Abstract

Orthodox stabilization plans in the member countries of the Organization for Economic Cooperation and Development (OECD)—such as the Reagan and Thatcher plans in the early 1980s—have had significant negative distributive and welfare effects. Unemployment and real wage reductions caused by the Phillips curve trade-off between inflation and the level of activity result from stabilization attempts that were based on contractionary monetary policies. The long stabilization in Chile, for example, although mixed with some incomes policy ingredients, had a strong orthodox flavor and was also associated with significant distributive and welfare consequences.

Orthodox stabilization plans in the member countries of the Organization for Economic Cooperation and Development (OECD)—such as the Reagan and Thatcher plans in the early 1980s—have had significant negative distributive and welfare effects. Unemployment and real wage reductions caused by the Phillips curve trade-off between inflation and the level of activity result from stabilization attempts that were based on contractionary monetary policies. The long stabilization in Chile, for example, although mixed with some incomes policy ingredients, had a strong orthodox flavor and was also associated with significant distributive and welfare consequences.

Heterodox experiments in Argentina, Brazil, and Peru during the 1980s, in contrast, had positive distributive consequences: real wages and employment grew during the initial phases of the plans. However, these plans failed to stabilize the economies in which they were applied. The conventional view is that they failed because policymakers neglected to deal with the “fiscal fundamentals” of the inflationary process, but they may have failed because they were associated with distributive shifts that were socially unsustainable. Recent plans in Argentina, Israel, and Mexico seem to combine the best of both types of previous experience: they have been successful in bringing down inflation and at the same time have had either nonnegative or positive consequences on the level of activity and real wages.

This paper examines the extent to which these plans did not have negative distributive and welfare consequences in the context of public finance issues.

Its objective is to explore the relationship between high and chronic inflation, successful and unsuccessful stabilization attempts, and distributive and welfare issues. The last mentioned are discussed in a broad sense, taking into account the behavior of real wages, poverty, and employment as well as the role of the level and structure of tax revenues and of government expenditures.

Unsuccessful stabilization attempts contribute to an inflationary crisis. Successful stabilization plans mark the end of such a crisis. That is why the analysis of distributive and welfare issues in this paper is separated into two basic contexts: that of an inflationary environment, which includes periods of accelerating inflation and of unsuccessful stabilization attempts, and that of successful stabilization. However, it is not true that the stabilization of an economy begins only when the final and successful plan is launched. Fiscal adjustments take time, and a few stabilization attempts may fail while the adjustment is in process. Therefore, the discussion of distributive issues associated with the stabilization period cannot overlook the distributive consequences of the inflationary period in which part of the required adjustments might have taken place. The extent to which distributive shifts should be associated with the inflationary crisis or with the stabilization process is a matter of interpretation.

Background

The origins and dynamics of high and chronic inflation have been analyzed in detail in the literature (see, for example, Bruno, 1993; Kiguel and Liviatan, 1992; and Calvo and Végh, 1992). However, in view of the specific objectives of this paper, a brief reconte, with minor variations, may be of some use.

In the recent Latin American experiences—in Argentina, Brazil, Mexico, and Peru, for example—the external debt crisis of the early 1980s had an important impact on the macroeconomics of those countries and can be seen as the original, cause of the ensuing inflationary crises.

Compared with the facts, a textbook response to the external shock would have led to a different, and certainly less tortuous, story. A real devaluation of the domestic currency would have shifted relative prices in favor of tradables, thus improving the balance of payments; a reduction in absorption would, at the same time, have reduced imports and inflationary pressures; and a more efficient government could have reduced the impact of the external debt on the internal debt where the so-called transference problem was relevant. However, textbook solutions are hard to reproduce because the social, political, and institutional constraints might prove hard to overcome. History is shaped by such constraints, and to disregard them, blaming inefficient, populist, or weak governments, is naive.

In Argentina and Brazil, for example, devaluations were attempted in the context of the redemocratization of national policies—a context in which unions were recovering their bargaining power and in which some degree of wage rigidity was inescapable. To blame the unions is to forget the political history of both countries.1 The dynamics of the exchange rate and wages, with the formation of prices in between, were at the core of the inflationary and distributive processes in those countries. Struggles over wage indexation rules and the institutional context in which wage bargaining takes place, and attempts to introduce concerted income policies and make wage policies more effective are all behind the vicissitudes of the real exchange rate, the real wage, and the path of inflation itself.

The link between the external and domestic debts has also had an important influence on the dynamics of the fiscal debt. Changes in the exchange rate and the interest rate have an immediate impact on government finances, reducing considerably the positive effects of the huge effort to produce primary surpluses. Through the connection between a government’s external and domestic accounts, the lack of capital inflows and the increase in international interest rates over the 1980s greatly diminished the results of fiscal efforts.

This is not to say that the accounts of governments are not affected by political factors. In democratic regimes, and in the processes of democratization in particular, the adjustment needed to reduce fiscal deficits requires complex political struggles. Social demands for public resources increase during periods of democratization, and if the governments in power do not have strong political support, they will find it difficult to increase taxes, reduce expenditures, and implement changes in priorities. Changes in priorities are opposed by conservative forces that supported the government during military periods in Latin America. The conflict between more universal and social demands, on the one hand, and corporatist demands, on the other, makes it very difficult to implement deficit reductions.

The failure to reduce inflation through wage and price freezes creates new dynamics for the inflationary process, with unions and firms trying to anticipate new policy initiatives. Uncertainty and expectations start to play a role. In this new environment, the dynamics of inflation and relative prices become relatively independent of so-called fundamentals. This is when institutions become important: the degree of centralization and synchronization of wage bargaining, the existence of wage and price setters, the role of business associations and central unions in supporting stabilization plans, and the efficacy of government agencies in providing price guidelines and implementing price controls all play an important role in making the stabilization effort work.

When nothing else works, the government is left with the interest rate as the only instrument to prevent price explosions. If, on the one hand, the policy prevents capital flight into real assets, on the other, it leads to very significant changes in the size of the domestic debt as well as in transfers of wealth from the government to holders of the treasury debt. In recent periods, high interest rates have attracted capital inflows that increase the stock of reserves and help prevent speculative attacks but at the same time fuel the increase in government debt because the central bank, to avoid monetizing the inflow of capital, issues treasury bonds.

Distributive Effects of Inflation

The distributive effects of the inflationary process are associated, in part, with the reduction in real wages as inflation accelerates.2 The degree of wage indexation to past inflation and the time between adjustments help explain the path of wages. It is easy to show that the average real wage over a year (Wt) depends on the peak real wage (w−1p), the degree of wages of indexation to past inflation (m), the rate of inflation over the year (Pt), and the length of the indexation period (d) (see Ros, 1989 and Amadeo, 1994):
wt=wp1(1+mp11+p1)(1(1+pt)d).

Given the rate of inflation, the shorter the indexation period and the greater the degree of indexation, the greater the real wage. An acceleration of inflation, all other things being equal, leads to a reduction in the real wage. That is why in economies with accelerating inflation, it is usual to see unions demanding a shortening of the indexation period. In these circumstances, if firms are able to keep their markup constant, the result will be a further acceleration of the rate of inflation. Vicious circles involving the rate of inflation, the degree of wage indexation, and the length of the adjustment period are part of the inflationary history of many Latin American countries and of Israel in the 1980s.

There are two instances in which wage indexation becomes a cornerstone of adjustment or stabilization efforts. In countries hit by external shocks, wage indexation eventually makes real devaluations difficult. Attempts to devalue in a situation involving wage indexation give rise to accelerated inflation. The other instance has to do with the rigidity imposed on the rate of inflation by backward-looking indexation in stabilization plans.

The distributive conflict between the exchange rate and the real wage is part of the perennial inflation story of some Latin American countries. As argued in the section on the distribution and demand for stabilization (pp. 69–70), resolving this conflict or alleviating it through trade deficits is at the core of many examples of successful and unsuccessful stabilization experiences.

Except for Israel, where real wages grew almost continuously before and after the successful stabilization plan of 1985, in all other economies with recent episodes of high inflation, indexation has never been able to protect wages against creeping inflation. Notwithstanding the natural tendency toward greater indexation and a shortening of the adjustment period, real wages in the industrial sector fell over the 1980s in Argentina (after 1984), Brazil (after 1986), Chile (until 1987), Mexico (until 1989), and Peru.

Devaluations, coupled with recession or restrictive wage policies, were instrumental in reducing wages in Argentina (1981–82), Brazil (1981–84 and 1990–93), Chile (1982–87), and Peru (1983–86). The political control over unions explains the continual reduction in wages in Brazil (1964–74), Chile (1970s), and Mexico (1980s). Hyperinflation explains the reduction in wages in Argentina in 1989 and 1990.

The trajectory of the minimum wage, which serves as a standard for the wages of the less organized workers and is usually determined unilaterally by the government, is even more dramatic. Over the 1980s, the real minimum wage fell by 30 percent in Argentina, 40 percent in Brazil, 30 percent until 1987 in Chile, 58 percent in Mexico, and more than 80 percent in Peru (Table 3.1).

Table 3.1.

Selected Latin American Countries: Real Minimum Wages

(1980 = 100)

article image
Source: Marinakis (1993).

Preliminary figures.

Two questions should be addressed in the context of the relationship between the path of real wages and inflation. First, to what extent should factors such as sluggish growth or external shocks, rather than inflation itself, be responsible for the reduction in wages? Second, to what extent does the reduction in wages necessarily improve macroeconomic performance?

It should be clear that it is almost impossible to separate the macroeconomic crisis of the 1980s from the inflationary process in most Latin American countries. The reduction in real wages was a deliberate policy measure to reduce absorption and increase the profitability of exports, thereby reducing the trade deficit in most Latin American countries in the early 1980s. It was relatively successful in this respect, but, at the same time, devaluations induced greater indexation of wages, thereby fueling inflation and creating the vicious circle mentioned above.

In stabilization plans, depending on the balance of political power and the degree of consensus about the appropriateness of the policy, it is possible to have wages carrying part of the adjustment burden. Consensual, or negotiated, wage restraint did not play a role in Latin America, although in Israel in 1985, it was an important element of the stabilization plan. In Latin America, there were situations under authoritarian governments in which the wage policy was deliberately used to curb inflation.

In a significant number of circumstances, however, the reduction in real wages is no more than an accommodation to the socioeconomic crisis. It is not a response to policy impulses; nor does it necessarily lead to macroeconomic improvement. The reduction in wages may be associated with a distributive shift owing to changes in the bargaining power of the agents in labor and goods markets. Episodes of overindexation of prices in anticipation of policy actions were common after a series of unsuccessful stabilization attempts in Argentina and Brazil. When inflation accelerates in this way, the reduction in wages does not improve macroeconomic performance. The effect may even be negative, to the extent that the wage bill is a major component of aggregate demand and thus has an important effect on the level of activity.

The prolonged inflationary crisis in some Latin American countries is at the root of the almost continual reduction in real wages. Such a crisis is characterized by repeated, unsuccessful attempts to adjust to external shocks and reduce inflation and by episodes of pre-emptive or anticipated inflationary hikes, all of which lead to the erosion of real wages in a sluggish economy. In sum, reductions in real wages make a strong redistributive dent. They may help economies adjust to external shocks or they may result in sluggish economic conditions. A policy-induced underindexation of wages helps curb inflation. Finally, reductions in real wages with distributive effects may result from endogenous inflationary spurs, and in such cases are either without purpose from the standpoint of macroeconomic performance or have negative effects.

The distributive effects of inflation also depend on the extent to which the poor pay the inflation tax on idle cash balances, and the rich benefit from high interest rates on government bonds. Cardoso (1992, p. 5) argues that middle-class families usually have their savings wiped out by the inflation tax. This, of course, is true only if savings are not indexed to inflation and if the real interest rate on savings is negative. If interest rates are positive, the inflation tax does not apply to savings. High interest rates are not an exception in countries with chronic inflation. On the contrary, to avoid capital flight into real assets, interest rates are usually kept very high. High interest rates not only have a negative effect on the effort to reduce the domestic deficit, but also have an important distributive effect to the extent that the holders of government bonds are banks, large firms, and rich families.

High interest rates reduce the demand of durables, establish a floor for profit margins and inhibit production, thus creating the conditions for stagflation. For firms, the higher the short-run rate of interest, the greater the profit margin required to induce production. In contrast, because inventories are kept very low, any time demand increases, excess demand develops, creating inflationary pressures. Hence, if, on the one hand, tight monetary policy avoids speculative attacks and hyperinflationary episodes, then, on the other hand, it has stagflationary side effects. Although government creditors obviously benefit from high interest rates, wage earners suffer the effects of the tight monetary policy on aggregate demand, employment, and real wages.

Welfare Effects of Successful Stabilizations

The main subject of this section is the successful stabilization programs implemented in Argentina (1991), Israel (1985), and Mexico (1988). However, before the welfare effects of these recent successful plans are analyzed, a few general observations based on Table 3.2 are presented.

Table 3.2.

Stabilization Plans

article image

Based on money anchors.

Based on different nominal anchors.

First, there are cases of successful stabilizations both in the orthodox and in the incomes policy columns. The role of incomes policy ingredients is relevant, particularly in cases of chronic inflation, in which the dynamics of prices become relatively independent of “fundamentals.” Second, although the recent plans in Argentina, Israel, and Mexico had important incomes policy elements, substantial fiscal and monetary restrictions preceded and followed the plans. Finally, when generally successful orthodox and incomes policy plans are compared, the former usually had worse distributive and welfare ratings than the latter. The same is true of unsuccessful orthodox and incomes policy attempts to stabilize, although the differences are not relevant to this discussion. The immediate distributive and welfare effects of the successful plans in Argentina, Israel, and Mexico were far from negative, considering the behavior of real wages, unemployment, and per capita consumption.

In Argentina, after almost a decade of negative rates of growth in total GDP and manufacturing GDP, both grew strongly after the convertibility plan of 1991 was implemented. Real wages in manufacturing fell continuously after 1985 and did not begin to recover until 1992 (Table 3.3).

Table 3.3.

Argentina: Macroeconomic Indicators

(In percent)

article image
Source: Inter-American Development Bank (1993).

In Israel, real wages, GDP, and employment grew faster, on average, in the years following the implementation of the stabilization plan (1986–91) than in the years characterized by high inflation that preceded the plan (1981–85) (Table 3.4). The average rate of unemployment was higher in the post-plan period, but this is, at least in part, the result of the massive immigration of Russians. Per capita consumption rose by about 20 percent between the first half of 1985 and the first half of 1987. The average annual rate of increase in consumption was 11.8 percent in 1986–87, 2.1 percent in 1988–89, 5.3 percent in 1990, and 7.6 percent in 1991.

Table 3.4.

Israel: Macroeconomic Indicators

(In percent)

article image
Source: Bruno (1993, p. 110).

In Mexico, the debt crisis and the period of fiscal adjustment were characterized by a reduction in GDP and investments, and a 27 percent reduction in real wages, compared with the golden years of 1977–80. After 1988, with the solidarity pact between government, trade unions, and businesses, investment and GDP recovered, and, after falling by 35 percent between 1978 and 1988, real wages grew by 14 percent between 1988 and 1991 (Table 3.5).

Table 3.5.

Mexico: Macroeconomic Indicators

(In percent)

article image
Source: Bruno (1993, p. 193).

Exchange-rate-based stabilization plans in Argentina, Israel, and Mexico were accompanied by an increase in the level of activity and real wages. In open economies, the real wage depends on the price of both tradables and nontradables. If, in an extreme case, only the exchange rate is under-indexed in relation to inflation, the appreciation of the real exchange rate will imply a reduction in the relative price of imported tradables and an increase in the real income of wage earners. As a result, there will be an increase in consumption, demand, and imports, leading to a deterioration of the trade balance. On the supply side, the reduction in the cost of production resulting from the fall in the domestic price of imported inputs will also induce an expansion of output, thus reinforcing the expansion of aggregate demand. The expansion in the level of activity will lead to an increase in the demand for labor, reducing unemployment and putting pressure on the real wage. In the experiences of Argentina, Israel, and Mexico, this account seems to be a plausible explanation of macroeconomic developments during the exchange-rate-based programs.3

In sum, comparing the behavior of GDP, employment, and real wages in the periods immediately preceding and following the stabilization effort in all three countries, stabilization had neutral or positive effects. The figures for real wages in Argentina are still very modest, but, in comparison with previous years, the situation did not worsen. Therefore, these three experiences did not duplicate the negative distributive and welfare performance of the “classical” orthodox stabilization programs.

However, the analysis of the distributive effects of stabilization efforts cannot be restricted to the movement of real wages and the level of activity, although this is certainly an important factor. The other factor is public finances, that is, the tax structure as it affects government receipts, the distribution of the tax burden through progressive forms of taxation, and the apportionment of government expenses among different programs. Because stabilization efforts are associated with the capacity to produce fiscal surpluses, the distribution of the tax burden and expenditure reduction have obvious welfare and distributive effects.

In this connection, perhaps the design of a “social welfare function” to study the distributive and welfare effects of stabilization should not give equal weight to government expenses accruing to the rich and those devolving to the poor. In an environment in which the incidence of poverty is high and the poor suffer from severe credit constraints, the impact of government expenses on the living standards of the poor is far from negligible. In such a context, uniformly distributed budget cuts, or cuts that reduce the share of social expenditures, will certainly have significant negative welfare, and even efficiency, consequences.

The increase in international interest rates in the early 1980s doubled the external debt of highly indebted Latin American countries. The inflation crisis that ensued further deteriorated public finances. The Oliveira-Tanzi effect reduces tax revenues. The internal transfer problem indicates that the size of the domestic deficit is linked to movements of the exchange rate and to the balance of payments, with devaluations and trade surpluses having a positive effect on the domestic debt.

Successful stabilization plans have been associated with major fiscal restrictions. As seen in Table 3.6, the significant increase in the interest rates on the domestic and external debts after 1981 resulted in an enormous effort by all countries to reduce the level of the public sector borrowing requirement. Current expenditures suffered major cuts in order to increase public sector saving, which, in Chile and Mexico, reached 10 percent of GNP in certain years. Public investments had to be reduced to diminish borrowing requirements.

Table 3.6.

Public Sector

(In percent of GNP)

article image
Sources: Argentina (Chisari and others, 1993), Brazil (Carneiro and Werneck, 1993), and Chile and Mexico (Amadeo, 1993).

The budget cuts associated with the fiscal effort led to significant reductions in the share of social expenditures in shrinking GNPs. Table 3.7 shows heavy cuts in social budgets for all countries except Brazil, where the per capita share of social expenditures in relation to GNP increased between 1979–85 and 1986–88. Figures reported by the UN Economic Commission for Latin America and the Caribbean show that the share of social expenditures in relation to GNP fell 14 percent in Argentina, 25 percent in Bolivia, and more than 30 percent in Mexico between 1979–81 and 1986–88. Per capita expenditures in education fell by 26 percent in Argentina, 47 percent in Bolivia, and 27 percent in Mexico in the same period.

Table 3.7.

Social and Education Expenditures per Capita

(In percent of GNP; 1979–81 = 100)

article image
Source: United Nations, Economic Commission for Latin America and the Caribbean (1992).

The distributive and welfare effects of such budget cuts cannot be overemphasized. Poor families are the main recipients of social services. It is a mistake to assume that budget cuts are uniformly distributed across programs and subprograms. The extent of the cuts is determined by the political leverage of the groups affected, and, again, the ability of the poor to resist budget cuts that have a direct effect on their welfare is negligible.

When the tax structures in Latin America, Southeast Asia, and OECD member countries are compared, it can be seen that, first, the tax burden in Latin America and Southeast Asia is around one-half that of the OECD countries and, second, that the share of direct taxes in GNP in Latin America is half that of Southeast Asia and about one-fourth of the share in OECD countries (Table 3.8). A “progressive” increase in the tax burden in Latin America would clearly mean an increase in direct taxes on income and wealth. Had such a tax reform been introduced together with the stabilization effort, the negative distributive and welfare effects of the fiscal effort would have been considerably smaller. Again, as in the case of budget cuts, the political aspect of tax reform is complex.

Table 3.8.

Tax Burden in Selected Regions

(Ratio of taxes to GNP in percent; average 1987–89)

article image
Source: United Nations, Economic Commission for Latin America and the Caribbean (1992).

Budget cuts with significant distributive effects may have an influence on the long-run efficiency of the economy. It can be argued that the decrease in the education, health, and other social budgets reduces the overall productivity of a country in the long run, not just the productivity of those workers directly affected. To the extent that educated and less educated workers perform complementary tasks in the economy, a reduction in the productivity of the latter will eventually hinder the productivity of the former and, by extension, will affect the economy and society as a whole. If this argument makes sense, budgetary cuts aimed at reducing deficits should be accompanied by redistributive measures both in the tax structure and in the structure of government expenditures in order to enhance equity and long-run efficiency.

Distributive Conflict and Exchange-Rate-Based Stabilizations

Successful and unsuccessful stabilization plans are usually associated with significant distributive shifts. In the case of chronic inflation, in which the formulation of prices and wages is relatively independent of fiscal fundamentals, the coordination of disinflation is difficult. Repeated, unsuccessful attempts to stabilize or the fear of firms or unions concerning price or wage losses lead to important price and wage rigidities. Such rigidities translate into “inflationary residuals” when a stabilization plan is in place. When coordination difficulties arise in promoting a “uniform deindexation,” policymakers are tempted to choose one (or more) key prices to “lead the way.” The already-mentioned Brazilian plan of 1964–67 is known for underindexing wages in order to promote stabilization. The exchange rate has also been used in many countries to guide the stabilization process. The fact is that the underindexation of one or more major prices in an environment of pervasive indexation helps promote stabilization.

Changes in relative prices affect economic agents in different ways. If only wages are underindexed, real wages fall and wage earners are hurt. If only the exchange rate is underindexed, the real exchange appreciates and exporters are hurt. In stabilization plans based on underindexation, the distributive conflict involving relative prices is a decisive element in the success of stabilization attempts.

It is a mistake to think that the only factor behind the failure of the Cruzado Plan in Brazil (1986) was the lack of a fiscal adjustment. The increase in real wages fueled by the expansion of the level of activity gradually began to squeeze profit margins in sectors where the price freeze was effective. These were precisely the oligopolistic sectors that, before the stabilization plan, were free to mark up costs but that, during the plan, were unable to raise prices. Hence, in the Cruzado Plan, the prices of the 18 oligopolistic sectors (and the exchange rate) were underindexed and, indeed, played the role of the main anchors of the plan. The change in relative prices implied by the dynamics of the Cruzado Plan was unacceptable to the large industrial and commercial firms, which demanded the relaxation of the price freeze.

The distributive conflict under stabilization plans based on indexation can be “suppressed” depending on the bargaining power of the agents involved. In democratic systems, a hegemonic political coalition may legitimize the underindexation of certain prices. Exporters may have to carry the burden of the stabilization in exchange-rate-based stabilizations if the majority favors such a plan. In authoritarian systems, the government can enforce the underindexation of certain prices and blue-collar workers may carry the burden.

Distributive conflicts can also be avoided, as they can under exchange-rate-based stabilization plans where capital inflows play the role of adjustment variables. The distributive shifts in the Southern Cone exchange-rate-based stabilization attempts in the late 1970s and early 1980s were not sustainable because the trade deficits resulting from the appreciation of the exchange rate were intolerable. Major distributive shifts and trade deficits—especially in Mexico (see Table 3.9)—are at the core of the success of recent stabilization plans based on exchange rate anchors. Huge capital inflows are required to equilibrate the balance of payments.

Table 3.9.

Argentina and Mexico: Selected Economic Indicators

article image
Source: Inter-American Development Bank (1993).

Capital inflows in Argentina increased from an average of less than US$2 billion between 1985 and 1990 to US$5.8 billion in 1991 and US$13 billion in 1992. In Mexico, they rose from an average of negative US$0.2 billion between 1985 and 1989 to US$20.4 billion in 1991 and US$26 billion in 1992.

It is likely that inflationary residuals in the recent exchange-rate-based plans were smaller than in the earlier Southern Cone plans—because wages and prices were not completely free in the former—and this obviously helped strengthen the program. But it is also true that the possibility of bypassing the distributive conflict between real wages and the real exchange rate through a massive inflow of capital helps sustain the plan. In this context, the long-run effects on the competitiveness of Argentine and Mexican industries, on the one hand, and the sustainability of capital flows into the two economies, on the other, are two important issues to keep in mind.

Distribution and the Demand for Stabilization

The recent literature on the political economy of inflation and stabilization (for example, Alesina and Drazen, 1991) emphasizes the distributive conflict between different social groups before and after stabilization programs are launched. Stabilization is delayed as long as the costs of inflation (associated with the existence of fiscal deficits and “distortionary finance”) to each group are perceived to be greater (owing to tax increases or budget cuts) than the costs associated with the stabilization plan.

Alesina and Drazen argue that there may be an “agreement on the need for a fiscal change but a political stalemate over how the burden of higher taxes or expenditure cuts should be allocated” (p. 1172); and that “when stabilization occurs, it coincides with a political consolidation.” Furthermore, they maintain, one side becomes politically dominant. The burden of stabilization is sometimes quite unequal, with the politically weaker group bearing a larger burden. Often this means the lower classes, with the burden of a successful stabilization being regressive” (p. 1173). The stalemate to which Alesina and Drazen refer makes sense in relatively homogeneous societies in which the level of information and bargaining power of the groups involved in the “war of attrition” are not unevenly distributed. Political representation in OECD countries is accorded to all citizens, which makes the distributive conflict more transparent and balanced. In Latin American countries, in contrast, the upper and middle classes have considerably more access to information and political leverage than the lower classes. Hence, the distributive battle involves a relatively smaller segment of society, and a disproportionate share of the burden of the adjustment falls on the least organized members of society.

Stabilization has been delayed in Latin America, in part, because there are internal conflicts among the most elite members of society. Stabilization is also delayed because the demand for it is weak. As seen above, the hardship of inflationary and stabilization processes falls disproportionally on the poor, who are less well organized and cannot voice their dissatisfaction. The agents who can protect their income and wealth against inflation do not have the incentives to demand stabilization actions or to act to sustain stabilization attempts. The technologies that permit an economy to function in spite of inflation and the ability of certain groups to protect themselves against adjustment costs therefore give rise to prolonged periods of inflation.

Conclusion

Distributive issues are at the core of inflationary and stabilization processes. They are associated not only with the path of real wages and other key relative prices, but also with the structure of taxes and government expenses. It has been shown that the “politically weak” tend to lose before successful stabilizations programs begin because (1) backward-looking wage indexation protects wages imperfectly, (2) the poor do not have access to indexed financial assets, (3) the poor do not have the political leverage to make their interests represented in budgetary disputes, and (4) high interest rates imply a major redistribution of wealth from the government to its creditors.

In Latin America, it has been shown that the politically weak lose under stabilization programs because of the major budget cuts in social expenses. Finally, the tax structure in these countries has not changed in recent years to alleviate and compensate for the burden of the stabilization effort that has primarily affected the poor. That recent stabilization plans based on exchange rate and other nominal anchors did not lead to real wage repression or unemployment hides the fact that the period preceding stabilization was associated with major distributive shifts against the poor. In other words, it can be argued that the effects of the adjustment are felt before the stabilization plan is fully implemented. Perhaps the stabilization effort would have failed if such adjustments had not taken place.

Hence, to argue that recent exchange-rate-based stabilizations in Latin America have been socially harmless is to forget that a redistribution of income and wealth took place before all elements of the stabilization program were in place. Moreover, the welfare consequences of budget cuts were not uniformly distributed between the rich and the poor. This is not to deny the importance of political conflicts over the distribution of income and government resources. Institutional and political factors, the history of relations between the state and society, and the conflict between the elite members of society and the newly empowered social groups in the wave of recent democratization processes in many Latin American countries complicate the solution to macroeconomic problems. To say that stabilization plans fail because they lack credibility or because of the populist attitude of governments oversimplifies the complexities associated with these factors and therefore cannot account for economic developments in the region.

Comment

Ricardo Ffrench-Davis

In his paper, Edward Amadeo provides a perceptive look at a broad range of relevant issues concerning the relationship between price stabilization programs and social welfare and income distribution.

The usefulness of his analysis lies from the outset in the fact that the choice discussed is not whether or not to stabilize, but rather when and how. Amadeo illustrates his article with many examples of the various “whens” and “hows” of stabilization. His study shows that the effectiveness and equity of the adjustment can be significantly affected by the whens and hows.

I will focus in this comment on a few aspects that, in my view, require greater attention.

Measuring the Effects

It is clear that measuring the effects cannot be limited to a comparison of situations with and without stabilization because to do so ignores the two essential questions: when and how. Furthermore, measuring the effects should not be limited to a comparison of situations before stabilization has begun and after it has been completed because this ignores both the characteristics of the initial economic situation and the evolution during the adjustment.

The results depend on the characteristics of the initial economic situation: it is essential to control for variables such as actual versus “equilibrium” wages (W) and exchange rates (ER), fiscal institutionality, and the relationship between productive capacity (PC) and effective GDP. For example, if there is a significant gap between PC and GDP, it is easier to harmonize stabilization with growth and an improvement in the standard of living, given that production, employment, and consumption can be increased without more investment. It is the ex post efficiency or productivity that increases as the economy moves toward the production frontier. This was the situation in Chile in 1975–80 and 1983–88, as Chart 3.1 illustrates.

Chart 3.1.
Chart 3.1.

Chile: Effective and Potential GDP, 1950–89

(In billions of 1977 U.S. dollars)

Sources: International Monetary Fund (1992, 1993). (Information for 1989 is unavailable.)

In addition, to measure welfare, it is important to examine what happens with, among other variables, employment, W, and GDP paths during the transition. The present value of the integral of the process of adjustment indicates what has happened to the welfare and wealth of the various agents until a new equilibrium is reached. The evolution of investment in human and physical capital also indicates how productive capacity may develop once the stabilization program has ended. It has a strong impact on the future level of employment and its productivity, and, consequently, on the ability of the economy to sustain rising wages. In reality, the hysteresis of the adjustment process affects both the interpretation of the past as well as projections for the future.

Removal or Creation of Obstacles to Growth with Equity

The mix of policies can lead to the removal of obstacles to sustainable development but can also create new distortions or obstacles. The latter outcome is not unusual, particularly when price stability as such is given absolute priority or is assumed to be the principal ingredient for spontaneous economic growth. If, as a result of this kind of approach, stabilization is achieved at whatever cost, the result may be stabilization with stagnation or short-term stabilization followed by instability.

For stabilization programs to be lasting and to contribute to development (as they undoubtedly do when they are well designed), the hysteresis of the adjustment process must be closely monitored: How are the basic ingredients of growth (for example, investment, training, innovation, and technology) being managed? And how do they affect capacity and the opportunities available to the various social sectors? Stabilization programs can have progressive or regressive results, depending on their quality.

Single Anchors

The results of programs that depend on a single variable to lead the stabilization process are generally not very satisfactory: they tend to be short lived and contribute little to economic growth. The two single anchors most often used are the fixing of the exchange rate and isolated control of the money supply.

Fixing the nominal exchange rate helps slow down inflation in the short term but tends to do so more strongly and rapidly for tradables, while the prices of nontradables react more slowly. The typical result is a significant exchange rate appreciation, as was the case in Argentina and Chile in 1978–81. Both countries were successful in reducing inflation, even achieving negative rates (Chile in 1982), but with an unsustainable distortion in the external sector.

For economic or political reasons, creating imbalances such as those described above may at times be unavoidable. The key then is to be aware of them and to provide an effective way out before the correction is too late and traumatic.

Something similar happens if the single anchor is the money supply. Except in cases of incipient inflation in economies with no indexation, total dependence on tight money tends to generate prolonged underutilization of productive capacity, even when such capacity is high initially. This was the case in Chile between 1974 and 1976. Inertial inflation operating in the aggregate supply pushed prices upward for a period of time although unemployment and underutilization of domestic productive capacity grew substantially.

Depending exclusively, or heavily, on a single monetary variable tends to lead to very high real interest rates (overshooting). It excessively rewards the financial side at the expense of the productive side and tends to place the economy below the production frontier. For both reasons, it discourages capital formation and the creation of productive jobs.

Multiple Anchors

To be successful, that is, to achieve sustainable stability with growth and equity, it is imperative to coordinate a series of variables or prices in order to avoid overlooking some or overemphasizing others: public and private wages, the exchange rate, utility rates, the money supply, interest rates, fiscal balance, and expectations of private price setters. Social cooperation efforts involving the more relevant economic sectors, such as those made in Mexico in 1987 and in Chile since 1990, while limited in scope, can be useful in achieving more sustainable balances with a higher effective productivity (associated with operating closer to the production frontier).

Macroeconomic Policies Designed to Place the Economy in the Production Frontier

One of the basic macroeconomic balances is the capacity utilization rate. Latin America was operating considerably below the production frontier during the 1980s, as a result of external restrictions. Chart 3.1, based on data for Chile, provides a perfect example.

A significant gap between utilization and capacity seriously discourages capital formation and the level of productive employment. In the 1990s, to a large extent because of the restoration of capital flows, there was a moderate economic recovery. In 1991–94, in fact, growth of per capita GDP once again exceeded population growth, with about one-third of GDP growth corresponding to use of capacity.

During the recovery stage in this decade, with the disappearance of the external restrictions that dominated in the 1980s, management of macroeconomic policy has been less demanding than it will be when the production frontier is reached. When this frontier is reached, any new aggregate demand and output to match it will require new investment to support it. Consequently, to sustain even the current moderate levels of growth (3.5 percent in 1991–94), investment must exceed the current level. To return to the average rate of growth of 5.3 percent experienced in Latin America from 1950 to 1981, tens of billions of additional dollars in capital formation would be required each year. It should be noted, underlining the intensity of that challenge, that the flow of foreign direct investment totaled some $15 billion in 1993 (approximately one-fourth of total capital flows).

The other important point is that when the production frontier is reached, more active policies will be required to regulate aggregate demand. If these are not applied, internal or external shocks (changes in interest rates, terms of trade, and availability of capital) will place inflationary or recessionary pressures on the economy. The result is a lower average net utilization of productive capacity, with the consequent negative impact on effective productivity (ex post) and discouragement of investment.

Both effects generate a decline in social welfare and lower productive employment (and/or lower domestic wages). This contrast is seen, in the industrial world, for example, between the automatic adjustments during parts of the nineteenth century and in the 1920s versus those of the years 1950–80. In the last period, the capacity utilization rate (or proximity to the frontier) was much higher, as was the rate of investment. The result was per capita GDP growth two to three times higher in 1950–80 than in the other periods (as well as a significantly greater improvement in social welfare). The shortcoming (not an input for growth but an undesirable side effect of increasing costs) was higher inflation, which culminated in the high levels of the 1970s (significantly higher in the United States than in Germany or Japan).

It is essential now to perfect the ability to design macroeconomic policies that harmonize the economy’s proximity to the production frontier with price stability.

It is clear that the national macroeconomy does not regulate itself spontaneously. Until there is a world central bank, as well, I would emphasize, as world welfare and finance ministries, national policies must be perfected. Their ever-declining ambit, be it monetary, fiscal, or exchange rate policies, must be broadened, and cooperation between the public and private sectors (management and labor) must be strengthened.

Comment

Carlos Noriega

The 1980s and early 1990s have been rich in terms of stabilization experiences, and, as a result, a consensus has gradually emerged among economists on how to proceed to eradicate inflation. However, there has not been a parallel advance in the analysis of the welfare and distributive effects of stabilization processes. Therefore, Amadeo’s paper, by focusing on these issues, enlightens us and advances our knowledge of inflationary episodes. Indeed, unless and until those issues are well understood, arguments favoring stabilization policies should be viewed with a certain degree of skepticism. This paper also serves as a stepping-stone for distributive and welfare studies by reviewing the impact in this context of the most common policy measures of a stabilization program.

The analysis in the paper, however, fails to take into account methodological considerations, thus weakening its conclusions. Let me refer first to some identification problems. At the outset, it must be acknowledged that the existence of inflation already denotes a previous macroeconomic disequilibrium. Therefore, the analysis should begin by measuring the impact of the original measures or shocks that led to the macroeconomic disequilibrium and that derive from inflation itself (already a difficult feat). Only then can one proceed to identify the effects stemming solely from the stabilization programs. Unless this point is addressed explicitly, it is difficult to give much weight to conclusions based only on the timing of the stabilization program and the ensuing evolution of distributive and welfare indicators.

A related issue is that the disinflation experiences analyzed by Amadeo were part of more comprehensive structural adjustment programs, and, therefore, the effects on distribution and welfare cannot be attributed only to the stabilization part of the programs. Furthermore, because the efficacy of stabilization depends on whether the program is supported by structural measures, disentangling their effects may be practically impossible without a more formal model.

Amadeo touches upon these issues tangentially when he states that “the discussion of distributive issues associated with the stabilization period cannot overlook the distributive consequences of the inflationary period in which part of the required adjustments might have taken place. The extent to which distributive shifts should be associated with the inflationary crisis or with the stabilization process is a matter of interpretation.”

The identification problem is relevant not only for measuring the effects of stabilization programs but also for establishing the issues to be analyzed. In this respect, welfare costs of stabilization across countries were not independent of such issues as the previous existence of a debt overhang problem, over- or undervaluation of the exchange rate, and, very important, the extent to which real wages had deviated in both directions from their equilibrium paths. In this context, little can be said about the comparative failures or successes of the stabilization programs surveyed in the paper.

Another set of issues related to the methodology employed in the paper is associated with the measurement of the distributive and welfare effects. Because stabilization processes do not occur overnight, it is difficult to determine the lag structure of the impact. It is interesting to note that even in the “successful” stabilization experiences, such as those of Israel (1985–89) and Mexico (1988- ), after an initial surge in the rate of growth of real GDP, a decline took place in later years. Real wages behave similarly. Thus, it might still be too early to assess the final impact of stabilization. The paper also lacks a discussion of the merits of the indicators used to measure the effects under study. Furthermore, the use of different indicators in the analysis prevents a coherent picture of these issues from emerging. Also, one could think of other potentially useful indicators that were not taken into account, such as the sources of income from national accounts statistics.

Turning now to a different topic, Amadeo raises a relevant question on the extent to which a reduction in wages necessarily improves macroeconomic performance. Although he responds that in a nonnegligible number of circumstances, reducing wages is not a response to policy impulses, nor does it necessarily lead to any macroeconomic improvements, the question deserves more in-depth consideration. Heterodox programs have attached great importance to nominal anchors, including nominal wages. Thus, besides reflecting shifts in bargaining power of the agents in the labor and goods markets, wage policy does, in the medium and longer term, seem to affect the final outcome of stabilization. In that context, real wage reductions may bear importantly on the speed of convergence to macroeconomic stability and ultimately on welfare gains.

The question could have been posed in the inverse. Is a reduction in real rents to capital necessary for improving macroeconomic performance? In answering the question from the perspective taken by Amadeo, it is worthwhile considering that the stabilization experiences under analysis generally took place together with significant trade liberalization. There is a presumption that in developing countries, trade protection generally served to increase the rate of return to capital at the expense of a reduction in the return to labor. Therefore, stabilization programs, in addition to profiting from the discipline that the international price level imposes on domestic prices, also play a critical role in reversing the bias in favor of returns to capital. In this context, one may hypothesize that price stabilization, by reducing in the longer run the average rate of return to capital without inhibiting productive investment, strongly contributes to a better distribution of income.

One final comment refers to the view of inflation as a tax. The theory of public finances would assert that, in the absence of other distortions, the inflationary tax entails a net welfare loss. It may thus be presumed that if that tax is eliminated, society as a whole will gain. In our imperfect world (which is the one Amadeo considers, for well-founded reasons), that result does not always hold, but this line of thought does set a guideline to measure welfare gains of disinflation.

Although my comments may be seen as highly critical of the paper, in fact, they only point to an obvious shortcoming of this type of analysis, which is the lack of a consistent model for measuring welfare and distribution effects. However, had Amadeo stopped in the face of this challenge, we would have been deprived of the opportunity to ponder these important issues.

References

  • Alesina, Alberto, and Allan Drazen “Why Are Stabilizations Delayed?” AmericanEconomic Review, Vol. 81 (December 1991), pp. 1170 –88.

  • Amadeo, Edward, “Adjustment, Stabilization, and Investment Performance: Chile, Mexico and Bolivia,” in Savings and Investment Requirements for the Resumptionof Growth in Latin America, ed. by E. Bacha (Washington: Inter-American Development Bank; Baltimore, Maryland: Distributed by Johns Hopkins University Press, 1993), pp. 161 –207.

    • Search Google Scholar
    • Export Citation
  • Amadeo, Edward, Institutions, Inflation, and Unemployment (Aldershot, Hants, England; Brookfield, Vermont: E. Elgar, 1994).

  • Amadeo, Edward, and Tariq Banuri, “Policy, Governance, and the Management of Conflict,” in Economic Liberalization: No Panacea: The Experiences of Latin Americaand Asia, ed. by T. Banuri (Oxford: Clarendon Press, 1991), pp. 29 –55.

    • Search Google Scholar
    • Export Citation
  • Bruno, Michael, Crisis, Stabilization, and Economic Reform: Therapy by Consensus (Oxford: Clarendon Press, 1993).

  • Calvo, Guillermo, and Carlos Végh, “Inflation Stabilization and Nominal Anchors,” IMF Paper on Policy Analysis and Assessment 92/4 (Washington: International Monetary Fund, 1992).

    • Search Google Scholar
    • Export Citation
  • Cardoso, Eliana, “Inflation and Poverty,” NBER Working Paper No. 4006 (Cambridge, Massachusetts: National Bureau of Economic Research, 1992).

    • Search Google Scholar
    • Export Citation
  • Carneiro, Dionísio, and Rogério Werneck, “Obstacles to Investment Resumption in Brazil,” in Savings and Investment Requirements for the Resumption ofGrowth in Latin America, ed. by E. Bacha (Washington: Inter-American Development Bank; Baltimore, Maryland: Distributed by Johns Hopkins University Press, 1993), pp. 57 –102.

    • Search Google Scholar
    • Export Citation
  • Chisari, Omar, José M. Fanelli, Roberto Frenkel, and Guillermo Rozenwurcel, “Argentina and the Role of Fiscal Accounts,” in Savings and Investment Requirementsfor the Resumption of Growth in Latin America, ed. by E. Bacha (Washington: Inter-American Development Bank; Baltimore, Maryland: Distributed by Johns Hopkins University Press, 1993), pp. 1 –56.

    • Search Google Scholar
    • Export Citation
  • Inter-American Development Bank, Situación Económica y Perspectivas de losPaises de América Latina y El Caribe (Unpublished; Washington: Inter-American Development Bank, 1993).

    • Search Google Scholar
    • Export Citation
  • Kiguel, Miguel, and Nissan Liviatan, “The Business Cycle Associated with Exchange-Rate-Based Stabilization,” World Bank Economic Review, Vol. 6 (May 1992), pp. 279 –305.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Marinakis, Andrés E., “Wage Indexation, Flexibility and Inflation: Some Latin American Experiences During the 1980s,” ILO Occasional Paper No. 6 (Geneva, Switzerland: International Labor Office, 1993).

    • Search Google Scholar
    • Export Citation
  • Ros, Jaime, “A Review of Literature on Inflation and Stabilization” (Unpublished; Helsinki: World Institute for Development Economics Research, 1989).

    • Search Google Scholar
    • Export Citation
  • United Nations, Economic Commission for Latin America and the Caribbean, Equidad y Transformación Productiva: Un Enfoque Integrado (Santiago, 1992).

    • Search Google Scholar
    • Export Citation

Note: I am very thankful to Joao Carlos Scandiuzzi for his excellent work as research assistant.

1

See Amadeo and Banuri (1991) for an analysis of labor market institutions and their role in the adjustment process in Latin America and Southeast Asia.

2

Cardoso (1992) estimates an equation using data from seven Latin American countries to show that annual “real wages fall by 14 percent when inflation doubles.” She also argues that in many countries the decline in real wages can be associated with the increase in poverty.

3

Calvo and Végh (1992) and Kiguel and Liviatan (1992) raise another hypothesis—usually based on the behavior of the interest rate—for the behavior of the level of activity in the course of exchange-rate-based stabilization plans.