Chapter One A New Approach to Analyzing Reforms: Micro-Micro Linkages In the last ten to fifteen years, the Latin American and Caribbean region has undergone the most significant transformation of economic policy since World War II. Through a series of structural reforms, an increasing number of countries have moved from the closed, state-dominated economies that characterized the import-substitution industrialization model to economies that are more market oriented and more open to the rest of the world. Complementary aspects of the process have accorded a new priority to macroeconomic stability, especially lower rates of inflation, and to increasing expenditure in the social area. Policymakers expected that these changes would speed up economic growth and increase productivity gains, at the same time that they would lead to the creation of more jobs and greater equity. Have those expectations been fulfilled? It is impossible to make more than a preliminary analysis at this point, since in many cases the reforms are less than a decade old. Even tentative conclusions will be useful, however. Governments must decide whether the new policies are moving in the right direction and, even if they are, whether they would benefit from some mid-course corrections. This requires more data and analysis than were available when we began this project. Moreover, any conclusions that can be drawn will be relevant beyond the boundaries of the region itself. In many parts of the world, including central and eastern Europe, the former Soviet Union, Africa, some parts of Asia, and even some industrialized countries with respect to specific policy areas, governments are experimenting with similar policy changes. Since Latin America has had a head start, others are interested in learning from its successes and failures. The Literature on the Reforms We are not, of course, the first to study the reforms and their impact. During the last decade, an extensive literature has developed on the topic. Early works tend to express quite rigid views--albeit with little evidence to back them up--that the reforms would either resolve most of the socioeconomic problems in the region or that they would result in disaster. With time and experience, opinions have converged to a certain extent, and a more nuanced evaluation has emerged. Substantial differences of opinion still remain, however, on whether additional reforms are needed, what role the state should play, and what can be expected of the new economic model. These differing evaluations lead to variations in policy recommendations. Most analysts have studied the reforms in terms of their impact on growth. While emphasizing that growth improved with respect to the 1980s, the majority holds that the impact was disappointing. Post-reform growth was lower than the region's past performance, lower than in some other regions, and lower than necessary to deal with the region's social problems. Others, however, say that the growth rate was as good as could be expected or that growth would have been even slower without the reforms. Most of this analysis concentrates on the aggregate level. An important exception is the United Nations Economic Commission for Latin America and the Caribbean (ECLAC), which points out that the lack of dynamism was accompanied by fundamental changes in sectoral and microeconomic momentum; in particular, exports from the region had a good performance, although they were concentrated in a few sectors and generally incorporated low levels of technology. With respect to employment, the consensus is that job creation was generally insufficient, more because of relatively slow growth of the gross domestic product (GDP) than because of problems of employment elasticities. The International Labour Organization (ILO) has taken the lead in stressing the problems in job quality. Thus Victor E. Tokman emphasizes that there was a strong expansion of employment during the economic recuperation of the first half of the 1990s, but these jobs were basically low paid and their productivity was low. In addition, employment became increasingly precarious, both because jobs were concentrated in the informal sector, generally without social security, and because modern firms outsourced the labor process, transferring the cost of adjustment and instability to small firms and the self-employed. Somewhat greater disagreement exists with respect to the impact of the reforms on equity. The Inter-American Development Bank (IDB) argues that the main deterioration in distribution occurred in the so-called lost decade of the 1980s and that the reforms helped to slow the tendencies of deteriorating income distribution and increasing poverty. In particular, it identifies a positive effect of the trade opening on the real income of the first three quintiles of the income distribution, and a negative effect on the richest 20 percent. In contrast, Bulmer-Thomas, on the basis of empirical evidence for the period up to the early 1990s, suggests that the results of the so-called new economic model were basically regressive in terms of income distribution. This was due to a decrease in real wages together with an increase in unemployment, real interest rates, the weight of the informal sector, and the concentration of wealth. The only element of the reform process that had a progressive role was fiscal reform, according to this analysis, in that it made possible the reduction of inflation. Similarly, Berry points to the negative role played by trade reform, which favored capital-intensive technology and subjected small firms to intense competitive pressures. Measuring the impact of the reforms on growth was very difficult since so many other things were happening simultaneously. In addition, the reforms themselves were a process, so it was hard to capture their impact at any particular point in time. This problem was even more acute in studying employment and equity than in analyzing growth, since data are more unreliable and the definition of relevant indicators is much more complex. Two main strategies have been used in the literature. ECLAC, Sebastian Edwards, and Shahid Javed Burki and Guillermo E. Perry, among others, describe the reform process and present indicators to capture their degree of implementation (such as the changing level of tariffs or number of privatized firms). This approach then relies heavily on examples, qualitative indicators, and case studies of firms or countries to evaluate the progress of the reforms. A different approach was followed by the IDB, which developed an index of reforms to summarize the reform process in five areas (trade, taxes, finance, privatization, and the labor market). The IDB designed these quantitative indicators to address the problem that there had been no systematic attempt to measure what had been reformed; without such a measure it was very difficult to evaluate the impact and to separate the reforms from other trends. Efforts to quantify the reforms often confuse policy and performance variables. Given that it is frequently easier to devise measures of the results of an action than to measure the action itself, many studies mix the two types of indicators. Thus studies on trade reform might use indicators like the ratio of exports plus imports to GDP, while attempts to measure fiscal reform might rely on measures like taxes as a share of GDP. In this context, despite its limitations, the IDB's reform index is a contribution in that it helps to concentrate attention on the reforms themselves. An author's approach to characterizing the reforms is linked to the methodology used for analyzing their effects. Authors who explicitly or implicitly reject the possibility of measuring the reforms tend to stress the richness and complexity of the processes and to make evaluations that are mainly historical. They make attempts to attribute results to individual elements of the new policy package. A particular variant of this approach is found in an analysis by the World Bank, which says: "[our] analysis ... of economic performance fulfills the dual purpose of assessing the state of reform and evaluating general economic development, under the hypothesis that in the recent experience of Latin American countries, economic development is largely a reflection of macroeconomic and structural reforms." In contrast to the historical approach are the authors who try to quantify the reforms; they then use their measures as independent variables in regression equations to "explain" economic growth or other variables of interest. Thus, for example, growth equations of the type popularized by Robert Barro tend to incorporate an additional variable. The IDB index serves this purpose for many analysts, but other ad hoc proxies have also been used. In principle, analysis of this type makes it possible to separate the influence of the reforms from that of other simultaneous events, especially macroeconomic policies; in practice, the variables are so interrelated that the results are questionable. Indeed, many problems exist with this literature, which is extremely diverse in coverage as well as methodology. The quantitative analyses of the IDB, in particular, have been criticized, both for the oversimplifications necessary to construct the reform indexes and for the attempt to draw conclusions about the effects that go well beyond what the data and the (sometimes debatable) econometric analysis will bear. Nonetheless, these criticisms should not obscure the advance that these efforts at quantification and methodological precision signify. Neither should they minimize the problems of alternative methodologies based almost exclusively on historical analysis that uses ad hoc examples, circumstantial evidence, and expert opinion. In addition to methodological problems of various sorts, four substantive limitations appear in the studies reviewed. (1) Failure to disaggregate the variables . The tendency of the analysis to focus exclusively on variables and processes at the aggregate level makes it difficult to capture the differences between countries, sectors, and socioeconomic actors with respect to the adoption, implementation, and impact of the reforms. If a possible result of the reform process is the growing heterogeneity of the productive structure and its performance, failure to disaggregate makes it impossible to recognize this, much less explain it. This problem is important at two levels. First, the lack of attention to processes and variables at the sectoral level hinders comprehension of the processes of specialization within the productive system, which may affect the growth capacity of the new structures generated by the reforms. These factors may also weaken domestic supplier chains, reducing the impact of investment and technological modernization. Second, the absence of analysis and evaluation at the microeconomic level makes it impossible to trace a key objective of the reforms: to modify the behavior of microeconomic actors, in particular, of firms. Moreover, no one has explored changes in entrepreneurial behavior in the areas of investment and the incorporation of technical progress across a wide range of countries, which would be a crucial indicator of the success or failure of the reform process. (2) Failure to emphasize the mechanisms of articulation between national economies and the international context . This problem is seen in the scarce attention that is paid to phenomena associated with the external financial opening and to the dynamics of foreign direct investment and the assimilation of technological progress. It is paradoxical because the preoccupation with an efficient integration into an increasingly globalized world economy is at the center of the new model of growth. Beyond a general recognition of the importance of the trade opening, however, the noncommercial dimensions of that integration do not receive adequate attention. Nor is much attention paid to the links between the reforms, macroeconomic policies, and the international context. (3) Failure to consider that the package of reforms and policies may be internally inconsistent . Although the framework of a market-friendly development model informed the design of the reforms and the policy instruments they engendered, detailed analysis of the internal consistency of the reform package and associated policy variables is lacking. This concerns relations among the reforms themselves, between the reforms and the macroeconomic and social policies that accompanied them, and between both the reforms and policies and the international environment in which the new economic model has to function. The assumption of coherence has led to the persistence of some serious problems. (4) Scant attention to the articulation of the dynamics of employment and the distribution of income with the rest of the model . Employment and equity, as well as their relation to investment, productivity, and growth, are absent from the original formulations of the Washington Consensus and the market-friendly development model. There was a subsequent attempt to link the reforms to employment and equity, but the mechanisms of transmission among the various components have yet to be adequately conceptualized. A New Conceptual Framework This book builds on the literature, but it also presents significant innovations to deal with the four problems identified above. The main characteristic that distinguishes it from other comparative studies of economic reforms is the focus on the interaction between macroeconomic and microeconomic processes. To make significant advances at this time, it is crucial to focus less exclusively on the macroeconomic and regional levels and more on countries and the microeconomic behavior of firms, grouped by sector, size, and ownership characteristics. Different countries and groups of firms are affected quite differently by government policies, including structural reforms, and by the increasingly globalized world economy. Some have been able to take advantage of the new opportunities created, while others have seen their situations become ever more precarious. The sum of these behaviors produces the aggregate trends that others have observed and measured. Knowing what lies behind the aggregates is essential for designing policy measures to improve future economic performance. Another way to characterize our approach is that it insists on the need to make economic actors central to the analysis and to try to understand their reactions to government policies in order to explain, predict, and (if necessary) modify their behavior. In particular, we focus on entrepreneurs' decisions on whether to invest and to incorporate new technologies. Under what domestic and international conditions will they make positive decisions? What will be the time frame for implementing investment decisions? Without a positive response on investment, and without increased productivity through technical progress and better skills for workers, medium- and long-term growth cannot take place, although economic recovery can occur. A disaggregated approach is also necessary for analyzing outcomes other than growth, especially the generation of employment and any change in the patterns of income distribution. Job-creation capacities vary widely for large and small firms and for firms in labor-intensive or capital-intensive or natural resource-intensive sectors. The skill differential that characterizes jobs in different categories of firms is then a crucial factor in determining patterns of income distribution. Given the centrality of economic actors for our approach, we start by considering the reforms as a set of signals in the form of government policy decisions. When governments want to change the way their economies (and societies) operate, they make policy decisions and transmit them to the relevant actors; these decisions constitute our signals. The governmental decisions are essential for creating a new environment in which the private sector can operate more dynamically. The environment is also influenced, both positively and negatively, by international forces that are generally beyond the control of most governments, including capital flows, interest rates, expansion of output in importing countries, and trade regimes. These will have to be taken into consideration as well. In addition to the signals and the environment they help to create, we need to be concerned about the reception of the policy signals at the microeconomic level in decentralized economies. A first issue with respect to reception obviously involves information. Do the actors know that policies have changed and what the government is trying to signal through its policy decisions? In the real world of imperfect information, not all actors have access to the same formal channels, and informal channels are even more skewed to favor some groups over others. We must therefore assume that information will be unequally distributed, and that less powerful actors will have less access. Information alone, however, is not sufficient. A second question concerns the credibility of the information and signals. Do the actors believe what the government says at any given moment, and do they believe that the new policies will remain in force for the foreseeable future? If they have doubts, they will not be willing to take the risks involved in making the new investments that are necessary to incorporate new technology and, ultimately, to pave the way for a new growth path. Lack of credibility with respect to economic reforms arises from three possible sources. First, the various components of the policy package may be inconsistent. If the reforms and policies are inconsistent among themselves, they create mixed signals, and the recipients are confused about what they should do. In an extreme situation, inconsistency may lead to a crisis that precipitates a change in policy stance. Second, key sectors of the population may not support the policies. Lack of support or, more important, active opposition may cause the relevant actors to surmise that the policies will be abandoned sooner or later. Third, the reforms may exist within an unfavorable international environment. Even with consistent policies and a reasonable degree of domestic support, an unfavorable international environment, such as a lack of external finance or a sharp deterioration in the terms of trade, will pose credibility problems. This is especially the case with a policy package that increases the openness of the economy and thus the dependence on external factors. These three sources of credibility problems are closely interrelated. A negative international environment can lead to, or increase, inconsistency among policies. Ironically, even a positive external context (such as abundant international finance) can create policy inconsistency. Either policy inconsistency or an unfavorable environment can undermine support for policies. Insofar as key economic actors perceive the presence of one or more of these problems, they are likely to draw the conclusion that the policies will be changed and will make their own decisions accordingly. Finally, beyond information and its credibility, the economic actors must be able to take advantage of the new circumstances. For example, they must have access to domestic or international financial resources that enable them to invest; they must have knowledge of the technological advances in their area; they must have access to the appropriate equipment that embodies the new technology; and they must have workers with the necessary training to use the new equipment they have acquired. Again, the resources at the disposal of different actors vary substantially across sizes and sectors of firms. To study the processes of signaling, response, and the resulting outcomes, we have worked with the conceptual framework shown in figure 1-1. Our analytical framework begins with the external context, which we model as variables related to international finance and the demand for Latin American exports. The past performance of these and other international variables has helped to determine the initial domestic conditions (economic, social, and political) in each country. In the present, the external context has a strong impact on government policy, making certain policy choices more likely than others. Beyond its impact in the policy arena, external finance facilitates investment and technical change processes, while international demand and the vagaries of financial flows have an impact on the dependent variables, especially growth rates. Initial conditions within each of the countries are mostly determined by domestic developments, although they are also influenced by external factors. We take these initial conditions as given, rather than trying to explain them, but they are crucial in determining both policy choice and response. From the perspective of policy choice, we are particularly interested in several economic variables, including growth and inflation rates, the structure of output and employment, and links with the world economy. Social characteristics of the population and the ability of governments to make and implement policy decisions are also important. At the firm level, the accumulated learning and productive capacity are elements that governments must take into account. Based on the initial conditions in each country and on external influence, governments make decisions on reforms (such as import liberalization, domestic financial liberalization, opening of the capital account, privatization, and tax reform), macroeconomic policies (including fiscal, monetary, and exchange rate measures), and social policies (especially with respect to education and health). Although the nine countries for this study had undertaken substantial reforms and changes in macroeconomic and social policies, policy decisions on individual items were not necessarily the same. On the contrary, one of the things we want to investigate is the difference in the choice and implementation of reforms and policies and how this affected the outcomes. The study analyzes investment and technical change at both the aggregate and sectoral levels to see what determines the response to the reforms. At the aggregate level, the response is related to the interaction of the reforms with macroeconomic policy and the international context. The uncertainty created by changes in the rules of the game and the volatility of key macroeconomic variables merit particular attention. The behavior of these variables, and the uncertainty they generate, may lead to a delayed response in investment and technological change. At the sectoral level, we study the transmission mechanisms between these same variables and the process of investment and the incorporation of new technologies. These operate through shifts in market structure, corporate strategy, and the entry of new actors into specific markets. The three dependent variables of the model are growth, employment, and equity. Growth in the post-reform period is compared with that of the 1950-80 base period, and its components (that is, capital accumulation, labor accumulation, and productivity) are analyzed in a growth-accounting framework. The characteristics of the growth process (in particular, the types of firms that are expanding output dynamically or lagging behind) and the decisions on the type of technology to be incorporated will determine employment generation; the latter is disaggregated by productive sectors and size of firm. Employment characteristics, especially the salary differential between skilled and unskilled workers, are important in determining the distribution of income, although this may be offset by other economic and social trends. Six Propositions We use this framework to explore a set of six propositions. They take as a point of departure the previous judgment on the impact of the reforms: growth has been modest; employment has grown slowly and with problems in job quality; and inequality has not improved and may even have gotten worse. The propositions, then, are directed toward answering the question of why the impact has not been more favorable. First, the initial conditions in the various countries were quite diverse and affected the extent to which reforms were adopted. Variables of particular importance include macroeconomic stability, economic distortions, past growth rates, and the degree of governability. The greater such problems in the recent past, the more willing countries are to undertake large-scale reforms in the hope of improving future economic performance. At the same time, a history of past turbulence and failed reform or stabilization attempts will increase the credibility problems of governments intent on changing policy course. Countries that have done well in the past have much less reason to undertake risky reforms, and there is likely to be less support for doing so. Second, governments frequently introduced reforms that were inconsistent with their macroeconomic and social policies. The most obvious case is when high rates of inflation and deficits continued, thus leading to uncertainty for economic actors. Even if stability is attained, however, other contradictions may exist. For example, exchange rate appreciation, stimulated by newly liberalized capital flows, prevents firms from taking full advantage of the new export opportunities presented by trade liberalization. The way in which budget deficits are controlled may also generate contradictions: cutting expenditures for social policies limits growth potential as well as possibly increasing inequality, while reducing public investment limits the expansion of essential infrastructure. Third, the reforms were slow to produce an impact at the microeconomic level because of the great uncertainty they generated, especially if they were combined with macroeconomic instability. This led to a hesitation on the part of investors to engage in large-scale projects, as they preferred to defer irreversible decisions. One possibility is that the timing of the response is arbitrary, depending on particular country conditions. It is also possible that a pattern can be found across countries, geared to the level of uncertainty and its impact on investors' risk-return calculations. Fourth, the uneven response of actors helped to explain both the less-than-hoped-for performance to date in most countries in the region, as well as the differential performance across countries. We may expect a heterogeneous reception of the signals--creating leading and lagging sectors--and thus the average response for the economies of the region would be modest. The percentage of actors receiving and responding positively to the signals also varies across countries (and across sectors and firms by size), so that country averages vary quite substantially. Within countries and sectors, economic actors' past experience results in a heterogeneous reception of, and response to, the signals from the reforms. The effects have implications for all three of our dependent variables: growth rates, employment, and equity. Fifth, the positive effects of reforms were frequently undermined by unfavorable trends in the international economy. Capital flows, for example, helped to sustain high growth, despite contributing to distortions such as overvalued exchange rates. However, sharp declines or reversals of--and great volatility in--private capital flows can overwhelm the positive effects. On the trade side, external demand can also contribute to economic growth, but volatility in export prices and sudden declines in demand can undermine the positive impact. The reforms have exacerbated both the positive and negative effects by increasing the openness of the economies to international flows. Sixth, the reforms were incomplete in that they lacked the proper institutional support typically found in the industrial world. A key example relates to the development of factor markets--namely, capital, technology, and labor. Although these markets are imperfect in all parts of the world, they are especially deficient in developing countries, including Latin America and the Caribbean. Capital markets lack long-term segments, labor markets lack training opportunities for unskilled or low-skilled workers, and technological progress is frequently limited to the largest firms. Another way in which the reforms were incomplete refers to the lack of a proper regulatory framework to complement privatization and liberalization reforms. For example, banking crises have been frequent in Latin American and Caribbean countries in recent years, since local governments have failed to establish and enforce adequate rules on reserves and lending practices. (Continues...) Copyright © 2000 The Brookings Institution Press. All rights reserved.
