economic performance of the Belaunde government led to

a total rout in the elections and thus brought Alan Garcia into power in
1985 with 45.8% of the votes. l8 The election which swept Garcia into office
in a landslide victory carried one clear message: growth first! [see Ortiz de
Zevallos (1989)]. This message, regardless of its feasibility, conditioned the
economic policies of the Garcia Administration from the very beginning and
is now taking the country to the brink.

5.2. Diagnosis and policies: The heterodox program

In July 1985, the last month of the Belaunde administration, inflation

reached 250% at an annual rate. Unemployment was pervasive, idle capacity
abounded, and real wages had been reduced in an effort to cope with the
external crisis. Against this background, Garcia’s government developed a
populist project of expansion, entitled ‘Growth with Redistribution’, which
emphasized economic recovery combined with disinflation.

Premises and policies. Three points are central to an understanding of why

economic policy took the form it did. The tirst is the impressively unequal
distribution of income. Fig. 4 shows the income distribution in Peru with a

IsSee Wise (1988) and Ortiz de Zevallos (1989) for a review of politics.
266 R. Dornhusch and S. Edwards, Macroeconomic populism

PER CAPITA PERCENT OF THE

INCOME (1979 INTIS) POPULATION

12,746 1 .O%

253 46.1%

/ \
223 13.8 23.5%

129 10.0 29.4%

Fig. 4. Peruvian income distribution

diagram widely used by the authorities. The striking fact is that l”/o of the
population receives nearly half the national income [see Glewwe (1988)].
The second is that policymakers were impressed with the large gap
between actual and perceived potential output. In Carbonetto et al. (1987, p,
41) it is estimated that actual output was only 66% of potential output in
1984, leaving a 34”,/, gap to be made up by the judicious choice of policy.
The third consideration is that Peruvian economists, in company with
economists throughout Latin America, were unimpressed with the effective-
ness of IMF programs. It should be remembered that at this time Israel and
Argentina made their heterodox stabilization efforts, as did Brazil shortly
afterwards. The orthodox approach to stabilization had been discredited by
the strong recessionary effects and the absence of any success stories. The
alternative, heterodoxy, had all the appeal of offering an end to inflation
without the attendant costs of unemployment.”
Against this background, the general theme of economic policy was
summarized in the Plan National de Desarrollo 1986-90 (p. 63) from which
we quote extensively.

The new economic policy seeks to pass from an economy of conflict and
speculation to one of production and consensus. In this economy it is

“See Dornbusch (1982, 1988a), Bresser Pereira and Nakano (1987), and Bruno et al. (1988).
 R. Dornbusch and S. Edwards, Macroeconomic populism 267

possible to make compatible stability, growth, distribution and develop-

ment in a context of national planning which finds concrete expression
in dialogue and social and economic concertation.
Planning of economic development will be full, decentralized and
participatory and concertation will center on the effort to make
compatible the generation of savings and productive investment with
attention paid to the undelayable priority of attending to social needs.
We need to reconcile economic efficiency with social equity in a
productive dynamics which is fundamentally sustained by domestic
resources.

Specific premises and prescriptions in the national development strategy can

be paraphrased as follows [Plan National de Desarrollo 198690 (pp.
63-65)]:

- The necessity to redistribute income as a means for sustained growth and

the possibility to bring together with the redistribution process the
necessary capacity to save and invest.
- Salaries and profitability: The generalized and open-ended restraint on
wages reduces profitability because it reduces workers’ purchasing power,
bringing about recessive effects that reduce demand and thus the benefits of
a dynamic economy.
- The fiscal deficit: The fiscal deticit is not necessarily inflationary. It only is
so if domestic demand exceeds potential output. With substantial idle
capacity there is a need for a certain deficit.
- Money creation: Must increase demand, thus allowing an increase in real
liquidity since the opposite would slow growth.
- Interest rates: Increased real interest rates do not raise savings, since the
latter depend fundamentally on income, but they discourage productive
investment.

The specific targets of the economic program are set out in table 4. The
new economic policy was based on four measures [Carbonetto et al. (1987, p.
15)] which in many ways resemble the Allende policies which we discussed
above:

- Rapid expansion of effective demand via real wage increases.

- Financial de-strangulation to give firms relief and contain cost pressures by
reducing financial costs of enterprises, effective interest rates, indirect taxes
and other elements of costs.
- Reestablishment of selective exchange rates and abandonment of the
devaluation policy.
- The external accounts were to be kept in balance by the growth in exports,
import substitution and limitations on debt service to be compatible with
acceptable growth. Of course, many of the foreign exchange and growth
268 R. Dornbusch and S. Edwards, Macroeconomic populism

Table 4
The 198690 Peruvian economic development plan.

1985 1986 1987 1988 1989 1990

Growth
GDP I .4 6.5 6.2 6. I 6.3 6.2
Consumption 0.1 8.5 6.3 5.7 5.0 5.2
Exports 3.5 - 13.6 4.8 5.8 5.9 5.9
Imports 24.6 5.0 13.7 11.9 5.6 7.4
Investment 112.3 12.0 14.4 11.8 11.6 10.7
Budget deficit” 0.6 2.3 2.3 2.1 1.8 1.6

“Percent of GDP.
Source: Presidencia de la Republica (1986).

Table 5
Peruvian macroeconomic indicators.

1985 1986 1987 1988

Inflation” 158 63 115 1722
Growth 2.5 9.5 6.9 ~ 8.4
Real wageb Ill 126 137 105
Trade balance (S) 1173 -67 -463 -84
Government revenues’ 42.7 32.1 25.9 23.5

“December-December.
‘Index July, 1985 = 100.
‘Percent of GDP.
Source: World Bank, Ministry of Finance, National Institute of
Planning.

policies on which any success of the plan might have depended never saw
the light of day. This is particularly the case for foreign exchange savings
strategies.

5.3. Phase I: The first two years

The immediate priority for the Garcia administration was to introduce a

‘heterodox’ program of stabilization: inflation reduction via an incomes
policy combined with a massive reactivation of the economy.
In the short term, the heterodox program was immensely successful.
Inflation declined sharply (as shown in table 5) employment increased, and
the real wage was pushed up substantially (as shown in fig. 5). In the last
quarter of 1987, the real wage stood 52 % above the level of 1985. Growth,
too, had been very substantial. In 1986, the economy grew by 9.5%; and, in
1987, by another 6.9’;/,.
It is important to recognize just how successful the heterodox approach is.
Given enough foreign exchange and a depressed economy, expansion of
 R. Dornbusch and S. Edwards, Macroeconomic populism 269

60
i

50

i
40- i

30 /,,,,,,,/,,,,,,,,,,,,/,,,/,,,,,,,,,,
1980 81 82 83 84 85 86 87 88 89

YEAR
Fig. 5. Peru: The real wage (index 1982:2= 100).

domestic demand can work. In fact, the success is broadly shared because the
recovery of demand can raise firms’ profitability by raising capacity utiliza-
tion. That was indeed the case in Peru. A year after the program started
Garcia was celebrated by the business class for the success of his recovery
strategy. Private investment increased by 24% in 1986 and another 18.6% in
1987.20
By early 1987, the program was at the peak of its success: real GDP had
grown cumulatively more than 20% since the third quarter of 1985, while
inflation had been reduced from 188% to only 75%. Despite the startling
success, the strain was starting to appear in rising cost pressures and a
growing loss of foreign exchange. However, as in the case of Allende’s Chile,
these strains were obvious to economists, but were far from alarming to
policymakers or the public.

5.4. Phase III: Inflation and collapse

The turning point came in July 1987, in the form of a political conflict that
erupted as a result of a proposal to nationalize the banking system. On July

“See Centro de Economia Aplicada (1988) and especially Iguiniz (1988) for an evaluation of
the first three years of the program.
270 R. Dornbusch and S. Edwards, Macroeconomic populism

27th, 1987, President Garcia declared [see Presidencia de la Republica

(1987b, p. l)]:
In Peru, today, the financial system is the most powerful instrument of
concentration of economic power and thus of political influence; it is the
major obstacle to the democratization of production and the accumu-
lation of the surplus.
At this time the right, with public appearances of Vargas Llosa, drama-
tized the shift in government policies toward socialism. This juncture also
represented the end of the recovery policy and the turning point toward
inflation and foreign exchange crises. Although the constraints tightened only
gradually and bottlenecks emerged only in few places, during the remainder
of 1987, it is fair to say that July-August of that year represented a crucial
period, after which the continuation of expansionary policies could no longer
be defended.
Populist programs such as that practiced in Peru fail when the economy
runs out of foreign exchange and when the controls that support the initial
redistribution and expansion have to be dismantled. By late 1987, growth
was petering out and inflation, brought about by external constraints,
bottlenecks and the adjustment of severe price distortions, exploded.
To understand what went wrong it helps to return to the basic philosophy
of the program. This is fully documented in EI Peru Hetrrodoxo: Un Modelo
Economico, a 500-page book published in mid-1987 by the economic
architects of the program [see Carbonetto et al. (1987)]. The most striking
revelation of this book is the extraordinary extent to which policymakers in
the Garcia administration, as had the Allende policymakers before them,
diverged from accepted economics. Thus we learn (pp. 75-76) that
An examination of the Peruvian record reveals that periods of moderate
inflation are associated with expansionary fiscal policies. And periods of
major inflation are associated with fiscal restraint. Thus, the record
shows exactly the opposite of what is predicted by a theory which
explains inflation by fiscal deficits.
And, to dispel any doubts (p. 82):
If it were necessary to summarize in two words the economic strategy
adopted by the government starting in August 1985 they are control
(meaning control of prices and costs and recognizing that this could be
done only temporarily for the first twelve months) and spend, transfer-
ring resources to the poorest so that they increase consumption and
create a demand for increased output, thus ‘justifying’ that idle capacity
be put to use.
It is necessary to spend, even at the cost of a fiscal deficit, because, if
this deficit transfers public resources to increase consumption of the
poorest they demand more goods and this will bring about a reduction
 R. Dornbusch and S. Edwards, Macroeconomic populism 271

Table 6
Peru: Real levels of controlled prices (July 1985= 100).

Dec. 1986 Dec. 1987 Dec. 1988

Average” 75 59 58
Rice 73 61 47
Light, electricity 49 40 16
Telephone 84 87 32
Gas 58 33 25

“Weighted by expenditure shares

Source: Apoyo.

in unit costs. Thus the deficit is not inflationary, on the contrary! This
constitutes without doubt the basic premise on which the economic
team acted and the major departure from the earlier strategy that had
emphasized adjustment from the demand side.

Peru learned in 1987-88 that there indeed are serious tradeoffs in policy:
continued rapid growth and massive real wage increases are incompatible
with moderate inflation21 In 1986 inflation was only 63%, far below the
level at which the program started. But the experience of high growth with
moderate inflation, after a while, became an artifact of the controls and
subsidies on public sector prices and on foreign exchange. Subsidies and
controls were used to avoid price increases in politically sensitive areas. In
1987-88, most petroleum products sold at one-third their July 1985 price.
Electricity prices, the price of rice, and bus fares had declined by more than a
third, as had the real price of foreign exchange (see table 6).

5.5. Phase III: The hyperinjlation

When adjustments in the controlled prices had to be made, inflation

accelerated rapidly. 22 Naturally, the massive increase in the budget deficit
also contributed to the acceleration of inflation. The deficit, in turn, greatly
increased as a consequence of a massive policy of subsidies, and as a result of
the extraordinary decline in real tax collection (see table 7). In 1975-86, tax
collection had averaged above 12% of GDP, and in 1985, it rose to 13.2%.
However, by 1988, and as a result of a sharp decline in compliance, tax
collection had fallen to only 7.5% of GDP. Additionally, the Central Bank’s
policy of multiple exchange rates contributed in an important way to the
deficit. This policy alone resulted in losses amounting to 2% of GNP from
buying foreign exchange at a high price from exporters and selling it at a low
or subsidized price to importers.
‘IThis of course, is exactly what we point out in our model in section 2
**See bornbusch and Edwards (1989) for further details on this issue.
272 R. Dornbusch and S. Edwards, Macroeconomic populism

Table 7
Peru: Public sector linancing requirement (percent of GDP)

1985 1986 1987 1988”

Overall public sector deficit 4.4 6.7 9.9 6.1
Budget deficit 2.4 4.9 6.5 5.3
Central Bank losses 1.9 1.8 2.8 0.8
Credit to development banks 0.8 1.4 1.9 0.8
Public sector ftnancing requirement 5.1 8.1 11.2 6.9
Domestic financing 1.2 5.1 9.8 6.9

“Estimate.
Source: World Bank, Central Bank and Ministry of Finance.

-0.2 -

-0.4 -

-0.6 -

-0.6 -

-l-
I I I1 I I , I I I, I I , T 1 I
1985 1986 1987 1988 1989

YEAR
Fig. 6. Peru: Net foreign reserves (billion $U.S.)

The large budget deficit not only contributed to inflation, but also affected
the allocation of credit and hence investment. While foreign exchange
reserves lasted, the government could sell foreign exchange rather than
borrow in the home market or print money. However, with reserves
precariously low, there was no room left for further foreign exchange sales to
finance the deficit (see fig. 6). 23 The declining reserves forced a more

Z3The net reserves included on the liability side $800 million of arrears to the IMF. The gold
position of the central bank had been revalued at various points and amounted to $659 billion
in December 1988.
 R. Dornbusch and S. Edwards, Macroeconomic populism 213

consistent exchange rate policy which immediately raised the inflation rate.
Thus external financing of the budget deficit ultimately had an inflationary
cost, even though it may have been delayed by a year or two.
As has traditionally been the case in populist experiences, the government
also resorted to the banking system to help finance the deficit. High reserve
ratios for banks or direct financing requirements effectively achieved this. The
counterpart of this policy was a 30% decline in real bank credit to the
private sector in three years.
Since September 1987, the World Bank and Peruvian authorities had been
discussing stabilization. But the political impetus for major policy changes
was missing. The only impetus for change came from the external balance
side where a crisis had been building up quite visibly. The trade surplus of
above $1 billion in 1985 had become a deficit by 1987. Reserves declined by
over $1 billion to the point where net reserves were negative at the beginning
of 1988. The response to the looming foreign exchange crisis was a major
devaluation in late 1987, which in the short run resulted in a small real
depreciation. This helped stem capital flight and the widening trade deficit
for some time. However, as in Allende’s Chile, shortly after this corrective
step, generalized and massive wage increases were granted. In a vicious cycle,
then, accelerating wage and price increases eroded the initial gain in
competitiveness. The renewed appreciation of the real exchange rate for
imports since the end of 1987, and the resulting exchange losses of the
central bank, signal the government’s inability to force a real depreciation.

5.6. Phase IV: What next?

Economic mismanagement need not be shortlived. If the politics are

supportive and external destabilization is not a factor, such a regime can last
another year or even more. At the time of this writing there seem to be few
avenues open to Peru. One possibility is that Garcia makes time by
radicalizing his own position and policies in the direction of socialism and
increasing government intervention. A massive program of nationalizations
and control, of the type Allende pursued in his last year, would give the
government more than a few months of breathing space, except if it were
overthrown.
But while there is apparent stability, the extreme reduction in living
standards and the growing number of strikes do suggest that surviving to the
elections in April 1990, will not be easy. 24 The very depth of the economic
collapse in late 1988 caused so sharp a fall in imports that the external
constraint, for a while, was lessened. The real exchange rate in the informal

24By April 1989 the decline in economic activity and the restrictions of imports had become
so massive that an actual reserve recovery had taken place. The reserve gains were sufficient to
feed a rumor of another reactivation program.

J.D.E. B
274 R. Dornbusch and S. Edwards, Macroeconomic populism

Table 8
Peru: Real wages and real per capita GDP: 1985-1989 (index July,
1985 = 100).

Real wages

Minimum Per capita

National average wage Government GDPb
1985 111 120 113 100
1986 126 124 118 107
1987 137 131 136 112
1988 105 118 101 110
1989” 56 64 43 81’

“July 1989.
‘Index, 1985= 100.
‘April 1989.
Source: Peru: Ministry of Economics.

market actually declined to pre-1987 levels. But the collapse of real wages
and activity was devastating. In fact, official estimates placed the decline in
per capita real income, between 1987 and April 1989, at 25% (see table 8).
Just how far the disintegration of the economy has gone is apparent from a
survey in January 1989 reported by Apoyo, an economic consulting service in
Peru. Middle and upper income respondents predominantly indicated that
they would buy dollars given extra income. Of the lower income groups,
more than half reported that they would spend the extra income on food.

6. Concluding remark

It is clear that the two instances of populism discussed here led to

disastrous consequences for those who were meant to be the beneficiaries.
The central question then is whether populist policies are outright unsustain-
able, or whether there is a variant which, properly executed can in fact
succeed. We leave to further research the elaboration of the thesis that
expansionary policies - somewhat similar in spirit to populist policies - can
succeed provided they stay far clear of foreign exchange constraints, empha-
size reactivation only for a brief initial period and then shift to growth
policies. Most important for success, expansionary policies need to be aware
of capacity constraints and have to rely for their financing on an extremely
orthodox fiscal policy and rigorous tax administration. Within those restric-
tions it would appear that there is room left for achieving redistributive
objectives in an effective way.”

‘sSee Dornbusch and Edwards (1989) for a detailed discussion on three alternative future
scenarios for Peru, See Edwards (1989) for a discussion on the inability of the Peruvian advisors
to learn from the Chilean experience.
 R. Dornbusch and S. Edwards, Macroeconomic populism 275

When looking forward into the next few years, one cannot avoid
wondering whether economic systems have ‘memory’, and whether the
mistakes of past populist regimes have been learned. This question is
particularly relevant for the Chilean case where in March 1990 a new
government - in all likelihood of centre-left persuasion - will face urgent and
immediate pressures to improve the social conditions of the poor. Needless
to say it is not possible to answer this question today. However, both the
writings of the economic team of the likely new Chilean government, as well
as the economic program of the Concertaci&, do suggest that some of the
more important lessions regarding the design of economic policy have indeed
been absorbed. Only the passing of time will allow us to verify this
conjecture.

References
Almeyda, C., 1987, Reecuentro con mi Vida (Ediciones del Omitorrinco, Santiago).
Alaluf. D. et al.. 1972. La Economia Chilena en 1971 (Institute de Economia. Universidad de
Chile, Santiago).
Baer, W. and 1. Kersternetzky, 1964, Inflation and growth in Latin America (Yale University
Press, New Haven, CT).
Bitar, S., 1979, Transicibn, Socialismo y Democracia (Siglo XXI, Mexico, D.F.).
Bitar, S., 1986, Chile. Experiment in democracy (Institute for the Study of Human Issues,
Philadelphia, PA).
Bresser Pereira, L.C. and Y. Nakano, 1987, Inertial inflation (Westview Press, Boulder, CO).
Boorstein, E., 1988, Allende’s Chile (International Publishers, New York).
Bruno, M., G. di Tella, R. Dornbusch and S. Fischer, 1988, Inflation stabilization. The
experience of Israel, Argentina, Brazil, Bolivia and Mexico (MIT Press, Cambridge, MA).
Carbonetto, D. et al., 1987, El Peru Heterodoxo: Un Modelo Economico (Instituto National de
Planificacion, Lima).
Centro de Economia Aplicada, 1988, Evaluation de dos adios y medio de Gobierno, Julio 1985-
Diciembre 1987 (CEA, Lima).
Conniff, M., 1982, Latin American populism in comparative perspective (University of New
Mexico Press, Albuquerque, NM).
de Vlyder, S., 1974, Allende’s Chile (Cambridge University Press, Cambridge).
Dornbusch, R., 1982, Stabilization policies in developing countries: What have we learned?,
World Development 10, no. 9 [reprinted in: Dollars, debts and deficits, MIT Press,
Cambridge, MA (1987)].
Dornbusch, R., 1988a, Inflation stabilization: The role of incomes policy and of monetization, in:
Exchange rates and inflation (MIT Press, Cambridge, MA).
Dornbusch, 1988b, Peru on the brink, Challenge, Sept./Ott.
Dornbusch, R. and J.C. de Pablo, 1988, Deuda Externa e Inestabilidad Macroeconomica en
Argentina (Editorial SodAmericana, Buenos Aires).
Dornbusch, R. and S. Edwards, 1989, Macroeconomic populism in Latin America, Working
paper no. 2986 (NBER, Cambridge, MA).
Drake, P., 1982, Conclusion: Requiem for populism?, in: M.L. Conniff, ed., Latin American
populism in comparative perspective (University of New Mexico Press, Albuquerque, NM).
Edwards, S., 1986, Stabilization with liberalization: An evaluation of ten years of Chile’s
experience with free market policies, 1973-1983, in: A. Choksi and D. Papageorgiou, eds.,
Economic liberalization in developing countries (Basil Blackwell. Oxford).
Edwards, S., 1989, The debt crisis and economic adjustment in Latin America, Latin American
Research Review 3, 172-186.
276 R. Dornbusch and S. Edwards, Macroeconomic populism

Edwards, S. and A. Cox-Edwards, 1987, Monetarism and liberalization. The Chilean experiment
(Ballinger, Cambridge, MA).
Falcoff. M.. 1989. Modern Chile 1970-1989 (Transaction Publishers. New Brunswick. NJ).
Ffrench-Davis, R. and E. Tironi, eds., 1974, El Cobre en el Desarrollo Economico National
(Ediciones Nueva Universidad, Santiago).
Foxley, A., 1981, Stabilization policies and their effects on employment and income distribution,
in: W. Cline and S. Weintraub, eds., Economic stabilization in developing countries
(Brookings, Washington, DC).
Foxley. A., 1983, Latin American experiments in neoconservative economics (University of
California Press, Los Angeles).
Foxley, A. and 0. Munoz. 1974. Income redistribution, economic growth and social structure:
The case of Chile, Oxford Bulletin of Economics and Statistics 36, no. 1.
Garcia, Norberto. 1972, Algunos Aspectos de la Politica de Corto Plaza en 1971, in: D. Alaluf et
al.. eds., La Economia Chilena en 1971 (Institute de Economia de la Universidad de Chile,
Santiago).
Geller, L. and J. Estevez, 1972. La Nacionalizacion del Cobre, in: D. Alaluf et al., eds., La
Economia Chilena en 1971 (Instituto de Economia de la Universidad de Chile, Santiago).
Glewwe. P.. 1988, The distribution of welfare in Peru 1985-1986, Mimeo. (World Bank).
Herrera, C. et al.. 1987, Reactivation y Politica Economica Heterodoxa 1985-1986 (Fundacion
Friedrich Ebert, Lima).
Iguiniz. J., 1988. Evaluation Critica de la Politica Economica Bajo Garcia: Balance a 10s Tres
Anos de Gobierno, Mimeo. (Fond0 de Region Andina, Lima).
Inostroza, A., 1979, El Programa Monetario y la Politica de Comercio Exterior de la Unidad
Popular, Panorama Economico. March.
Kuczynski, P.-P., 1977, Peruvian democracy under economic pressure (Princeton University
Press, Princeton, NJ).
Lopez, J., 1972. La Economia Politica de la Unidad Popular: Una Evaluation de su Primer Atio
de Gobierno, in: D. Alaluf et al.. eds., La Economia Chilena en 1971 (Instituto de Economia
de la Universidad de Chile, Santiago).
Martner, G., ed., 1971, El Pensamiento Economico del Gobierno de Allende (Editorial
Universitaria, Santiago).
Ministerio de Economia, 1987. Programa Trienal 198881990 (Ministerio de Economia, Lima).
Nove, A., 1976, The Political economy of the Allende Regime, in: P. O’Brien, ed.. Allende’s Chile
(Praeger, New York).
Novoa. E., 1971, Vias Legales para Avanzar al Socialismo, Revista Mensaje 167.
ODEPLAN, 1971, Objetivos del Plan 1971, in: G. Martner, ed., El Pensamiento Economico del
Gobierno de Allende (Editorial Universitaria, Santiago).
Oppenheim. L., 1989, The Chilean road to socialism revisited, Latin American Research Review
24, no. I, 1555183.
Ortiz de Zevallos, F.. 1989, The Peruvian puzzle (Twentieth Century Fund, New York).
Orrego Vicuna, F., ed., 1975, Chile: The balanced view (Institute of International Affairs,
University of Chile. Santiago).
Pazos, F., 1972, Chronic inflation in Latin America (Praeger, New York).
Presidencia de la Republica, 1986, Plan National de Desarrollo 19861990 (Instituto National
de la Republica, Lima).
Presidencia de la Renublica, 19X7a. Plan National de Desarrollo 1988 (Instituto National de
Planificacion, Lima).
Presidencia de la Reoublica, 1987b. Porque la Estatizacion de1 Sistema Financier0 Peruano
(Instituto Nacionai de Planilicacion. Lima). Aug.
Presidencia de la Republica. 1987~. Diagnostic0 y Programacion 1987 (Instituto National de
Planihcacion, Lima).
Ramos, J., 1980, The economics of hyperstagflation, Journal of Economic Development,
467488.
Ramos, J., 1986, Neoconservative economics in the southern cone of Latin America, 1973-1983
(Johns Hopkins University Press, Baltimore, MD).
Rosenstein-Rodan, P., 1974, Why Allende failed, Challenge, May-June.
 R. Dornbusch and S. Edwards, Macroeconomic populism 211

Sachs, J., 1989, Social conflict and populist policies in Latin America, Working paper
no 2897, March (NBER, Cambridge, MA).
Seers, D., 1964, Inflation and growth: The heart of the controversy. in: W. Baer and I.
Kerstenetzky, eds., Inflation and growth in Latin America (Yale University Press, New
Haven, CT).
Solimano, A., 1988, Politica de Remuneraciones en Chile: Experiencia Pasada, Instrumentos y
Opciones a Futuro, Coleccion Estudios Cieplan, no. 25, Dec. (CIEPLAN, Santiago).
Stallings, B., 1978, Class conflict and economic development in Chile, 1958873 (Stanford
University Press, Stanford, CA).
Stallings, B., 1988, Self-destruction of an auspicious initiative: Peruvian debt policy under Alan
Garcia, Mimeo. (University of Wisconsin, Madison, WI).
Sunkel, O., 1960, Inflation in Chile: An unorthodox approach, International Economic Papers,
no. 10, 107-131.
Thorp, R., 1987. Trends and cycles in the Peruvian economy, in: P. Bardhan et al., eds.,
International trade, investment, macro policies and history (North-Holland, Amsterdam).
Thorp, R. and G. Bertram, 1978, Peru 189@1977 (Columbia University Press, New York).
Unidad Popular, 1969, Programa Basic0 de Gobierno (Editorial Universitaria, Santiago).
Vuskovic. Pedro, 1970, Distribution de1 Ingreso y Opciones de Desarrollo, Cuadernos de la
Realidad National, Sept.
Vuskovic, Pedro, 1973, The economic policy of the Popular Unity government, in: J. A:Zammit,
ed., The Chilean road to socialism (University of Texas Press, Austin, TX).
Wise, C., 1988, Peru in the 1980s: Political responses to the debt crisis, Mime,. (Columbia
University, New York).
World Bank, 1985, Peru. Country economic memorandum (World Bank, Washington, DC).
World Bank, 1988, Peru: Policies to stop hyperinflation and initiate economic recovery (World
Bank, Washington, DC).
Yanez, J., 1978, Una Correction de1 lndice de Precios al Consumidor Durante el Period0 1971-
73, in: F. Contreras. ed., Comentarios Sobre la Situation Economica (Instituto de Economia
de la Universidad de Chile, Santiago).
Zammit, J.A., ed., The Chilean road to socialism (University of Texas Press, Austin, TX).