POLICY RESEARCH WORKING PAPER 2b90 Estimating the Poverty Impacts of Trade Liberalization Jeffrey J. Reimer The World Bank RPM Development Research Group E Trade February 2002 I POLICY RESEARCH WORKING PAPER 2790 Abstract As a new round of World Trade Organization perhaps the most important link between trade and negotiations is being launched with greater emphasis on poverty, since households tend to be much more developing country participation, a body of literature is specialized in income than they are in consumption. emerging which quantifies how international trade Meanwhile, survey data on the income sources of affects the poor in developing countries. In this survey of developing-country households has become increasingly the literature, Reimer summarizes and classifies 35 trade available. As a result, this survey gives particular and poverty studies into four methodological categories: emphasis to the means by which studies address factor cross-country regression, partial-equilibrium and cost-of- market links between trade and poverty. living analysis, general-equilibrium simulation, and The general conclusion of Reimer's survey is that any micro-macro synthesis. analysis of trade and poverty needs to be informed by These categories include a broad range of both the bottom-up and top-down perspectives. Indeed, methodologies in current use. The continuum of recent "two-step" micro-macro studies sequentially link approaches is bounded on one end by econometric these two types of frameworks, such that general analysis of household expenditure data, which is the equilibrium mechanisms are incorporated along with traditional domain of poverty specialists, and sometimes detailed household survey information. Another labeled the "bottom-up" approach. On the other end of methodology in a similar spirit and also increasingly used the continuum are computable general equilibrium involves incorporating large numbers of surveyed models based on national accounts data, or what might households into a general-equilibrium simulation model. be called the "top-down" approach. Although most of these studies have so far been limited Another feature of several recent trade and poverty to a single region, these approaches can be readily studies-and one of the primary conclusions to emerge adapted for multi-region modeling so that trade and from the October 2000 "Conference on Poverty and the poverty comparisons can be made across countries within International Economy" sponsored by Globkom and the a consistent framework. World Bank-is the recognition that factor markets are This paper-a product of Trade, Development Research Group-is part of a larger effort in the group to understand the impact of trade reform on poverty. Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Paulina Flewitt, room MC3-333, telephone 202-473-2724, fax 202-522-1159, email address pflewitt@worldbank.org. Policy Research Working Papers are also posted on the Web at http:II econ.worldbank.org. The author may be contacted at reimerj@purdue.edu. February 2002. (35 pages) The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the countries they represent. Produced by the Research Advisory Staff Estimating the Poverty Impacts of Trade Liberalization Jeffrey J. Reimer Purdue University Contact information: Center for Global Trade Analysis and Department of Agricultural Economics, Purdue University, West Lafayette, Indiana, 47907-1145, USA. Email: reimerigpurdue.edu Acknowledgements: This work was supported by the World Bank under the project on Poverty and the International Economy. I am very grateful to Tom Hertel, Will Martin, L. Alan Winters, Martin Ravallion, David Tarr, Anne-Sophie Robilliard, and John Cockburn for helpful input at various stages of this project. Table of Contents I. Introduction .................................. 1 II. The Factor Income Link ..................................4 III. Cross-Country Regression ...................................7 IV. Partial-Equilibrium/Cost-of-Living Analysis ..................................8 V. General-Equilibrium Simulation .................................. 12 VI. Micro-Macro Synthesis .................................. 17 VII. Summary and Conclusions .................................. 21 References..23 List of Tables Table 1. Cross-Country Regression .29 Table 2. Partial-Equilibrium/Cost-of-Living Analysis .29 Table 3. General-Equilibrium Simulation .31 Table 4. Micro-Macro Synthesis .35 I. Introduction A wave of trade liberalization over the last decade has positioned many developing countries to increasingly participate in world markets. This new openness has been accompanied by concern that the poor will be adversely affected, and that the distribution of income in developing countries will deteriorate. Indeed, suggestions have been made to emphasize poverty reduction in the next round of World Trade Organization (WTO) negotiations, and even to label it the "development round". Accordingly, the issue of trade and developing-country poverty has become the focus of much research in the last several years. This paper is a survey of recent studies that analyze how trade policies and other types of external shocks affect the incidence of poverty in developing countries. The objective is to summarize the methodologies of the diverse strands of research that are currently being conducted. It is not yet possible to provide a comprehensive synthesis of findings, since many of the studies are very much in the draft stage. Not surprisingly, a variety of methodologies have been proposed to analyze the trade/poverty issue, which suggests that the range of findings will be nearly as diverse. The most obvious methodological gulf is between researchers who have come at these issues from a tradition of measuring poverty using detailed household expenditure data, and those who are of a trade background and more accustomed to dealing with economywide data. One might refer to these as "bottom-up" and "top-down" approaches, respectively. The former emphasizes the heterogeneity of individuals and households as revealed through surveys, while the latter builds on the microeconomic assumption of a representative agent. Since most studies focus on a single country, it is difficult to distinguish the degree to which findings are driven by methodological assumptions as opposed to characteristics of the particular population in question. In spite of the methodological diversity of the studies, there appears to be increasing recognition that any analysis of trade, trade policy, and poverty needs to come to grips with the issue of factor market 1 effects. This was one of the key conclusions of the October 2000 Conference on Poverty and the International Economy, organized by the Swedish Parliamentary Commission on Global Development and the World Bank'. This observation has also been made in recent empirical studies by Coxhead and Warr (1995), Harrison, Rutherford, and Tarr (2000), Hertel, Preckel, Cranfield, and Ivanic (2001), and Warr (2001), each discussed in more detail below. The importance of factor market effects arises because households tend to be much more specialized with regard to factor earnings (that is, income derived from productive factors such as labor, capital, and land) than they are with regard to consumption. Accordingly, this paper emphasizes how current analyses address the factor income side of the trade/poverty issue. For papers that incorporate factor earnings effects, particular attention is given to the sources of the data, and how researchers link factor income to individual households or household types. The primary means of describing this is via Tables 1 through 4 at the end of the paper, which summarize the objective, use of earnings data, general methodological approach, and conclusions of each paper. To keep the survey manageable, and to avoid undue repetition of what has already been covered in other surveys, a set of criteria for inclusion was adopted. First of all, papers from the extensive literature on trade and wages are excluded since they are typically concerned with labor market and income distribution issues in developed instead of developing countries. Moreover, a number of excellent overview papers on that topic already exist, including Wood (1995) and Slaughter (1999). Within the realm of trade/poverty studies, this survey places emphasis on analyses that involve some sort of "counterfactual" simulation or regression analysis, as opposed to those only documenting how poverty has evolved over time. Simulation is stressed because it facilitates understanding of the links between a specific shock and poverty, holding other factors constant (indeed, the vast majority of trade/poverty studies employ some form of simulation analysis). While all the studies in this survey focus on the poor in developing countries, they do not all involve a change in trade policy. For example, a few papers on technical change and economic growth were included because their frameworks are highly relevant to XA web site describing the conference is: http://wwwl .worldbank.or2lwbiep/trade/povertvconf.html. 2 trade policy analysis. Additionally, this survey places emphasis on studies that have an empirical rather than theoretical focus, and have been carried out within the last 10 years or so. Indeed, most of the papers surveyed are not yet in published form. In collecting studies to include in this survey, it quickly became apparent that there is no obvious or ideal means to categorize them, since they differ in a number of significant ways. Studies vary across many dimensions, such as whether the analysis is carried out for representative households or actual households (i.e. microsimulation), whether it is static or dynamic, single- or multi-region, partial- or general-equilibrium, and so forth. Out of these possibilities, four broad categories of study were identified based upon the principal methodology employed. The first methodological classification is for studies that undertake cross- country regression analysis. These studies test for correlations among trade, growth, income, poverty, and inequality variables observed at the national level. The second category encompasses a wide array of partial-equilibrium and/or cost-of-living approaches. These studies are typically based on household expenditure data, and generally emphasize commodity markets and their role in determining poverty impacts, or at least as a measure of poverty across time. Studies in the third category all involve some form of general equilibrium model that accounts for commodity, terms of trade, and factor market effects. These studies are usually based on a disaggregated economywide Social Accounting Matrix. The fourth and final category represents a relatively recent approach - general equilibrium simulation coupled with some form of post-simulation analysis based on household survey data. These studies may be thought of as micro-macro synthesis. While the term "micro-macro" has been used differently in other contexts, in this paper it is meant to refer to the sequential linking of a model based on micro-level data with a model based primarily on macro-level data. 3 II. The Factor Income Link Before describing each of the methodological approaches in turn2, it is useful to consider the linkages that exist between trade, trade policy, and poverty. In a comprehensive paper on this topic, L. Alan Winters (2000) identifies several key linkages, which are reiterated in large part by Bannister and Thugge (2001). Potential links include changes in: (a) the price and availability of goods; (b) factor prices, income, and employment; (c) government transfers influenced by changes in revenue from trade taxes; (d) the incentives for investment and innovation, which affect long-run economic growth; (e) external shocks, in particular, changes in the terms of trade; (f) short-run risk and adjustment costs. Most studies focus on only one or two of these linkages, while abstracting from the rest. Nearly all of the studies in this survey consider the consumption side of the trade-poverty linkage (a). Linkages (b) through (f) tend to be less frequently considered. A study by Levin (2000) focuses on transfers, link (c). A number of economy-wide analyses account for terms of trade effects, link (e). Each study typically abstracts from at least two of the linkages in order to keep the model tractable, and because the necessary data may not have been available. When reading a paper one should keep in mind which linkages are excluded from the analysis, and how this may influence the results. As suggested in the introduction, the factor price, income, and employment link (b) may have the greatest relative importance of all the links between trade and poverty. Household survey data used in the Hertel, Preckel, Cranfield, and Ivanic study (described in section VI) as well as casual observation suggests that people tend to be much more heterogeneous with respect to income than with respect to consumption. In other words, two households may have identical commodity budget shares, and the same level of income, but entirely different sources of income; e.g., one derives all income from 2 Papers included in the survey are categorized and summarized at the end in Tables I through 4. 4 agricultural labor, while the other relies on transfers from a relative who works abroad. This point is underscored by the fact that opposition to free trade initiatives often arises from groups with highly specialized income, such as steel workers and sugar farmers in the U.S., to name just two examples. Within the world of classical trade theory, income effects are key to the famous Stolper- Samuelson theorem, which relates international trade to the domestic distribution of income (Dixit and Norman). By the Heckscher-Ohlin theorem, a country has a comparative advantage in the good that intensively uses the country's relatively abundant factor. Free trade will increase the relative price of that good and so, by the Stolper-Samuelson theorem, increase the real return of the relatively abundant factor by an even larger percentage. At the same time, trade will reduce the return to the relatively scarce factor, though to a smaller degree. As a result, it can be said that changes in commodity prices due to trade liberalization magnify the resulting changes in factor prices. The presence of this Magnification Effect (due to Jones, 1965) in theoretical trade models is one reason why trade economists tend to focus on factor market effects when analyzing trade liberalization and poverty. Some (e.g. Winters, 2000) have argued that the practical relevance of the Stolper- Samuleson/Magnification result is negligible, since it rests on so many restrictive assumptions as to be a special case. Nevertheless, this theoretical insight underscores the importance of considering factor earnings effects when examining the relationship between trade liberalization and poverty. Three empirical studies reinforce this view. A general equilibrium analysis of technical change in the Philippines by Coxhead and Warr (1995) found earnings effects to be substantially more important than consumption effects. In particular, income effects accounted for two-thirds of poverty alleviation when there was a rise in agricultural productivity. While this is not a trade liberalization study, the nature of the shock is not dissimilar since the adjustments are transmitted through commodity and factor markets. Harrison, Rutherford, and Tarr (2000) find that factor price changes drive the incidence of trade liberalization in Turkey. They demonstrate this by employing three counterfactuals in which the 40 representative households in the analysis (differentiated by rural/urban orientation and by income level) 5 have (i) identical consumption shares, (ii) identical factor income shares, and then (iii) identical consumption and factor income shares. Since counterfactual (i) provided nearly identical results to those generated when the heterogeneity of the 40 households is left intact, the authors conclude that "clearly, for the poor it is the source of income, not the pattern of expenditure that is driving the adverse impact relative to the average household" (p. 12). A general equilibrium analysis by Warr (2001) of Thailand's proposed rice export tax also suggests that factor eamings effects are the driving force behind welfare and distributional effects. Although an export tax generates government revenue and lowers the price of rice for consumers, it also lowers the return to unskilled labor, which is used intensively in the Thai rice industry. Because both the rural and urban poor derive more than 40 percent of their income from unskilled labor (according to the Thai survey upon which the stylized households are based), the negative income effect ends up outweighing the consumption benefit, such that both the rural and urban poor are harmed by the export tax. Despite the apparent importance of factor earnings effects, they are often not accounted for in studies that quantify the effects of external shocks on the poor in developing countries. This is particularly the case for analyses based on detailed household surveys, at least historically. Because abstracting from this particular linkage may be quite misleading, this survey will pay particular attention to how each analysis deals with the income side of the story. At the same time, the issue of whether a focus on "factor markets" is the same as a focus on "income" is not explored in depth here. It can be argued that many of the poor are subsistence farmers and largely disconnected from markets, or that their well being is largely determined by their net trade position in a food commodity such as rice. Studies that explore this latter issue in more detail include Ravallion (1990) and Ravallion and van de Walle (1991). As to the importance of thinking about a household's income in terms of commodities versus factors, Hertel, Preckel, Cranfield, and Ivanic (2001) provide interesting survey evidence on this issue for seven developing countries in Figures 1-21 of their paper. 6 III. Cross-Country Regression As discussed above, four general methodologies are in current use for estimating the poverty impacts of trade liberalization. The first approach considered in this survey is cross-country regression, as exemplified in a recent paper by Dollar and Kraay (2001). These authors first categorize developing countries as either globalizers or non-globalizers based on changes in trade volumes and tariff rates since 1980, then carry out case study as well as statistical analysis. Looking at anecdotal evidence on poverty, including time-series Gini coefficients and income growth rates for average households versus the poorest quintile, they find no general trend in inequality among countries classified as globalizers. Globalizers, however, tend to have higher rates of growth than non-globalizers. This leads to the conclusion that globalization tends to be associated with a decline in absolute poverty. Verifying these findings in a more rigorous manner, the authors undertook cross-country regression analysis, and determined that no systematic relationship exists between changes in trade volumes and changes in the income share of the poorest. Additionally, no statistical relationship between changes in trade volumes and changes in income inequality could be found. Rodrik (2000) offers a cogent critique of Dollar and Kraay's study. In general his remarks relate to issues with the data, to the difficulty of distinguishing between correlation and causation in cross- country regression analysis, and to the challenge of obtaining results that are robust to specification changes. Estimating the relationships that exist between trade policy, growth, and poverty depends critically on finding appropriate measures of these variables, and carefully sorting out omitted variable and endogeneity problems, all of which are quite challenging given the very limited data available. The fact that Dollar and Kraay include results obtained using Instrumental Variables provides some reassurance against Rodrik's critique. Most trade and poverty researchers forego cross-country regression analysis and instead carry out some form of simulation analysis (this can be verified by comparing the very limited number of papers in 7 Table 1 with the much larger number of papers in Tables 2 - 4). The hallmark of simulation analysis is the use of a counterfactual, which literally means "contrary to the facts" and enables investigation of "what might have been" had a certain shock taken place. The great advantage of counterfactuals is that the effects of a specific shock can be isolated from the effects of all other events occurring during the period of interest. Counterfactual analysis, therefore, provides an elegant means of avoiding the identification problems inherent to cross-country regression, while allowing the researcher to pose specific policy questions once the appropriate simulation model has been operationalized. The cross-country regression approach nevertheless has a number of advantages for understanding the links between trade and poverty. First of all, it enables the use of traditional statistical tools for testing results and hypotheses, as opposed to only making predictions3. Secondly, cross-country regression results are typically much more general than the country-specific results of many applied simulation models. Thirdly, cross-country regression may be able to account for some of the dynamic aspects of trade reform that are missed by static simulation models. Given the differing advantages and disadvantages associated with the cross-country regression and simulation approaches, they should probably be viewed as complementary forms of analysis as opposed to substitutes. IV. Partial-Equilibrium/Cost-of-Living Analysis The second general methodology identified as a means of estimating the poverty impacts of trade liberalization is partial-equilibrium/cost-of-living analysis. The awkwardness of this characterization reflects the fact that more than one type of study is included in this category. In general, however, all of the studies in this category are "partial equilibrium" in nature, since they focus on one or a limited number of markets in an economy. Additionally, most can be considered "cost-of-living" studies since they tend to focus on household expenditure as a measure of poverty. 3At the same time, techniques such as Systematic Sensitivity Analysis are available and increasingly used for assessing the robustness of results from calibrated simulation models. 8 The majority of studies in this category can also be regarded as microsimulation models. Microsimulation is distinguished by a focus on behavior at the individual or household level, as opposed to using any sort of representative household. As such, individual or household survey data are key to applications of the microsimulation approach. It should be noted that microsimulation is sometimes associated with general equilibrium contexts as well (see, for example, the studies by Cogneau and Robilliard, and Cockbum). A great many papers fit into the partial-equilibrium/cost-of-living analysis category (see Table 2 for the complete list). One fairly representative approach is by Levinsohn, Berry, and Friedman (1999), who examine how the Indonesian economic crisis affected poor households in that country. The authors combined 1993 consumption data for 58,100 households from the Susenas survey, along with price changes due to the 1997-1998 crisis, to compute household-specific cost-of-living changes. The salient findings were that very low income households were not insulated from the international shocks, and in fact tended to be hurt the most. Regardless of being urban or rural, households at lower expenditure levels experienced larger cost-of-living increases (a relationship that is monotonic). Additionally, the consumer price impacts of the crisis were greater for urban than for rural areas, and greatest overall for the urban poor. From a methodological perspective, the Levinsohn, Berry, and Friedman analysis has two principal drawbacks. First, focusing only on the consumption side of the crisis (link (a) in section II) precluded calculation of its real effects. This may not have been so critical for this particular application, since increases in nominal wages were overshadowed by increases in general commodity prices (an average of 26.0% versus 92.5% according to Levinsohn, Berry, and Friedman). However, studies focusing on trade liberalization generally find factor market effects to be at least as important as commodity market effects. Secondly, the Levinsohn, Berry, and Friedman analysis did not allow the effects of the crisis to be isolated from other phenomena, including the El Nino drought and widespread 9 forest fires that occurred in the same period as the crisis. This drawback could have been avoided in a model enabling the specification of counterfactual simulations4. Another methodological limitation of significance to this survey was that household expenditure shares in the Levinsohn, Berry, and Friedman study were assumed to stay fixed throughout the crisis. Changes in demand due to changes in income or the prices of other goods were ignored. In terms of a household's demand schedule for a given good, movements along as well as shifts in the demand curve were precluded. Estimation of a demand system, particularly one that is non-homothetic, would have avoided this issue. Another limitation is that the expenditure shares were outdated, since Indonesia in 1997 had changed substantially from where it had been in 1993. The authors point out, however, that the consumption baskets of pdor households relative to rich ones, and rural households relative to urban ones, likely did not vary much over this period. This is relevant because the authors were primarily interested in assessing the relative impact of the crisis across income levels, and between rural and urban areas. Another approach to trade, price changes, and poverty is provided by Case (1998). She quantifies the extent that trade reform in South Africa will affect households as consumers, using household budget shares and estimates from a Linear Expenditure System estimated separately for Africans and Whites. Budget shares and the demand system estimates were calculated using the nationally representative 1993 South African Living Standards Survey, which covers 43,794 individuals in 8,848 households drawn from 360 clusters. Using outside estimates of the price changes following tariff reform, it is found that the cost of reaching the household's initial level of utility falls by roughly 2 percent for African households and by 1 percent for White households. As with the Levinsohn, Berry, and Friedman study, potential factor eamings effects do not enter into Case's analysis, despite the availability of employment and income information in the household survey. A third example of how partial equilibrium models are being used to address trade and poverty issues is Minot and Goletti (2000), who offer an extensive examination of how rice market liberalization 4 See, for example, Robilliard, Bourguignon, and Robinson (2001) for an alternative assessment of the Indonesian 10 in Viet Nam may affect income and poverty in that country. They employ a variety of methods to reach their research objective, including descriptive analysis based on surveys of rice producers, traders, and other market participants; time-series analysis of rice prices and production; and estimation of household demand behavior based on the nationally representative 1992-93 Viet Nam Living Standards Survey of 4,800 households. Households in poverty are defined as those below the 25th percentile in terms of per capita expenditure, and results are provided in terms of the Foster-Greer-Thorbecke poverty index as well. The centerpiece of Minot and Goletti's analysis is a multimarket spatial equilibrium model that is used to conduct a series of policy experiments, including (i) removing the rice export quota, (ii) changing the quota level, (iii) replacing the quota with a tax, and (iv) removing restrictions on the internal movement of food. The distributional consequences of these counterfactuals are determined by way of the net rice sales position of different household classes. It is found that export liberalization raises rice prices within the country, particularly in the country's rice exporting areas. The higher prices have a positive effect on rural incomes, and are generally favorable with regard to the number of people in poverty. Relaxing the restrictions on the internal movement of rice from south to north generates net benefits for the country, without increasing most measures of poverty. Since rice production is quite labor intensive in Viet Nam, a rise in rice prices should increase demand for agricultural labor, and consequently the agricultural wage rate. Higher rice prices would then lead to a greater decrease in poverty, particularly in households that derive a share of their income from agricultural labor. Unfortunately, Minot and Goletti's counterfactual analysis assumes that labor demand and wage rates remain constant. While they point out that landlessness and the use of hired labor are not widespread in Viet Nam, inclusion of a factor earnings link (b) would have quantified this perception. crisis. V. General-Equilibrium Simulation If a researcher is interested in how trade liberalization will affect only a limited number of an economy's markets, needs to incorporate a great amount of sectoral detail, or has limited time available, then partial-equilibrium/cost-of-living analyses are logical approaches. They also have the advantage of being easier to understand than general equilibrium modeling. When examining the question of poverty, however, partial-equilibrium/cost-of-living analysis usually requires a researcher to abstract from the income side of the issue, or limit the analysis to consideration of a single factor (typically labor). The focus on commodity markets is due in part to a traditional lack of good data on household earnings, since in household surveys income information tends to be less complete and less reliable than expenditure information (Cockbum; Hertel et al.). However, several recent empirical studies provide evidence that - regardless of the data limitations - this abstraction is not innocuous5. General equilibrium analysis of poverty and distribution issues in a developing country context has its origins in work by Adelman and Robinson for Korea (1978), along with Lysy and Taylor for Brazil (1980). General equilibrium models are now widely used to assess the impact of economic shocks that reverberate across sectors and, in some cases, regions of a country or even the world. They are capable of producing disaggregated results at the microeconomic level, while providing a consistency check on macroeconomic accounts. A general equilibrium model is generally calibrated to a Social Accounting Matrix, which is a complete, consistent, and disaggregated data system. The salient feature of Social Accounting Matrixes is that they quantify - at a single point in time - the interdependence of sectors and regions in an economy. General equilibrium models are typically based on neoclassical theories of firm and household behavior, and have a time frame long enough to achieve equilibrium in markets. While most are comparative static in nature, dynamic versions have also been developed to address certain types of issues. A study by Lofgren (1999) is representative of how applied general equilibrium models are 12 currently being used to analyze trade and poverty issues. Lofgren investigates how reduced agricultural and industrial protection will affect representative Moroccan households in the short run. The general equilibrium model is multi-sector, single-region, static, and calibrated to a 1994 Social Accounting Matrix6, which captures the pronounced rural/urban disparity in economic structure, wages, and education that is characteristic of Morocco. Four household groups are distinguished according to whether they are rural or urban, poor or non-poor. Unlike two studies examined below in this section, the distribution of income within the groups is not modeled, as the study does not seek to make statements about the total income distribution. Based on information in the Social Accounting Matrix, Lofgren divides factor markets into four types of agricultural resources, five types of capital, and a variety of labor types, differentiated by skill, urban/rural orientation, and use in agriculture. Production is specified as a Leontief function of aggregate value-added and an aggregate intermediate input, which are a constant elasticity of substitution (CES) function of primary factors, and a Leontief function of intermediate inputs, respectively. Consumer demand is represented by the Linear Expenditure System. The model relies on standard neoclassical assumptions and is set up in "real" terms, such that there are no asset markets, money is neutral, and all agents make decisions as a function of relative prices. Lofgren's simulations assess the impact of removing border protection under different assumptions about labor market rigidity. The essential results are that trade liberalization in agriculture will result in gains for the country as a whole, while the rural poor loses out. Compensation in the form of government transfers as well as education and infrastructure investments for rural areas would likely be needed if liberalization were to be pursued. 5See Coxhead and Warr (1995), Harrison, Rutherford, and Tarr (2000), Warr (2001), and Hertel, Preckel, Cranfield, and Ivanic (2001). 6 The Social Accounting Matrix - upon which the credibility of the results hinges - was developed using a numerous data sources, including statistical publications from the Moroccan government, the World Bank (including the RMSM data base), the Food and Agricultural Organization, the International Monetary Fund, various Royaume du Maroc statistical volumes (see Lofgren's paper for details). The procedures for designing the stylized households from survey data are not discussed. 13 On the methodological side, Lofgren finds that the results are strongly influenced by the commodity, factor, foreign exchange, and government budget links between agriculture and the rest of the economy, which correspond to links (a), (b), (c), and (e) listed in section II of this survey. Of all the potential linkages identified by Winters (2000), Lofgren's analysis excludes only the investment and innovation link (d), and risk and adjustment cost link (f). Ignoring these two effects would likely result in systematic underestimation of the long-run benefits and short-run costs of trade liberalization, respectively. Determining the ultimate importance of these linkages would require specification of a dynamic model. Lofgren's general approach is more or less representative of a large number of trade and poverty studies carried out over the past decade (see Table 3). One variant of this basic paradigm is to address in greater detail how external shocks affect the total income distribution of a country. For this purpose it is necessary to postulate a distribution of income for each representative household type (as in Adelman and Robinson, 1978) or to work at the level of actual households (as in Cogneau and Robilliard, and Cockburn). If a distribution is assumed a priori, it can then be used in conjunction with the general equilibrium model to assess the impact of exogenous shocks on the income distribution of a country, as well as poverty. In this framework, it is typically the case that the mean and total income levels for a household group are endogenous while the higher moments of the distribution are fixed. In an interesting paper, Decaluwe, Patry, Savard, and Thorbecke (1999) consider this basic approach and provide some refinements- to it. They model an archetype African economy with two agricultural activities, four non-agricultural activities, and six representative household groups. One of the innovations is the use of a flexible Beta functional form to model the income distribution within household groups, instead of the more common - and restrictive - lognormal or Pareto distributions. The parameters of the Beta distribution are specified to conform to observed socio-economic characteristics of each household type, and it is shown that the shape of the distribution may indeed vary markedly across them. Another of the model's refinements is the specification of a poverty line in the LES demand 14 system based on a unique and fixed bundle of basic-needs commodities. Because commodity prices are endogenously determined, the poverty line is as well. Although no empirical results are presented, Decaluwe, Patry, Savard, and Thorbecke suggest that their innovations will "help shed more light on the black box pertaining to the behavior of poverty following a shock". The authors also emphasize in the first part of their paper that Social Accounting Matrixes can be used on their own to analyze issues related to income distribution, and to a lesser extent, poverty. This involves the use of accounting multipliers in conjunction with information on the factor income of disaggregated household types. Another approach to trade, poverty, and income distribution modeling is offered by Cogneau and Robilliard (2000). In many ways their general equilibrium model is fundamentally different from the general equilibrium models described above. Their aim is to assess the impact of different growth strategies on welfare and poverty in Madagascar. To meet this goal they embed an econometrically estimated labor allocation model based on 4,508 households within a general equilibrium framework. The combination of a microsimulation and general equilibrium model facilitates the modeling of a country's overall income distribution, since it is no longer necessary to a priori assume an income distribution for each household type7. The combination of approaches also allows endogenous variables to be determined at the level of individual households, thereby eliminating the representative household assumption (for the most part) and its associated theoretical shortcomings. Three aggregate sectors of the Madagascar economy are modeled: a formal sector that produces a tradable commodity, an informal sector that produces a non-tradable, and an agricultural sector that produces both a tradable and non-tradable. Productive factors include labor, agricultural capital, and formal sector capital. Agricultural and informal activity is endogenous and determined at the household level, as is agricultural labor demand. Informal labor demand is determined at the aggregate level based on demand for the informal good and agricultural labor. The supply of labor to the agricultural and 15 informal sectors is endogenous and determined at the individual level using the labor allocation model. Consumer demand is modeled with the Linear Expenditure System. Macroeconomic data are from a 1995 Social Accounting Matrix constructed by Razafindrakoto and Roubaud (see the paper for more details). Microeconomic data, covering 4,508 households, are from the 1993 EPM (Enqu8te Permanente aupres des Menages) survey carried out by INSTAT (the Institut National de la Statistique) on behalf of the Malagasy government. Although endogenous variables are based on individual household behavior via the microsimulation model, results of the simulations are presented in terms of 14 representative households. Four of the representative households are urban, and differentiated according to educational attainment and gender. Eight households are rural, agricultural, and differentiated according to region and farm size. The remaining two types of households are rural, nonagricultural, and distinguished according to wealth. Although it is not clear from the paper, these typologies appear to be based on the 1993 EPM survey data set and the 1995 Social Accounting Matrix. These same data sources also provide information for the disaggregation of household income. Earnings are based on receipts from agricultural labor, informal labor, formal sector labor, capital dividends, sharecropping income, and transfers from other households or the government. Cogneau and Robilliard consider six counterfactuals, including (i) an increase in formal sector labor demand, (ii) an increase in formal sector wages, (iii) an increase in agricultural productivity, (iv) an increase in food crop productivity, (v) an increase in cash crop productivity, and (vi) an increase in the world cash crop prices. While relative income and price changes are significant in most simulations, the effect of shocks on poverty and inequality are small. The authors identify several reasons for this finding, including the unequal distribution of productive factors across households, and the ability of households to diversify their income sources through reallocation of productive activity. 7 It is typically assumed that the income distribution within a representative household group is lognormal with endogenous mean and fixed variance. 16 Cogneau and Robilliard's analysis is a unique melding of microsimulation and general equilibrium modeling8. Basing the analysis on actual households facilitates the study of income distribution, since restrictive assumptions about within-group distributions and certain other aggregation issues can be avoided. Working with actual households also lends an air of realism, and allows for the possibility that there is considerable heterogeneity across households. Meanwhile, incorporation of general equilibrium mechanisms captures the redistributive effects of shocks on both sectors and households. These accomplishments do entail higher data requirements and computational costs, however.. Working with 4,508 agents requires other model dimensions to be scaled back, since, for example, the income of each agent needs to be tied to each commodity represented. As a result, the sectors and commodities of Cogneau and Robilliard's model are highly aggregated, and a number of critical macroeconomic features are ignored. Another consideration is that it is not practical to inspect the impact of a simulation on each of several thousand households. Accordingly, results must be aggregated and analyzed for a limited number of representative households, just as in conventional general equilibrium models. VI. Micro-Macro Synthesis While the approach of Cogneau and Robilliard is innovative, there are other ways to capitalize on the detail of household survey data while availing the ability of general equilibrium models to capture the numerous links between trade and poverty. A somewhat simpler, more pragmatic means to the same end is offered by the studies in this fourth and final category of the survey, which, for lack of a better label, is entitled "micro-macro synthesis". An alternative description might be "general equilibrium simulation with post-simulation analysis". This approach is best characterized by its sequential, two-step nature. In general, a general equilibrium model is first shocked to get commodity and factor price changes. These s Two other studies that incorporate large numbers of actual households into a relatively standard general 17 are then fed into or calibrated to a post-simulation framework that calculates the effects on actual or highly disaggregated representative households; Various poverty measures can then be applied to assess the distributional effects of the shocks. This two-step approach is similar to that employed in some partial equilibrium studies (e.g. Case 1998, Deaton 1989), except that in partial equilibrium analyses the price changes are typically for consumer goods only, and are purely hypothetical or based on real-world observations - in other words, not from a counterfactual simulation. A limitation of post-simulation analysis, at least in the view of general equilibrium practitioners, is that the reactions of households to commodity and factor price changes in the post-simulation analysis are not transmitted back to the general equilibrium model. Although this absence of feedback is not satisfactory from a theoretical point of view, the resulting error is likely to be small. Robilliard, Bourguignon, and Robinson (2001) is one of a growing number of micro-macro studies to recently emerge. As in the Levinsohn, Berry, and Friedman paper, the authors study the effects of the 1997 Indonesian crisis on poor households. The general equilibrium model is based on a single- region Social Accounting Matrix that captures macroeconomic constraints along witlx intersectoral flows for 38 sectors and 15 factors of production. The post-simulation analysis is a microsimulation model based on the 1996 Susenas survey, with 33,000 individuals in 9,800 households. Conducting the analysis with actual households facilitates calculation of changes in the income distribution, since one can avoid strong assumptions about intragroup distributions and certain other aggregation issues. Robilliard, Bourguignon, and Robinson's microsimulation model represents the way in which households generate their income, by focusing on how earnings are determined and how occupational choices are made. Workers are divided into eight groups according to skill, gender, and area of residence. Functions corresponding to wage worker earnings, farm and non-farm worker profits, and occupational equilibrium model are G0rtz, Harrison, Nielsen, and Rutherford (2000), and Cockbum (2001). 18 choices are estimated. Labor supply is modeled as a discrete choice between inactivity and full time work. The general equilibrium model relies on standard neoclassical assumptions and is set up in "real" terms, with no asset markets, neutral money, and decisions based on relative prices. The model is dualistic in that it distinguishes between formal and informal activities in each sector, both of which produce the same good. Eight labor categories, six types of capital, and 10 household types9 are distinguished, along with macro accounts for enterprises, govemment, the rest of the world, and for savings-investment. The real wage is assumed to be fixed in formal-sector labor markets, while informal- sector labor markets absorb any labor displaced from the formal sectors. The general equilibrium model is linked to the microsimulation model through (i) the wage level in each wage labor market, (ii) the income level for the informal self-employed sector, (iii) the number of wage workers and self-employed by labor market segment, and (iv) consumption prices. The microsimulation model is solved so that it generates equilibrium values and changes that are consistent with the results from the general equilibrium model. Simulations were carried out to (a) decompose and reproduce the crisis impact, (b) examine how the Indonesian economy would have fared with the same adjustment in trade balance but no credit crisis, and (c) examine different policy options, including a food price subsidy, a public work program for unskilled workers, and transfers to target groups. It is found that poverty increases over the 1997-98 period were due in equal measure to the El Nino drought and to the financial crisis (a very different perspective from that of the Levinsohn, Berry, and Friedman paper). The second set of experiments suggests that some of the available policy options would have resulted in a smaller increase in poverty. On the methodological side, the Robilliard, Bourguignon, and Robinson approach is somewhat costly since the unit of analysis is an actual household, and a great deal of estimation work is required. To assess the benefits of this approach, they carried out the analysis using representative households to 19 compare. They determine that a representative household assumption biases most experiments, and leads to incorrect results in the case of targeted policies. In particular, the representative household approach appears to systematically underestimate the effect of the shocks on income inequality and poverty. Another interesting approach to trade and poverty issues is offered by Hertel, Preckel, Cranfield, and Ivanic (2001). They examine how global trade liberalization affects poverty in each of seven different developing countries. While they center their analysis on factor market effects, they also allow for commodity market and terms of trade effects (altogether incorporating links (a), (b), and (e) described in section II). The first step of the authors' analysis involves conducting a policy experiment in the Global Trade Analysis Project (GTAP) model of trade (Hertel, 1997) to generate a vector of factor and commodity price changes for 17 regions of the worldl'. Since the GTAP database is designed for broad country coverage, it is limited to one representative household per region - clearly not adequate for an investigation of poverty. The price changes are therefore fed into a post-simulation framework that characterizes households according to factor income and consumption profiles, which are based on International Comparison Project data, and household surveys for seven countries, respectively One of the authors' most striking findings is the extent to which households in each of the seven countries are specialized in terms of factor earning profiles. To capture the consequent vulnerability to trade liberalization, households are categorized into five strata, including those getting at least 95% of income from (i) transfers, (ii) agriculture, (iii) non-agricultural business, (iv) wages, and then (v) a stratum for households that have diversified income sources. Within each stratum, the differences across income levels are preserved. 9 Since factor price changes are passed on directly to the microsimulation model, disaggregation of households in the CGE model is not necessary, but is used for comparison with the microsimulation approach. 10 The experiment involves complete elimination of merchandise tariff barriers as well as textile and apparel quotas in place in 1997. 1 " The household surveys are for Brazil, Chile, Indonesia, Philippines, Thailand, Uganda, and Zambia, and are available through the World Bank. 20 Changes in real household incomes are calculated, and demand response is simulated by feeding commodity price changes into an estimated global AIDADS demand system12. The demand system is used to calculate the poverty level of utility for each region. Equivalent variation (EV) and a first-order compensating variation (CV) measure are then calculated at both the per capita and poverty line levels. Since the CV approximation proves to be quite accurate compared to the exactly computed EV, it is used to decompose the results into underlying commodity and factor market adjustments. The Foster-Greer- Thorbecke measure of poverty is used to calculate the total transfer required to lift all households above the poverty level of utility, as a proportion of the poverty level of income. Hertel, Preckel, Cranfield, and Ivanic's findings suggest that multi-lateral trade liberalization will reduce overall poverty in Indonesia, Philippines, Uganda, and Zambia, but increase overall poverty in Brazil, Chile, and Thailand. Within regions, the results vary considerably by household group. The largest poverty reduction occurs among agriculture-specialized households in Brazil, while the largest increase occurs among non-agricultural, self-employed, and wage-labor households in Brazil, Chile, and Thailand. VII. Summary and Conclusions Quantifying the poverty impacts of trade liberalization and related extemal shocks is currently an area of intense research, and a variety of methodologies are being employed to address the issues involved. This survey provides a review of methods in current use, and classifies them into four broad categories, namely (i) cross-country regression analysis, (ii) partial-equilibrium/cost-of-living analysis, (iii) general-equilibrium simulation, and (iv) micro-macro synthesis (also referred to as general- equilibrium simulation with post-simulation analysis). These four groups encompass both the "bottom- up" and "top-down" traditions that are associated with poverty and trade specialists, respectively. The former approach builds on detailed survey information, and emphasizes the heterogeneity of individual 12 AIDADS is a generalization of LES, allowing for the possibility of non-linear, non-monotonic Engel effects. 21 households as well as commodity market linkages between trade and poverty. The latter approach begins with the representative household assumption from microeconomic theory, and generally incorporates additional linkages between trade and poverty such as factor earnings and terms of trade effects. The general conclusion of this survey is that any analysis of trade and poverty needs to be informed by both perspectives. Indeed, the most recent and innovative studies sequentially link the top- down and bottom-up approaches in a two-step procedure, such that general equilibrium mechanisms are incorporated along with detailed household survey information. While not necessarily elegant in a theoretical sense, this approach accounts for the majority of trade-poverty linkages, is based on solid empirical foundations, and is compatible with multi-region trade modeling. Another methodology in the same spirit and also increasingly used involves the incorporation of large numbers of actual households into a general equilibrium simulation model. This framework allows for the possibility of substantial heterogeneity across households within a region, while maintaining feedback effects between those households and the rest of the economy. Although these studies have so far been limited to a single region, there is no reason this approach cannot be adapted to multi-region modeling, such that trade-poverty comparisons can be made across countries in a consistent manner. These developments have come through increased recognition that households are typically much more specialized in income than in consumption - making factor markets the key linkage between trade liberalization and poverty - and through the increased availability of household survey information on factor earnings. 22 References Adelman, Irma, and Sherman Robinson. 1978. Income Distribution Policy. A Computable General Equilibrium Model of South Korea Stanford: Stanford University Press. Adelman, Irma, and Sherman Robinson. 1988. "Macroeconomic Adjustment and Income Distribution: Alternative Models Applied to Two Economies." Journal of Development Economics 29(1): 23- 24. Appleton, Simon. 2001. "Poverty Reduction During Growth: the Case of Uganda, 1992-2000." Unpublished manuscript, University of Nottingham. Bannister, Geoffrey, and Kamau Thugge. 2001. "International Trade and Poverty Alleviation." Finance and Development, Vol. 38, (December) No. 4. Bautista, Romeo M., and Marcelle Thomas. 1997. "Income Effects of Alternative Trade Policy Adjustments on Philippine Rural Households: A General Equilibrium Analysis." IFPRI TMD Discussion Paper 22. Bourguignon, Francois, Jaime de Melo, and Akiko Suwa. 1991. "Distributional Effects of Adjustment Policies: Simulations for Two Archetype Economies." World Bank Trade Policy Working Papers, WPS 674. Bourguignon, Francois, William Branson, and Jaime de Melo. 1989. "Adjustment and Income Distribution: A Counterfactual Analysis." World Bank Trade Policy Working Papers, WPS 215. Case, Anne. 1998. "Income Distribution and Expenditure Patterns in South Africa." Paper prepared for the Conference on Poverty and the International Economy, organized by World Bank and Swedish Parliamentary Commission on Global Development, Stockholm, October 20-21, 2000. Cockburn, John. 2001. "Trade Liberalisation and Poverty in Nepal: A Computable General Equilibrium Micro Simulation Analysis". CREFA Working Paper (01-18). Available at: www.crefa.ecn.ulaval.ca/cahier/0 118.pdf. 23 Cogneau, Denis, and Anne Sophie Robilliard. 2000. "Growth, Distribution and Poverty in Madagascar: Learning from a Microsimulation Model in a General Equilibrium Framework." IFPRI TMD Discussion Paper 61. Coxhead, Ian and Peter Warr. 1995. "Does technical progress in agriculture alleviate poverty? a Philippine case study." Australian J Agr. Econ., 39(1), April, pp.25-54. Cranfield, John A.L., Thomas W. Hertel, and Paul V. Preckel. 2000. "Multi-lateral Trade Liberalization and Poverty." Paper prepared for the Conference on Poverty and the International Economy, organized by World Bank and Swedish Parliamentary Commission on Global Development, Stockholm, October 20-21, 2000. Deaton, Angus. 1989. "Rice Prices and Income Distribution in Thailand: A Non-parametric Analysis." The Economic Journal 99 (395, Conference Supplement): 1-37. Deaton, Angus, and Alessandro Tarozzi. 2000. "Prices and Poverty in India." Unpublished manuscript. Available at: http://www.wws.princeton.edu/rpds/working.htm. Decaluw6 B., A. Patry, L. Savard and E. Thorbecke. 1999. "Poverty Analysis within a General Equilibrium Framework." CREFA Working Paper 9909. Universite Laval. Available at: http://www.crefa.ecn.ulaval.ca/cahier/liste99.html. Dercon, Stephan. 2001. "The Impact of Economic Reforms on Households in Rural Ethiopia 1989-1995." Unpublished manuscript, Centre for the Study of African Economies, Dept. of Economics and Jesus College, Oxford University. Dervis Kemal, Jaime De Melo, and Sherman Robinson. 1982. General Equilibrium Models for Development Policy. Cambridge: Cambridge University Press. Devarajan, Shantayanan, and Dominque van der Mensbrugghe. 2000. "Trade Reform in South Africa: Impacts on Households." Paper prepared for the Conference on Poverty and the International Economy, organized by World Bank and Swedish Parliamentary Commission on Global Development, Stockholm, October 20-21, 2000. 24 Dixit, Avinash, and Victor Norman. 1980. Theory of International Trade. Cambridge: Cambridge University Press. Dollar, David and Aart Kraay. 2000. "Growth is Good for the Poor." World Bank Policy Research Working Paper 2587. Dollar, David and Aart Kraay. 2001. "Trade, Growth, and Poverty." World Bank Policy Research Working Paper 2615. Evans, David. 2001. "Identifying Winners and Losers in Southern Africa from Global Trade Policy Reform: Integrating Findings from GTAP and Poverty Case Studies for a Zambian Example." Paper prepared for the ESRC Development Economics/International Economics Conference, Nottingham University, April 5-7, 2001. Fofack, Hippolyte, Celestin Monga, and Hasan Tuluy. 2001. "Household Welfare and Poverty Dynamics in Burkina Faso: Empirical Evidence from Household Surveys." World Bank Policy Research Paper 2590. Friedman, Jed. 2001. "Differential Impacts of Trade Liberalization on Indonesia's Poor and Non-Poor." Paper prepared for the Conference on Poverty and the International Economy, organized by World Bank and Swedish Parliamentary Commission on Global Development, Stockholm, October 20-21, 2000. Gortz, Mette, Glenn W. Harrison, Claus Kastberg Nielsen, and Thomas F. Rutherford. 2000. "Welfare Gains of Extended Opening Hours in Denmark." Economics Working Paper B-00-03, The Darla Moore School of Business, University of South Carolina. Harris, Rebecca Lee. 2001. "A Computable General Equilibrium Analysis of Mexico's Agricultural Policy Reforms." IFPRI TMD Discussion Paper 65. Harrison, Glenn W., Thomas F. Rutherford, and David G. Tarr. 2000. "Trade Liberalization, Poverty and Efficient Equity." Economics Working Paper B-81-02, Moore School of Business, University of South Carolina. 25 Hertel, Thomas W., ed. 1997. Global Trade Analysis: Modeling and Applications. New York: Cambridge University Press. Hertel, Thomas W., Paul V. Preckel, and Jeffrey J. Reimer. 2001. "Trade Policy, Food Price Variability, and the Vulnerability of Low-Income Households." Paper prepared for Meeting of the American Agricultural Economics Association, Chicago, Illinois, August 6-8, 2001. Ianchovichina, Elena, Alessandro Nicita, and Isidro Soloaga. 2001. "Trade Reform and Household Welfare: The Case of Mexico," Policy Research Working Paper No. 2667, The World Bank Development Research Trade Group, August. Jones, Ronald W. 1965. "The Structure of Simple General Equilibrium Models," Journal of Political Economy 73, (December), 557-572. Levin, Jorgen. 2000. "Kenya - Poverty Eradication Through Transfers." Paper prepared for the Conference on Poverty and the International Economy, organized by World Bank and Swedish Parliamentary Commission on Global Development, Stockholm, October 20-21, 2000. Levinsohn, James, Steven Berry, and Jed Friedman. 1999. "Impacts of the Indonesian Economic Crisis: Price Changes and the Poor." Paper prepared for the Conference on Poverty and the International Economy, organized by World Bank and Swedish Parliamentary Commission on Global Development, Stockholm, October 20-21, 2000. Forthcoming in Managing Currency Crises in Emerging Markets, Michael Dooley and Jeffrey Frankel, ed., The University of Chicago Press. Lofgren, Hans. 1999. "Trade Reform and the Poor in Morocco: A Rural-Urban General Equilibrium Analysis of Reduced Protection." IFPRI TMD Discussion Paper 38; L6fgren, Hans with Osten Chulu, Osky Sichinga, Franklin Simtowe, Hardwick Tchale, Ralph Tseka, and Peter Wobst. 2001. "External Shocks and Domestic Poverty Alleviation: Simulations with a CGE Model of Malawi." IFPRI TMD Discussion Paper 71. 26 Lysy, Frank, and Lance Taylor. 1980. "The General Equilibrium Income Distribution Model", in Lance Taylor, Edmar Bacha, Eliana Cardoso and Frank Lysy, eds., Models of Growth and Distribution for Brazil. Oxford: Oxford University Press, Ch. 6. McCulloch, Neil, and Michele Calandrino. 2001. "Poverty Dynamics in Rural Sichuan between 1991 and 1995." Paper prepared for the ESRC Development Economics/Intemational Economics Conference, Nottingham University, April 5-7, 2001. Minot, Nicholas, and Francesco Goletti. 2000. "Rice Market Liberalization and Poverty in Viet Nam." IFPRI Research Report No. 114. Ravallion, Martin. 1990. "Rural Welfare Effects of Food Price Changes Under Induced Wage Responses: Theory and Evidence for Bangladesh." Oxford Economic Papers 42: 574-585. Ravallion, Martin, and Dominique van de Walle. 1991. "The Impact on Poverty of Food Pricing Reforms: A Welfare Analysis for Indonesia." Journal of Policy Modeling 13(2): 281-299. Rimmer, M.T. and A.A. Powell. 1996. "An Implicitly Additive Demand System." Applied Economics 28:1613-1622. Robilliard, Anne-Sophie, Francois Bourguignon, and Sherman Robinson. 2001. "Crisis and Income Distribution: A Micro-Macro Model for Indonesia." Paper prepared for the ESRC Development Economics/Intemational Economics Conference, Nottingham University, April 5-7, 2001. Rodrik, Dani. 2000. Comments on "Trade, Growth, and Poverty," by David Dollar and Aart Kraay. Available at: http://ksghome.harvard.edu/-.drodrik.academic.ksgInapers.html. Sadoulet, Elisabeth and Alain de Janvry. 1992. "Agricultural Trade Liberalization and Low Income Countries: A General Equilibrium-Multimarket Approach." American Journal of Agricultural Economics 74 (2): 268-80. Slaughter, Matthew J. 1999. "Globalization and Wages: A Tale of Two Perspectives," World Economy, 22, 5 (July): 609-630. 27 Taylor, Lance, ed. 1990. Socially Relevant Policy Analysis: Structuralist Computable General Equilibrium Models for the Developing World. Cambridge: The MIT Press. Warr, Peter G. 2001. "Welfare and Distributional Effects of an Export Tax: Thailand's Rice Premium." American Journal ofAgricultural Economics. 83(4) (November): 903-920. Winters, L. Alan. 2000. "Trade, Trade Policy, and Poverty: What are the Links?" Centre for Economic Policy Research Paper No. 2382. Wood, Adrian. 1995. "How Trade Hurt Unskilled Workers," The Journal of Economic Perspectives, Vol. 9, No. 3. (Summer): 57-80. 28 Table 1. Cross-Country Regression Authors Title of study Type and source of earnings data Notes on approach Principal findings Dollar and Trade, Growth, Income distribution summary statistics from Determine the statistical relationship Trade volumes are not correlated with Kraay and Poverty UN-WIDER (2000), Deininger and Squire between trade volume and inequality inequality measures, but trade is positively (2001) (1996), Ravallion and Chen (2000), measures such as Gini coefficient correlated with economic growth. Since Lundberg and Squire (2000). Per capita growth reduces poverty, trade also must GDP from Summers and Heston Penn reduce poverty. World Tables, and World Bank data. Dollar and Growth is Good (Same as above) Attempt to explain deviations around the There is no systematic relationship between Kraay for the Poor one-to-one relationship between changes in average incomes and the share of income (2001) average income and changes in poorest one- accruing to the poorest fifth of the income fifth's income using regressions that include distribution. Thus, economic growth tends variables such as trade volume, education, to reduce poverty. and rule of law. Table 2. Partial-Equilibrium/Cost-of-Living Analysis Authors Title of study Type and source of earnings data Notes on approach Principal findings Deaton Rice Prices and Expenditure data from the 1981-2 Thailand Combines expenditure data and Higher rice prices are favorable to rural (1989) Income Socioeconomic Survey of the Whole hypothesized price changes to study the Thai households at all income levels. The Distribution in Kingdom serve as an indicator of income. distributional effect of higher rice prices principal beneficiaries are middle-income Thailand: A Non- that would result from export liberalization rural households. parametric in Thailand. Analysis Ravallion Rural Welfare Aggregate income statistics by source Econometrically estimate welfare effects of An increase in the price of rice will likely (1990) Effects of Food (wages, business, non-market) for 4 rice price increases using time series data have an adverse effect on rural households Price Changes different income groups are from 1984 for Bangladesh; account for effect of higher for 3 to 4 years. In the long run, the rice Under Induced Bangladesh Bureau of Statistics Statistical rice prices on wage rates. price increase is more likely to benefit the Wage Responses: Yearbook of Bangladesh. poorest households than those who are less Theory and poor. Evidence for Bangladesh Ravallion The Impact on Money income is estimated using household Estimate welfare distributions associated With hypothesized full producer income and van de Poverty of Food consumption expenditures in 1981 Susenas with demand response to price and income effects, uncompensated trade liberalization Walle (1991) Pricing Reforms: data set. variability, using 3 stylized examples of has adverse effect on poverty. Poverty A Welfare how incomes could be affected by rice price orderings depend critically on the definition Analysis for changes. and measurement of poverty. Indonesia Levinsohn, Impacts of the No earnings data used, except for summary Calculate change in cost-of-living by Disregarding self-produced agriculture and Berry, and Indonesian statistics from Badan Pusat Statistik on how income decile using actual consumer good owned housing, mean increase in cost-of- Friedman Economic Crisis: labor wages changed during the time of the price changes. living is 130%, with the rural poor suffering (1999) Price Changes and Indonesian economic crisis. most. Accounting for these effects, it is the the Poor urban poor who were most adversely affected. 29 Authors Title of study Type and source of earnings data Notes on approach Principal findings Case (2000) Implications of Expenditure data from the 1993 South Calculate compensating variation for Preliminary results suggest that tariff reform Trade Policy African Living Standards Survey, with Africans and Whites using household results in gains to both groups. A 75% tariff Reform Given information on 8848 households in 360 budget shares, LES estimates, and outside reduction lowers cost of reaching initial Income clusters. Data set includes unspecified estimates of consumer good price changes utility by roughly 2% for Africans, and by Distribution and income and employment information that due to trade reform. 1% for Whites. Expenditure was not used in the study, except for income Pattems in South distributions of different ethnic groups. Africa Deaton and Prices and Expenditure data from 1987-88 and 1993-94 Calculate price indices for 1987-88 and Results generally show agreement with Tarozzi Poverty in India Indian National Sample Survey - no 1993-94 across Indian states, and compare official price indices with respect to rate of (2000) earnings data discussed or used in the with official poverty statistics. increase in price index over time. However, analysis. problems are found with current procedures for calculating official poverty lines. Minot and Rice Market Information on income and employment A multimarket spatial equilibrium model is Export liberalization raises rice prices Goletti Liberalization and from 1992-93 Viet Nam Living Standards simulated for different policy scenarios; within the country, giving a positive effect (2000) Poverty in Viet Survey of World Bank, but factor earnings abstract away from factor market effects of on rural income and a mixed but slightly Nam not explicitly modeled. higher rice prices. favorable impact on poverty. Dercon The Impact of 1989 and 1994-95 panel data set on 362 Changes in welfare of households between In general growth is pro-poor, but results (2001) Economic rural households in 6 communities (not the two time periods of the survey are are mixed. One group with good land, Reforms on necessarily representative of Ethiopia). explained by regression on household labor, and location improved substantively Households in Household income disaggregated into characteristics. over the period; another less well-endowed Rural Ethiopia, earnings from land, wages, livestock, group experienced little change. 1989-1995 business income, and transfers. McCuloch Poverty Dynamics 1991-1995 panel data set of 3311 Provides measures of poverty across 5 years Households are highly vulnerable to falling and in Rural Sichuan households in rural Sichuan collected by the based on household surveys. Measures into poverty even when their average Calandrino between 1991 and Chinese National Bureau of Statistics. vulnerability to moving into and out of consumption is over 20% above the poverty (2001) 1995 Household income disaggregated into poverty. line. earnings from wages, business profits, transfers, and assets income. Appleton Poverty Reduction Data set consists of 6 nationally Household surveys are used to estimate The results suggest that poverty fell for all (2001) During Growth: representative surveys from Uganda from changes in average living standards, poverty decile groups and socio-economic the Case of 1992-2000. Earnings information not and inequality during 1992-2000, in which categories, regardless of the poverty line Uganda, 1992- discussed or used. there was rapid economic growth. used. 2000 Fofack, Household Use 1994 and 1998 household surveys, Calculate Foster-Greer-Thorbecke poverty Economic growth resulting Burkina Faso's Monga, and Welfare and (Enquete Prioritaire I and 11) which appear measure and Gini inequality measure using exchange rate devaluation was undermined Tuluy (2001) Poverty Dynamics to include limited information on household expenditure data for 1994 and by increasing inequality. Suggested causes in Burkina Faso: employment and income of survey 1998. are disparities in human capital, wages, and Empirical participants. access to productive factors. Evidence from Household Surveys 30 Table 3. General-Equilibrium Simulation'3 Authors Title of study Type and source of earnings data Notes on approach Principal findings Adelman and Income 15 household types based on 15 First general equilibrium model to be Most policies tend to have little effect on Robinson Distribution occupational groups, including skilled and applied to a developing country. While size distribution of income (as given by an (1978) Policy: A CGE unskilled workers in different industries, neoclassical, it contains a number of non- estimated 2-parameter lognormal Model of South agricultural workers, government workers, neoclassical features. Models the functional distribution). However, when ownership of Korea capitalists. Disaggregation based on income distribution assuming a simple human capital and land is equitable, a labor reconciling information from the Korean lognormal income distribution for each and skill-intensive export-oriented strategy Special Integrated Household Expenditure household type. can improve the distribution of income. Employment Survey, Urban household survey, Farm household survey, and others. Lysy and The General 12 different types of income recipients In the structuralist tradition. Society is made Reaches opposite conclusion of Adelman Taylor Equilibrium spread across 25 sectors, for a total of 130 up of conflicting groups that differ by and Robinson - public policy contributed to (1980) Income classes in total. Income arises from one of 3 employment. Model includes a forced the deterioration in the distribution of Distribution possible sources: labor, capital, or a saving mechanism and limits the amount of income over the period studied. A later Model proprietorship. Based largely on a 1959 1-0 output adjustment. study by Adelman and Robinson (1988) matrix by Willy van Rijckeghem (1969) and attempts to reconcile the different findings, on Brazilian income distribution data from and suggests they are due to differing Albert Fishlow (1972, 1973). definitions of the income distribution. Bourguignon, Adjustment and None (archetype economy). Links the micro elements (relative price Stabilization packages which do not have Branson, and Income shifts) by which structural adjustment specific components targeted towards the de Melo Distribution: A policies affect income distribution with the poor will have a noticeable adverse effect on (1989) Counterfactual macro elements (asset prices) of adjustment the distribution of income, which is likely to Analysis packages that affect income distribution result in some form of permanent damage through level of economic activity. for those below the poverty line. Sadoulet and Agricultural None (archetype economy). Construct integrated multimarket general The expected 20% increase in world cereal de Janvry Trade equilibrium model for 3 archetypical groups prices following OECD trade liberalization (1992) Liberalization and of developing countries, distinguished on results in rising food import bills and Low Income the basis of whether domestic food crop exchange rate depreciation in African Countries: A GE- production is competitive with imported countries with noncompetitive cereal Multimarket cereals. Incorporates 2 types of labor and 5 imports; the opposite happens in other Approach types of households. African and Asian countries. Coxhead and Does Technical 7 household types, distinguished based on General equilibrium model of Philippines Factor market adjustments are substantially Warr (1995) Progress in ownership of mobile labor and capital, and 4 with 3 commodities and 4 sectors is used to more important than spending effects when Agriculture sector-specific factors. Based on a 1985 measure and decompose the effects of technological progress occurs. When Alleviate Philippines National Statistics Office technical change on poverty. General poverty sensitive welfare weights are used, Poverty? A Family Income and Expenditure Survey. equilibrium closures are compared to partial partial equilibrium analysis predicts smaller Philippine Case equilibrium closures. welfare gain than general equilibrium Study analysis. 13 With regard to the column on earnings data (column 3), the general equilibrium analyses generally employed multiple stylized households unless otherwise indicated. Additionally, the method used to link a representative household's income to productive factors was almost never described in the studies. It is generally implied that factor earnings links are based on an underlying Social Accounting Matrix. 31 Authors Title of study Type and source of earnings data Notes on approach Principal findings Bautista and Income Effects of 3 rural (large-farm, small-farm & other) and General equilibrium model is based on 3 Tariff liberalization is superior to policy Thomas Alternative Trade 2 urban (metro Manilla & other) household types of trade policy adjustments dealing options that are more restrictive to trade, (1997) Policy types. Based on a 1979 Philippines Social with an unsustainable current account deficit since there are larger benefits to small-farm Adjustments on Accounting Matrix with 4 primary factors are simulated. and other rural households relative to the Philippine Rural (unskilled labor, skilled labor, land, capital). more affluent Metro Manilla, other urban, Households: A and large-farm households. GE Analysis L6fgren Trade Reform and 4 household types, distinguished according A special feature of the Social Accounting Trade liberalization in agriculture will result (1999) the Poor in to whether they poor/non-poor, and Matrix is the attempt to capture pronounced in gains for the country as a whole, while Morocco: A rural/urban orientation. Based on a 1994 rural-urban distinctions in skills, wages, and the rural poor loses out. Compensation for Rural-Urban GE Social Accounting Matrix, for which the sectors that are prevalent in Morocco. rural areas is likely needed for liberalization Analysis of population, consumption, and labor force to be pursued. Reduced information is derived from various Protection Royaume du Maroc statistical volumes, as well as World Bank and IMF statistics. Decaluwe, Poverty Analysis None (archetype economy). First part focuses on how Social Accounting The analysis is for an archetype economy, Patry, within a General Matrixes can be used to analyze income so no numerical results are available. The Savard, and Equilibrium distribution. In second part, a general highlights are the relatively general Thorbecke Framework equilibrium framework is developed in functional form for the within-group income (1999) which the poverty line is endogenous, and a distribution, and methodology for flexible Beta functional form is used to incorporating an endogenous poverty line. model intragroup income distribution. Cogneau and Growth, Earnings data are split out across 6 income This is a novel approach, since endogenous While relative income and price changes are Robilliard Distribution and sources for 4,508 households are from the variables are determined by 4,508 actual generally significant, the impact of the (2000) Poverty in 1993 EPM survey carried out by INSTAT households, and incorporated into an overall various shocks on the aggregate indicators Madagascar: on behalf of the Malagasy government. general equilibrium framework. Results are of poverty and inequality tend to be small. Learning from a This is linked to macroeconomic data from a presented in terms of 14 household types. A Microsimulation 1995 Madagascar Social Accounting Matrix key benefit is that within-group income Model in a constructed by Razafindrakoto and distribution variance is endogenized. General Roubaud. However, only 3 sectors are modeled, and Equilibrium there is limited macroeconomic structure. Framework Harrison, Trade 19 rural households and 21 urban Addresses the possibility of combining Trade liberalization leads to some groups of Rutherford, Liberalization, households, distinguished by income level. policies so that no poor household is harmed the rural and urban poor being worse off in and Tarr Poverty and 8 types of labor and 2 types of capital. from trade policy reform. the absence of compensation. Direct (2000) Efficient Equity Based on a 1990 Social Accounting Matrix compensation mechanisms appear to work constructed by De Santis and Ozhan. The well in the case of trade reform in Turkey, authors note that 'it does not appear that even when accounting for costs of raising production labor is mapped into the sectors revenue. that use it using defensible economic criteria'. Gortz, Welfare Gains of None, although use is made of a survey Has nothing to do with trade and poverty, There are substantial consumer welfare Harrison, Extended quantifying how 613 Danish households but illustrates how 613 households can be gains from liberalizing the regulations on Nielsen, and Opening Hours in allocate their time. directly incorporated in a general shopping hours in Denmark. Poorer Rutherford Denmark equilibrium model. households tend to benefit more than (2000) wealthier households. 32 Authors Title of study Type and source of earnings data Notes on approach Principal findings Devarajan Trade Reform in 24 households comprised of 6 income Four model specifications concerning the Removal of tariffs will likely reduce the and South Africa: categories and 4 ethnic groups. Within each revenue replacement issue that arises with average welfare of white households but Mensbrugghe Impacts on ethnic group, labor is fuither divided into 13 complete removal of tariffs. improve the average welfare of black (2000) Households categories, making for 52 types of workers. households. Within ethnic groups, richer Based on a 1992 South African Social black and poorer white households are Accounting Matrix. likely to benefit. Levin (2000) Kenya - Poverty 6 rural households are distinguished by size Use general equilibrium model with 9 A transfer to a specific group can alleviate Eradication and type of landholding. 4 urban households sectors to calculate transfers needed to yield poverty in that group, but depending on how Through are distinguished by education level. Based zero headcounts in the Foster-Greer- this is financed it may affect non-targeted Transfers on a 1986 Social Accounting Matrix of Thorbecke poverty index. groups negatively. Kenyan Ministry of Planning and National Development, which was constructed using Kenya's Urban Household Budget Survey, Rural Household Budget Survey and Labour Market Survey. Lee Harris A Computable 15 households are differentiated according In contrast to most other studies in this A system of lump sum payments to farmers (2001) General to region and income level. 4 types of non- section, this general equilibrium model is preferable to the old system of subsidies Equilibrium ag labor (professional, white-collar, blue- distinguishes 4 regions within the economy. and price supports. However, in the event of Analysis of collar, and unskilled) and 4 types of ag. an exchange rate devaluation, the old system Mexico's labor categories. Based on a 1996 Mexican performs better in terms of rural incomes Agricultural Social Accounting Matrix - household and output. Policy Reforms income and expenditure data come from the 1994 Encuesta Nacional de Ingresos y Gastos de Hogares, INEGI. Lofgren, External Shocks 8 households distinguished by land holdings Simulations designed assess impact of tax- Real depreciation has a pro-rural bias and is Chulu, and Domestic (some have no ag. income), education, based land reform and its sensitivity to effective at eliminating balance-of-payment Cichinga, Poverty urban/rural orientation. Income is from ag. various assumptions. difficulties. An expanded public works Simtowe, Alleviation: labor, land, non-ag. labor, ag. capital, and/or program generates significant gains for the Tchale, Simulations with govt. transfers. Based on 1998 Malawi rural poor, but negatively affects non- Tseka, Wobst a CGE Model of Social Accounting Matrix from Chulu and agricultural households. (2001) Malawi Wobst (2000), which is based partly on the 1997-98 Malawian Integrated Household Survey published by the Malawi National Statistical Office. Evans (2001) Identifying Global Trade Analysis Project (GTAP) Examine the effect of trade liberalization on Regionally based trade policy reforms Winners and database distinguishes I household in poverty in Zambia using both a case study generally have a neutral or adverse impact Losers in Zambia, that receives income from 5 approach and the GTAP general equilibrium on poverty. Favorable distribution impacts Southern Africa factors. Results are also presented in terms model. The study's hypothesis is that these on poverty from trade policy reform are from Global of 4 stylized households distinguished by two approaches are highly complementary. anticipated in a global WTO liberalization Trade Policy location and income. Five alternative policy scenarios are scenario, and in a EU/SADC7 scenario. Reform considered. 33 Authors Ti¢Ke of study l a -z -ce source of emrmiungs dna Notees on appironclP IrniPHnin f&indlngs Cockburn Trade 1995 Nepalese Living Standards Survey Similar in spirit to the framework of Results suggest that impact of trade (2001) Liberalization and (NLSS) gives income sources for 3,373 Cogneau and Robilliard, a general liberalization is complex and varies Poverty in Nepal: households, although income data not equilibrium model is constructed that substantially by household, but that effect A Computable clearly distinguished between labor (skilled explicitly models 3,373 households. This on overall income distribution is small. The General and unskilled) and capital (land, ag. capital enables determination of how trade absolute impact of liberalization - whether Equilibrium and non-ag. capital) as used in the 1986 liberalization impacts individual households positive or negative - generally increases Micro-Simulation Social Accounting Matrix. Income shares and how these results feed back into the with the level of income. Approach of these factors from the Social Accounting general equilibrium of the economy. Matrix were applied to the NLSS data in order to separate out these sources. Warr (2001) Welfare and 10 household types distinguished by income Examine the distributional impacts of a rice The negative effect of lower unskilled Distributional quintile, and rural versus urban orientation. export tax for Thailand, which has a small wages outweighs the benefit of lower rice Effects of an Earnings are based on holdings of skilled degree of market power in world market. A prices, causing both the rural and urban poor Export Tax: and unskilled labor (distinguished by method is demonstrated whereby the to be harmed by an export tax. Thailand's Rice education), land, agric. capital, and non- optimal tax can be determined. Premium agric. capital (mobile and fixed). Based on a 1988 Thai NSO Socio-economic survey. 34 Table 4. Micro-Macro Synthesis Authors Title of study Type and source of earnings data Notes on approach Principal findings Friedman Differential Income sources for each of 62,010 households in Simulate GTAP model under Indonesian Under both trade liberalization (2000) Impacts of Trade 1996 Susenas survey include wages, ag. and non-ag unilateral trade liberalization, and then scenarios, few or no households are Liberalization on business income, house rent, pensions, dividends, global trade liberalization. Calculate made worse off, but distribution of Indonesia's Poor transfers, and asset sales. Global Trade Analysis household-specific compensating gains from liberalization tends to be and Non-poor Project (GTAP) database gives income for a variation associated with commodity skewed towards the urban rather than representative Indonesian household based on 5 price changes. Adjust this for income the rural, and the wealthy rather than factors. Paper describes assumptions made to changes (as given by GTAP model's the poor. reconcile GTAP factor income with Susenas factor price changes) to get net effect per income sources. household. lancho- Trade Reform and Information on 47 income sources for each of Use price changes from GTAP model Tariff reform will have a positive effect vichina, Household 14,042 households provided by 1996 Mexican simulated under the NAFTA agreement on welfare for all income deciles. An Nicita, and Welfare: The National Household Income and Expenditure in conjunction with household survey assumption of homothetic preferences Soloaga Case of Mexico Survey. These 47 income sources are mapped to 5 information for Mexico to compute results in larger welfare gains, while a (2001) Global Trade Analysis Project (GTAP) factors changes in income distribution and non-homothetic assumption results in using a number of arbitrary assumptions. poverty. Avoids representative household the poor gaining the most. assumption Robilliard, Crisis and Income Information on earnings (various wage work and Simulate Indonesian financial crisis with Poverty increases over the 1997-98 Bourguignon Distribution: A self-employment) for sub-sample of 9,800 general equilibrium model to get period were due in equal measure to the and Micro-Macro households provided by 1996 Susenas survey. aggregate price changes, and then solve El Nino drought and to the financial Robinson Model for General equilibrium model is based on 1995 Social an estimated microsimulation model of crisis. Use of some available policy (2001) Indonesia Accounting Matrix that has 8 labor categories, 6 household income such that it generates options would have resulted in a types of capital, and land. Mapping between factors aggregate changes consistent with the smaller increase in poverty. in general equilibrium model and the household general equilibrium model - avoids data is not described. representative household assumption. Hertel, Poverty Impacts Use household surveys with earnings information Simulate global trade liberalization with Trade liberalization generally results in Preckel, of Multilateral for Brazil, Chile, Indonesia, Philippines, Thailand, Global Trade Analysis Project (GTAP) reduced poverty in each of the 7 focus Cranfield, Trade Uganda, and Zambia (available from World Bank). general equilibrium model to get factor countries. Indonesia experiences the and Ivanic Liberalization Across the surveys, categorize households into 5 and commodity price changes by region. largest national reduction in poverty. (2001) strata, including those getting at least 95% of These are then used in conjunction with The experience of particular groups income from (i) transfers, (ii) agriculture, (iii) non- estimated expenditure functions and within the countries is mixed. ag. business, (iv) wages, and then (v) a stratum for factor eamings information across 5 households that have diversified income sources. strata (and income levels within a strata) to assess the welfare of marginal households in each of 7 countries. 35 Policy Research Working Paper Series Contact Title Author Date for paper WPS2765 Inequality Aversion, Health Adam Wagstaff January 2002 H. Sladovich Inequalities, and Health Achievement 37698 WPS2766 Autonomy, Participation, and Learning Gunnar S. Eskeland January 2002 H. Sladovich in Argentine Schools: Findings and Deon Filmer 37698 Their Implications for Decentralization WPS2767 Child Labor: The Role of Income Rajeev H. Dehejia January 2002 A. Bonfield Variability and Access to Credit in a Roberta Gatti 31248 Cross-Section of Countries WPS2768 Trade, Foreign Exchange, and Energy Jesper Jensen January 2002 P. Flewitt Policies in the Islamic Republic of David Tarr 32724 Iran: Reform Agenda, Economic Implications, and Impact on the Poor WPS2769 Immunization in Developing Countries: Varun Gauri January 2002 H. Sladovich Its Political and Organizational Peyvand Khaleghian 37698 Determinants WPS2770 Downsizing and Productivity Gains Martin Rama January 2002 H. Sladovich In the Public and Private Sectors Constance Newman 37698 of Colombia WPS2771 Exchange Rate Appreciations, Labor Norbert M. Fiess February 2002 R. lzquierdo Market Rigidities, and Informality Marco Fugazza 84161 William Maloney WPS2772 Governance Matters II: Updated Daniel Kaufmann February 2002 E. Farnand Indicators for 2000-01 Aart Kraay 39291 Pablo Zoido-Lobat6n WPS2773 Household Enterprises in Vietnam: Wim P. M. Vijverberg February 2002 E. Khine Survival, Growth, and Living Jonathan Haughton 37471 Standards WPS2774 Child Labor in Transition in Vietnam Eric Edmonds February 2002 R. Bonfield Carrie Turk 31248 WPS2775 Patterns of Health Care Utilization in Pravin K. Trivedi February 2002 R. Bonfield Vietnam: Analysis of 1997-98 31248 Vietnam Living Standards Survey Data WPS2776 Child Nutrition, Economic Growth, Paul Glewwe February 2002 E. Khine and the Provision of Health Care Stefanie Koch 37471 Services in Vietnam in the 1990s Bui Linh Nguyen WPS2777 Teachers' Incentives and Professional Gladys L6pez-Acevedo February 2002 M. Geller Development in Schools in Mexico 85155 Policy Research Working Paper Series Contact Title Author Date for paper WPS2778 Technology and Firm Performance Gladys L6pez-Acevedo February 2002 M. Geller in Mexico 85155 WPS2779 Technology and Skill Demand Gladys L6pez-Acevedo February 2002 M. Geller in Mexico 85155 WPS2780 Determinants of Technology Adoption Gladys L6pez-Acevedo February 2002 M. Geller in Mexico 85155 WPS2781 Maritime Transport Costs and Port Ximena Clark February 2002 E. Khine Efficiency David Dollar 37471 Alejandro Micco WPS2782 Global Capital Flows and Financing Ann E. Harrison February 2002 K. Labrie Constraints Inessa Love 31001 Margaret S. McMillan WPS2783 Ownership, Competition, and Geroge R. G. Clarke February 2002 P. Sintim-Aboagye Corruption: Bribe Takers versus Lixin Colin Xu 37644 Bribe Payers WPS2784 Financial and Legal Constraints to Thorsten Beck February 2002 A. Yaptenco Firm Growth: Does Size Matter? Asli Demirgu,-Kunt 38526 Vojislav Maksimovic WPS2785 Improving Air Quality in Metropolitan The Mexico Air Quality February 2002 G. Lage Mexico City: An Economic Valuation Management Team 31099 WPS2786 The Composition of Foreign Direct Beata K. Smarzynska February 2002 P. Flewitt Investment and Protection of 32724 Intellectual Property Rights: Evidence from Transition Economies WPS2787 Do Farmers Choose to Be Inefficient? Donald F. Larson February 2002 P. Kokila Evidence from Bohol, Philippines Frank Plessmann 33716 WPS2788 Macroeconomic Adjustment and the Pierre-Richard Agenor February 2002 M. Gosiengfiao Poor: Analytical Issues and Cross- 33363 Country Evidence WPS2789 "Learning by Dining" Informal Somik V. Lall February 2002 Y. D'Souza Networks and Productivity in Sudeshna Ghosh 31449 Mexican Industry