Who Gains from Free Trade

Export-Led Growth, Inequality and Poverty in Latin America

Rob Vos

• Publisher: Taylor & Francis
• Year: 2007
• Print ISBN: 9780415632386
• eBook ISBN: 9781135987015
• Edition: 1
• Topic: Business
• Subtopic: Business General
• Index: Business

The issue of the pros and cons of free trade from the point of view of developing countries refuses to dissipate, and in Latin America, the debate rages most fiercely. Argentina is still licking its wounds after a catastrophic past five years, and Brazil and others have hardened their line – even going so far as to initiate the influential new G20 group of the most powerful LDCs.

Who Gains from Free Trade examines the extent to which trade reforms have been an important source of the slowdown of economic growth, rising inequality and rising poverty as observed in many parts of the region. This volume presents a comprehensive analysis of this important topic, utilizing:

• research based on sixteen country narratives of policy reform and economic performance
• rigorous general equilibrium (CGE) modelling of the economy-wide effects of trade reform for all country cases
• application of an innovative method of microsimulations to assess the employment and factor income distribution impact of policy reforms on poverty and inequality at the household level.

This important study, a valuable resource for postgraduate students of development economics and political economy, examines all the current issues and brings together some of the world's leading experts.

1 Introduction

Rising exports, slower growth and greater inequality: is trade liberalization to blame?

Rob Vos, Enrique Ganuza and Samuel Morley

1.1 Reforms and dismal growth

Since the late 1980s, almost all Latin American countries have gone through a process of far-reaching economic reforms, featuring in particular trade, financial and capital account liberalization. Economic opening has gone hand in hand with large financial inflows, particularly in the first half of the 1990s. Increased openness has brought new sources of economic growth but it has also increased volatility and sensitivity to external shocks. At first, the reforms seemed to be working as promised. Economic growth increased, inflation declined and there was a big surge in foreign capital inflows. But somewhere around 1995 or thereabouts, growth faltered, particularly in the countries of South America. So did exports. They had been expected to be the leading sector in the post-reform growth model, but unlike the Asian experience, export-led growth in Latin America has proved so far to be anything but a development miracle. Not only has overall growth been far less than during the period of import substitution, but export growth has also slowed down and is still dominated by primary products. Also, income inequality has risen in a context where inequality was already very high to begin with. Not surprisingly, little progress has been made on poverty reduction since 1990. Indeed, poverty has been rising in the region as growth slowed down after 1995.

An important and natural question to ask is whether the reforms are the reason or at least a significant contributing factor for the poor performance of the region since the mid-1990s? Did they contribute to the growth slowdown and the rise in poverty and inequality? These are crucial questions and the more so because the countries in the region are considering further trade integration under the flag of the Free Trade Area of the Americas (FTAA), bilateral free trade treaties with the USA and/or in the context of regional economic integration areas such as Mercosur, the Andean Pact, CAFTA and NAFTA. The countries of the region also have become members of the World Trade Organization (WTO), subscribing to further trade liberalization, including the elimination of export subsidies. Key players from the region, like Brazil, were instrumental in blocking a new global trade agreement under the aegis of WTO in September 2003 precisely because of concerns about the developmental and equity effects of freer world trade.

In this book, we take a dual approach to trying to find answers to what was the impact of the reforms and the switch to an export-led growth strategy. Since we believe that countries differ in structure and that structure matters, we have collected together case studies for 16 of the countries in the region, covering over 90 per cent of its population and of its GDP. The comparative analyses in Chapters 2–3 covers all the 16 countries, while in Chapters 4–12 selected country cases are presented whereby for each country we first present a historical narrative of recent growth and poverty trends.¹ Then, we develop a country-specific computable general equilibrium (CGE) model with which we can simulate the effects on output and employment of trade liberalization, export promotion and various external shocks, holding everything else constant. We also develop a microsimulation model to translate these sectoral impacts into changes in poverty and inequality.

Our brief answer to the main questions posed above is that trade liberalization and the switch to export-led growth are not the cause of the growth slowdown in the region. Nor are they the cause of rising poverty and inequality. On the contrary, for most countries their impact is mildly positive for both growth and poverty reduction. But overall, the impact of the reforms and export growth, while positive, are small. So while the reforms are not to blame for the region’s woes, they do not appear to be the solution for them either. The case studies also show that trade reforms increase skill-intensity and that some social groups win (mostly the better educated workers and profit earners) and some lose (often agricultural and unskilled workers) in the process providing the explanation for the rising inequality observed in most country experiences.

1.2 The analytical approach of this study

There are no facile answers about the impact of trade reforms on growth, poverty and income distribution but, whatever they are, they are bound to be country specific, requiring in-depth analysis of the nature and timing of reforms and the prevailing economic structure. Cross-country econometric studies are often used as an alternative way to find answers to the questions raised.² Such studies typically are not very enlightening about the mechanisms by which trade liberalization and export promotion affect growth and income distribution and by their nature have little to add to country-level policy debates. Instead, we will rely on country case studies to obtain policy-relevant answers. But even then, how can we separate the effects of specific policy changes from other factors such as external shocks and other policy initiatives? We faced this standard problem of economic analysis by combining four methodological elements.

First, we present country narratives for 9 out of 16 countries discussing policy changes and observed outcomes in a before-and-after approach. The country stories start with a basic set of questions and hypotheses and a simple analytical framework suggesting possible channels of causation of the type outlined above.

Second, still within the realm of before-and-after, to identify the importance of changes in the external and domestic policy environment we applied a set of standardized decompositions of sources of output growth and of the forces driving balance-of-payments adjustment (see Chapter 2 for the related methodologies). We use the former type of decomposition to identify to what extent trade has become a more forceful engine of growth since trade liberalization. The second type of decomposition is a standard decomposition of the current account of the balance of payments, which enables us to show the observed relative importance of various types of external shocks (in terms of trade, world demand, interest rates, etc.) and domestic responses (through adjustments in aggregate demand, and import and export ratios) in the pre- and post-liberalization episodes.

Third, to isolate the effects of different types of trade reforms and external shocks, a standardized CGE model framework was set up (see Löfgren et al. 2001 and Chapter 3). The country models capture the structural features of each country through a social accounting matrix, country-relevant elasticities of commodity and factor market adjustment to relative price changes and country-specific ‘closure rules’, which define policy regimes (such as a fixed or flexible exchange rate regime, a fixed or endogenous fiscal balance) and behavioural rules in key markets (endogenous or exogenous capital flows, price or quantity adjustment in different segments of the labour and other factor markets and the degree of factor mobility across sectors). We then perform a common set of coun-terfactual model simulations for each country (with-and-without) of different trade policies and external shocks and assess the (static) impact on (sectoral) output and export growth, employment, labour–income inequality and household welfare.

Finally, the poverty and inequality impact of the counterfactual CGE analysis is estimated using a microsimulation methodology. This sequential modelling approach fits a recent tradition of assessing the poverty impact of economic policies in developing countries (see Bourguignon et al. 2002b). One drawback of the typical CGE model, including the ones used in this study, is that simulation results are only obtained for between-group differentials in terms of employment and factor remuneration. The assessment of the impact on household incomes and consumption is also limited to differentials between highly aggregated groups. In consequence, the CGE model captures only a part of the distribution effects and therefore lacks sufficient detail for a meaningful assessment of the poverty impact of the policy simulations. The microsimulation approach helps to overcome this shortcoming by taking the outcomes of the CGE model simulation and applying the simulated changes in labour market parameters to the full income distribution using household survey data. The microsimulation methodology also resolves the issue of which individuals in the distribution are likely to move position in the face of the counterfactual labour market adjustment (e.g. which worker will loose his or her job if the unemployment rate moves up). Most country studies follow the microsimulation approach as developed in a previous study (Ganuza et al. 2002), but in a few cases (Argentina, Mexico) the approach as suggested by Bourguignon et al. (2002a) was followed. These methodologies are described in Chapter 3.

1.3 Overview of the main findings

Chapter 2 analyses growth trends and the vulnerability to external shocks of the countries in Latin America and the Caribbean over almost a quarter of a century since 1980. It is shown that as a consequence of the process of economic opening to world markets, growth has become export-led in virtually all countries of the region. However, unlike the experience with export-led growth in East Asia, Latin America’s new growth strategy seems to come with a number of less virtuous characteristics. First, while more reliant on exports, economic growth did not significantly increase after trade opening. Instead, growth slowed down and economic performance was worse in the second half of the 1990s than in the first and most countries slipped to negative per capita income growth at the turn of the century. Second, vulnerability to fluctuations in global commodity markets (i.e. terms-of-trade shocks and volatility in global demand) remains high and is a first indication of insufficient diversification of trade. This vulnerability to trade shocks cannot by itself explain the dismal growth performance, as in fact for most countries terms of trade and world demand for their exports improved during the 1990s.

Third, for most countries of the region export growth has been below that of world trade, implying lower export penetration in global markets as a result of losses in competitiveness. At the same time, import dependence has risen more strongly than the capacity to export. As a result, capital flows have become more important to sustain a growth path built on this paradoxical combination of increasing reliance on exports and a structural rise in the trade deficit. Capital flows in turn have both initiated (to the extent exogenous) and reinforced this pattern by pushing up real exchange rates, cheapening imports and squeezing profits for exporters in the short run. As capital flows themselves have been volatile, to a large part for reasons exogenous to the economic conditions of the countries of the region, macroeconomic adjustment has become more difficult to steer, resulting in short-lived booms as access to foreign borrowing eases and demand deflation as it contracts with important implications for employment and wages. Trade reforms therefore must be studied in conjunction with such macro-economic constraints.

The CGE model framework laid out in Chapter 3 deals with this by the explicit specification of alternative macroeconomic closure rules under which trade reform policies work their way through the economy, labour markets and eventually poverty. Such closure rules define whether in a particular context there is easy or constrained access to foreign savings, whether investment demand is savings constrained or not and whether there is a binding fiscal constraint or not. These macroeconomic conditions also reflect alternative policy regimes: a fixed versus flexible exchange rate regime, a Keynesian demand-constrained versus neoclassical supply-constrained policy environment and a fixed versus flexible target for the fiscal deficit, respectively. By applying the same closure rules across countries we are able to detect the differential impact of trade reform due to differences in economic structure and degree of flexibility in market adjustment. Subsequently, we compare this with the type of macroeconomic closure rules considered relevant for each country case to understand how trade reforms interact with varying macroeconomic constraints. We find that trade liberalization increased output in almost every country in our sample. This finding is consistent across alternative macroeconomic constraints, but output growth is stronger under a regime with easy access to foreign capital and subsequent real exchange rate appreciation. Under the latter scenario, the country models reproduce the pattern observed in reality for most countries, that is, trade liberalization increases the growth rate but it is led by the non-traded goods sector and imports rather than exports. The trade deficit increases. Output effects are dampened visibly if the level of foreign savings is assumed to be fixed before and after trade opening.

In the same vein, trade liberalization leads either to increasing wages or employment depending on the labour market closure used in the country-specific models. Consistent with this, poverty declined in all but one country (Ecuador) in the unilateral trade liberalization scenario. Rising labour inequality, particularly between skilled and unskilled workers, emerges in the majority of cases, but does not necessarily translate into more inequality in per capita household incomes because of offsetting positive employment effects. The inequality effects tend to be stronger in natural resource abundant economies like Argentina, Bolivia, Costa Rica, Ecuador and Peru. The effects on poverty and inequality of further unilateral trade liberalization are small, but positive.

These results are broadly consistent with moving to completely free trade under WTO rules or to a region-wide multilateral trade agreement under FTAA. Both these changes also reduce poverty and inequality in most of the countries, but not in all cases. Poverty rises (modestly) under these scenarios in Costa Rica (only WTO), Ecuador, Paraguay and Venezuela, mainly due to the negative impact on agriculture in these countries, which is not compensated by sufficient employment and income growth in other sectors. Gains from further trade integration are clearly not evenly distributed within countries and this is therefore a legitimate concern in trade negotiations.

With a few exceptions (Brazil and Argentina), unilateral export promotion via export subsidies is also found to be poverty reducing in apparent contradiction with the WTO scenario. In a sense, it works like a tariff cut stimulating aggregate employment as mostly more labour-intensive (e.g. agriculture) sectors benefit from subsidies that are increased in the scenario. Skill-differentials rise in some countries and fall in others. Thus, one cannot say that choosing a more export-led growth strategy will in general favour either the skilled or the unskilled. This depends on the export structure of individual countries. The country stories in Chapters 4–12 present the details.

These model results have to be interpreted with some caution. The reader should note that our findings are general equilibrium, comparative static results that do not reflect the costs of adjusting to a changed production structure. If we assume that the exchange rate is fixed, the simulation determines the long-run or permanent impact of lowering the tariff rate while bringing in more foreign capital to permanently finance a bigger balance-of-payments deficit. As indicated, the latter change is expansionary (though growth is led by non-traded goods rather than exports). If foreign saving is constrained, the exchange rate has to depreciate to allow exports to expand enough to pay for additional imports. But total output and employment increase in both cases and poverty declines. The simulation results also suggest that if no poverty reduction was observed in practice after trade liberalization, it is either because a lot of other poverty-increasing factors were changing at the same time (most typically dealing with macro shocks; see Taylor and Vos 2002) or because the economies are still in the process of adapting their production structures.

Notes

1 The full set of country cases can be obtained from United Nations Development Programme (UNDP) or downloaded at the following website: http://www.undp.org/rblac/drafts/

2 See, e.g. the recent debate spurred by papers of Sachs and Warner (1995), Rodriguez and Rodrik (2000), Dollar and Kraay (2001a, b), Easterly (2001) and Rodrik et al. (2002). And see Berg and Krueger (2003) for a ...

Table of contents

• Cover
• Halftitle
• Title
• Copyright
• Contents
• List of tables and figures
• About the authors
• Preface
• 1 Introduction: Rising exports, slower growth and greater inequality: is trade liberalization to blame?
• 2 Bad luck or wrong policies?: External shocks, domestic adjustment and growth slowdown in Latin America and the Caribbean
• 3 Are export promotion and trade liberalization good for Latin America’s poor?: A comparative macro–micro CGE analysis
• 4 Argentina – the convertibility plan, trade openness, poverty and inequality
• 5 Bolivia – export promotion and its effects on growth, employment and poverty
• 6 Brazil – the impact of trade openness on employment, poverty and inequality
• 7 Costa Rica – export orientation and its effect on growth, inequality and poverty
• 8 Cuba – export promotion, poverty, inequality and growth in the 1990s
• 9 Ecuador – dollarization, trade liberalization and poverty
• 10 Mexico – do the poor benefit from increased openness?
• 11 Peru – impact analysis of trade liberalization on poverty and inequality
• 12 Uruguay – export growth, poverty and income distribution
• Index