Chapter 5. Poverty and Distribution: Thirty Years Ago and Now
- Benedict Clements, Ruud Mooij, Sanjeev Gupta, and Michael Keen
- Published Date:
- September 2015
- Ravi Kanbur
In 1987 the IMF Staff Papers published “The Measurement and Alleviation of Poverty: With an Application to the Impact of Macroeconomic Adjustment” (Kanbur 1987). This paper, which was written in 1985 during a stay at the IMF as a Visiting Scholar, spoke to the key analytical and policy issues of the day as I saw them. The present chapter highlights what has happened in the past three decades in the poverty and distribution discourse from an analytical and policy-oriented perspective. There is no attempt to be comprehensive. Rather, this somewhat idiosyncratic take on the literature of the past 30 years is offered as a way of sparking a discussion on where we stand and where we need to go.
The mid-1980s were the culmination of 15 years of intensive discussion on the analytical and philosophical aspects of poverty and inequality conceptualization and measurement. Publications by Tony Atkinson (1970) and Amartya Sen (1973, 1976) ignited a technical literature on inequality and poverty indices that overlapped with and complemented economists’ engagement with an emerging philosophical literature on inequality and distribution. Kenneth Arrow (1973) formulated Rawls (1971) in terms that economists could relate to, and Nozick’s (1974) counter also made it to economists’ reading lists. In poverty measurement, the paper by Foster, Greer, and Thorbecke (1984) introduced a poverty index (the Foster-Greer-Thorbecke [FGT] measure) that has become the workhorse of empirical work because of the intuitive way in which it captures differing degrees of poverty aversion, and because it is additively decomposable—thereby allowing an easy and intuitive accounting disaggregation of national poverty into policy-salient components. Kanbur (1987) used these decomposability properties to analyze the possible effects of expenditure-switching policies on poverty, while Besley and Kanbur (1988) analyzed the targeting of food subsidies to minimize the FGT poverty index, drawing on a literature that went back to Akerlof (1978). More generally, the public economics literature on taxation and expenditure advanced rapidly on the theoretical front, with a whole range of results being derived on distributionally sensitive fiscal policies (for example, Atkinson and Stiglitz 1980; Newbery and Stern 1987).
This chapter takes up the story in the mid-1980s and brings it up to date. It considers developments in three categories: facts and empirics, concepts and theory, and policies and interventions. While issues that came to the fore particularly in the past 30 years will be addressed, there were indeed antecedents in an earlier period. Moreover, a clean categorization into empirics, theory, and policy is also not possible, and some issues cover all three to varying degrees. Finally, this treatment is clearly idiosyncratic rather than comprehensive; others will surely have different takes on the development of the literature.
Facts and Empirics
Just More Facts
A key difference between now and 30 years ago is quite simply the availability of more distributional data sets—for more countries, more time periods, and more dimensions.1 The World Bank’s Living Standards Measurement Survey (LSMS) website lists more than 100 surveys for more than 35 countries.2 All of these surveys were developed after 1985. There are a host of surveys for the transition economies of Europe and central Asia dating from the 1990s, as these countries joined the market economic system. In Latin America and the Caribbean, where wage and income surveys dominated in the past and still do in the present, expenditure surveys are increasingly being conducted and becoming the norm.
Another major source of information that has expanded dramatically in the past 30 years comes from the Demographic and Health Surveys (DHS). The website3 lists more than 300 surveys, most from the 1990s onward. Although the DHS do not collect information on household income and expenditure, they do collect information on assets, which some analysts have used to cross-correlate with health and nutrition outcomes (for example, Filmer and Pritchett 1999). In any event, the availability of health and nutrition data has spurred further work on these dimensions of poverty and human development. In particular, the new data have helped clarify and quantify the prevalence and spread of HIV/AIDS, an issue that was hardly on the policy agenda in developing countries in the mid-1980s.
Perhaps the most remarkable transformation occurred for Africa. When Kanbur 1987 was written, no modern household income expenditure surveys were available for African countries. Now surveys for the following countries are listed on the LSMS website: Côte d’Ivoire, Ghana, Malawi, Niger, Nigeria, South Africa, Tanzania, and Uganda. But these are just the surveys on this website, with microdata made available for public use through the LSMS project. The World Bank’s PovcalNet project makes available data on various distributional aggregates for many more African countries, and for 161 countries globally.4 The DHS website has microdata available for more than 35 sub-Saharan African countries.
Consider now the world’s two most-populous countries, India and China. For India, every five years the National Sample Survey’s “thick round” provides distributional data that are central to the policy and dialogue in that country. The controversy on data for the 1990s, and the role of this controversy in debates on the distributional impact of economic liberalization, shows the ongoing use that is made by researchers and policymakers of the time series data that are available for the country (Deaton and Kozel 2005). The emergence of China from its economic isolation in the 1980s also meant greater availability of distributional (and other) data for China. Data availability increased significantly in the 1990s, feeding into an intensive debate on growing inequality in China (for example, Chen and Ravallion 2007; Zhang and Kanbur 2005).
Greater availability of household surveys has allowed the application of techniques that had been developed before the mid-1980s, but which awaited appropriate micro-level data for implementation. Prominent examples include incidence analysis of public expenditure and distributional analysis of price and subsidy reform. Such application is now routine, including in the IMF (Coady, Dorosh, and Minten 2007). It would not have been possible before the mid-1980s. Time series of distributional data for a large number of countries have permitted intertemporal analysis of the evolution of inequality and poverty, and have fueled a debate on globalization, growth, and distribution. For instance, recent work at the IMF (for example, Ostry, Berg, and Tsangarides 2014; IMF 2014) has made use of the World Income Inequality Databases, which are explained and assessed by Jenkins (2014). Such comprehensive data sets were simply not available three decades ago.
For a smaller but still significant number of countries, household panel data are now available. Baulch (2011) provides an inventory of more than 60 panel data sets covering more than 30 developing and transition countries. Most of these data sets are from the 1990s onward. Along with these new panel data has come an interest in analysis of the policy implications of the findings of these panels, especially for income and poverty dynamics, and the high degree of risk and vulnerability faced by the poor in developing and transition economies.
Trends in Poverty and Inequality
From the vantage point of the debates of the mid-1980s, two sets of facts seemed to be important. First, a group of countries in East Asia had managed to achieve growth with equity, leading to dramatic declines in poverty. This seemed to suggest that increasing per capita income could exist simultaneously with declining or constant inequality, contrary to Kuznets’s (1955) hypothesis that inequality would first increase and then decrease only after some time as per capita incomes increased. Second, in the absence of time series data, the testing of this hypothesis in the 1970s and 1980s relied on cross-country econometric regressions, with all the problems that those entail. Counter to the estimates of Ahluwalia (1976), work in the 1980s by Anand and Kanbur (1993a, 1993b) argued that there was no “Kuznets curve” to be found in the cross-country data. This became the conventional wisdom, and has been found to be largely corroborated by the literature of the 1990s and the 2000s, using more, and more recent, distributional data, as from the compilation by Deininger and Squire (1996, 1998). The result was interpreted by some as suggesting that there was no systematic relationship between growth and inequality change, so policymakers should go all out for growth, since inequality would be expected to remain constant and thus poverty would decline (Dollar and Kraay 2002). An alternative interpretation would simply have been to caution against the use of cross-country regression analysis to draw time series conclusions for any one country.
From the vantage point of today, however, East Asia, and Asia in general, tells a very different story for the past three decades. Growth rates have been at historically unprecedented highs, but inequality has risen sharply. The growth has aided poverty reduction, but at a slower pace than if inequality had not risen. One estimate suggests that at constant inequality Asian growth of the past two decades would have lifted 240 million more people out of poverty (Kanbur and Zhuang 2012). Rising inequality has also occurred in the Organisation for Economic Co-operation and Development economies, the United States in particular (Stiglitz 2013). The phenomenal response to the book on inequality by Piketty (2014) highlights the policy and public concerns to which this persistent inequality has given rise. The global trend appears to be led by fundamental forces of technological change that are favoring capital over labor and skilled labor over unskilled labor, and the forces for rising inequality are being transmitted throughout the world by the globalization of trade, investment, and migration (Kanbur 2014b). One part of the world has indeed reversed this trend and produced falling inequality. The Latin American story highlights the way in which purposive policy intervention can address global forces (Kanbur 2014b; Kanbur and Zhuang 2012). But the global forces have raised distributional concerns among policymakers and civil society, despite the reductions in poverty that have accompanied the high growth rates.
Rising inequality dissipates the impact of growth on poverty. A given growth rate applied to a high level of inequality will lead to a lower level of poverty decrease than the same growth rate applied to a lower level of inequality, even if inequality does not change with growth. In this sense, inequality is bad for poverty reduction (Ravallion 2008). But might rising inequality impede growth itself in future periods? On the face of it, if increased inequality and high growth rates have persisted for more than a decade for a significant number of countries—two decades or more for some countries—then it does seem that rising and high inequality may not be an impediment to growth. However, the time series may not be long enough to draw confident conclusions. The literature has once again had to fall back on cross-country regressions, this time of growth as a dependent variable and inequality as an independent explanatory variable. Despite ample room for agnosticism given the usual pitfalls of cross-country regressions, a consensus does seem to be developing that high inequality may be detrimental to growth.5
Micro and Macro Evaluation
The 1980s saw the peak, and then the decline, of two methods of distributionally oriented evaluation of government interventions. The first, on project evaluation and shadow prices, dated back to the developments in public economics of the late 1960s and the 1970s. But by the mid-1980s intellectual interest had waned, and operational interest never reached critical mass.6 The second, on evaluation of macro policies, and their distributional consequences in particular, using computable general equilibrium models also had its day in the 1980s and the early 1990s and then faded.7
One empirical strategy that was prevalent in the mid-1980s and before can be argued to have been used even more intensively in the subsequent period. This strategy is the use of cross-country regressions to test hypotheses about the impact of interventions and policies on growth and distributional outcomes. Cross-country regularities were brought to the fore in the 1960s and the 1970s in the work of authors such as Adelman and Morris (1974) and Chenery and Syrquin (1975). This chapter has already referred to the Kuznets curve literature of the 1970s and 1980s. But the availability of expanded data sets for more countries, the macroeconomic focus of the “structural adjustment period” of the 1980s and the 1990s, and the revival of growth theory in the 1990s assisted in an explosion of econometric analyses using cross-country data. Indeed, cross-country regressions have in many ways been central to the development debates of the past two decades.
This literature has not been without its controversies. Much attention has focused on the impact of globalization on growth, inequality, and poverty. Many of the technical issues highlighted in the Kuznets literature of the earlier period are equally present in these debates—most prominent among them, the extent to which inference can be drawn about development processes from cross-section relationships across countries with very different structural conditions. The endogeneity of key explanatory variables, such as trade ratios or tariffs to measure openness, has also been much debated.8 The “institutional quality” variables that have increasingly been used in these regressions are open to similar questions. However, despite the debates, the cross-country regressions approach to evaluating development policies and interventions continues to be a significant presence in the current development discourse.
However, one empirical strategy was largely absent in the mid-1980s and before, but rapidly gained prominence in the past decade. This strategy takes the “medical drugs testing” approach seriously and implements it for development interventions. The basic problem with the standard econometric approach is the difficulty of controlling for unobserved heterogeneity, which may have correlation patterns that bias the estimated results. The “randomized evaluation” methodology from pharmaceuticals testing has framed the response of a growing number of development economists, especially in the 2000s. Some have argued that this should be at the heart of what organizations like the World Bank do in the development business (Banerjee and He 2003). The movement has had an impact in the academic literature, but also in the fact that evaluation procedures are much more carefully discussed in project design. This is project evaluation of a type, then, that is very different from the type that rose to prominence in the 1970s and faded in the 1980s—the design of control groups has replaced the estimation of shadow prices. Invariably, of course, there has been a backlash. The very strength of randomized evaluation—namely, the controlling of specific conditions to isolate the impact of the intervention in question—raises questions about the generalizability of the conclusions, particularly for policy purposes. There is also the concern that the “big questions” in development are being sidelined in favor of the small detail of the design of randomization. The debate continues, and is one of the live and vital issues in development economics today.9
Concepts and Theory
Poverty Dynamics and Risk
The greater availability of panel data sets has led to a greater empirical appreciation of the general “churning” that takes place around the poverty line. In rural KwaZuku-Natal in South Africa, 44 percent of the population either moved into or out of poverty during a five-year period. The number was 28 percent in rural Nicaragua during a three-year period, and similar numbers can be found for a range of countries (Chronic Poverty Research Centre 2005, Table 11.1). In countries for which panels have been developed for a period of a decade or more, the volatility of household incomes has been well documented.10
Leaving aside the empirical literature that has developed on the basis of these panel data sets, the issues raised by them have led to a new wave of theorizing on and conceptualization of poverty in a dynamic and risky setting. The standard snapshot view is by now relatively straightforward. There is a threshold value of well-being, and poverty is below this threshold. Various axioms capture intuitions about aggregation into a single poverty index, and from these are derived families of poverty indices, like that put forward by Foster, Greer, and Thorbecke (1984). But now consider dynamics, or risk. Suppose the individual is in poverty in one period, and out of poverty in the next. How is this individual’s poverty to be assessed? One possibility is to compare the individual’s present discounted value of well-being (as measured by consumption, say) to the present discounted value of the poverty line (chronic poverty). But this still leaves open the issue that an individual not in poverty by this measure is still in poverty in one of the two periods (transient poverty [Jalan and Ravallion 2000]). An alternative is to classify as chronically poor those who are poor in both periods.11
All of the above led to a lively literature on dynamics and poverty measurement in the 1990s. But risk and vulnerability must also be added to the measurement mix. What is the risk of income and consumption fluctuations faced by individuals and households, and how does this risk vary at different income levels? Moreover, how is individual risk to be aggregated to provide an overall level of risk for the society as a whole—what weights are to be used? Empirical issues arise in estimating risk from the fairly short runs available for panels so far (two or three observations for most, half a dozen or so for a small number), but the theoretical literature has developed quickly in response to these questions, and a range of dynamic and risk-encompassing measures of poverty are now available (for example, Ligon and Schechter 2003; Calvo and Dercon 2009; Foster 2009).
An extreme case of the variability caused by risk is mortality. What happens to standard measures of poverty when a poor person dies? All else being held constant, the measures decrease! This is surely an unacceptable property of our poverty measures, and raises a fundamental question about their axiomatic structure. All of these measures are derived assuming that the population set is unchanged. But when individuals disappear because of mortality, or new individuals appear because of birth, new rules are needed to make the disappearance or appearance commensurate with the ongoing presence of individuals. Various methods for addressing this issue have been proposed (for example, Kanbur and Mukherjee 2007), but the basic point is that while the “Sen axioms” of the 1970s served us well in helping to derive operational poverty measures, during the past decade these axioms have increasingly been questioned in the conceptual literature.
Gender and Intrahousehold Inequality
In the mid-1980s, the discourse on the systematic incorporation of gender and intrahousehold allocation into the poverty distribution was just beginning. Sen (1983) introduced the issue of gender inequality within the household and in general during the development process. The 1980s also saw the beginning of theoretical and conceptual work on intrahousehold allocation models and their application to the newly available household survey data sets.
A major data constraint in mapping out the extent and nature of intrahousehold inequality in consumption, commensurate with consumption-based measures of overall inequality and poverty, is that individual-level data on consumption are not available. Indeed, it is difficult to see how this could be done comprehensively, since a significant portion of household consumption (like the house itself) is a “public” good within the household. In any event, information, even for those items like food that are individualized, is collected at the household level. The standard practice is then to divide the monetary measure of total household consumption by the number of household members, and allocate to each individual the per capita household consumption. Sometimes (though not very often in official statistics), household size is adjusted for composition and for economies of scale. But throughout, in the absence of individual consumption data, the assumption is that real inequality within the household is not present. In a first attempt to measure the effect of ignoring household inequality, Haddad and Kanbur (1990) use a specially collected data set with individual-level food-consumption information from the Philippines. They find that overall inequality and poverty could be understated by as much as 30 percent.
Apart from its impact on measures of inequality and poverty, neglect of intrahousehold inequality also affects policy interventions that try to target individuals within the household, such as young children or women. A model of intrahousehold allocation as a function, among other things, of overall household resources is needed. It would be fair to say that well into the 1980s the standard “unitary model” ruled the roost. The key theoretical prediction of this model, with implications for empirical testing and for policy, is that of “income pooling”: the household’s consumption pattern, including that of individuals in that household, depends only on the total budget constraint, not on which individual brings in what amount of resources to the budget. But from the 1980s onward, and especially in the 1990s, new theoretical models, and new types of econometric testing, began to question systematically the income-pooling hypothesis. For example, intrahousehold bargaining models were developed to highlight the implication that the resources brought to the household by an individual would partly determine how much benefit that individual could draw from the available household consumption (Manser and Brown 1980; Bourguignon and Chiappori 1992; Ghosh and Kanbur 2008). The econometric theory of testing for “collective” as opposed to unitary models using only household-level consumption data was developed and applied. And the impacts of differential sources of household income on observable individual outcomes such as anthropometrics were also tested and found to be significant (for example, Browning and Chiappori 1998; Quisumbing and Maluccio 2003). By the mid-1990s, a group of economists was already working in this area and had issued a manifesto arguing that the burden of proof would now be on those who would support the unitary model (Alderman and others 1995). Since then the theory and the evidence against the unitary model has continued to grow.
Although developed-country policy issues have also propelled these conceptual changes, and their empirical testing, development-economics issues have played a central role in motivating these analytical advances, which started slowly in the 1980s, accelerated in the 1990s, and have matured in the 2000s. These issues are surveyed and synthesized in the World Bank’s World Development Report 2012: Gender Equality and Development (World Bank 2011).
Multidimensionality and Interdisciplinarity
What exactly is poverty? If it is the lack of an adequate standard of living, what exactly is this standard of living? In the 1980s the discussion, at least in the international agencies, was dominated by economic concepts and measurement—essentially the monetary value of goods and services consumed. Sure, there were concerns and methods for addressing at least some items of consumption for which there were no market prices (like production of food for home consumption), but the core approach and method were very clearly tied to “money metric utility.”
In the 1990s this sole focus on the monetary value of consumption (or income) shifted to a systematic concern with a broader range of items, even though income and consumption retained a central role. Starting in 1990, the United Nations Development Programme’s Human Development Report introduced the Human Development Index, which combined income with education and health to produce an overall index for ranking countries. Despite many technical criticisms of the index,12 its introduction changed the terms of the debate. It brought education and health to the forefront as key independent components of assessment, not just as inputs to income enhancement, thereby cementing these dimensions of the “basic needs” approach of the 1970s. Part of the rationale for such broadening was to argue that education and health were components of the standard of living that could not be proxied by income (conceptually or empirically). Another part was that an explicit discussion of education and health also broadened the policy focus to incorporate direct interventions in these sectors rather than relying solely or primarily on instruments seen to increase incomes more directly. The development of the AIDS pandemic, not really foreseen clearly in the 1980s, also served to bring health goals to the fore. The multidimensionality wrought by the Human Development Index is seen in fuller form in the eight Millennium Development Goals of the UN system, which were signed on to by world leaders in 2000 for the period 2000–15, and their extension in discussions on the “post-2015” agenda (United Nations 2014).13 At the time of writing, the world is looking to the post-2015 agenda under the heading of Sustainable Development Goals. Whatever the outcome of this complex process, it looks as though multidimensionality will play a larger role.
One dimension of the standard of living that has not received as much support and consensus as education and health is “voice”—the extent to which poor people participate in and influence the decisions that affect them. There are, of course, difficulties in the conceptualization and measurement of voice, or of “empowerment,” another term often used to capture this constellation of concerns. And the causal, instrumental role of “democracy” and “governance” variables in growth is perhaps even less clear than the roles of education and health. However, the importance of these factors in an evaluation of development outcomes is at least implicitly recognized in the conditionality of many bilateral donors, who put a high weight on the poor having voice and accountability. As the discussion matures, and concepts and measurement are sharpened, no doubt this cluster of issues will take its rightful place in the broadening of the standard of living from narrow, income-focused approaches.14
Development has, of course, seen interest from many of the social science disciplines. Historically, these disciplines have largely gone their separate ways, with their own literatures, journals, and conferences. Before the mid-1980s, economics and philosophy interacted somewhat on issues of equity and social welfare. However, since the mid-1990s the amount of cross-fertilization between the disciplines in the social sciences in the study of development has increased significantly, and a lot if it has been driven by distributional concerns.
At the World Bank and in the donor agencies, the hiring of noneconomists was initially prompted by the need to assess the social and environmental consequences of infrastructure projects, such as population resettlement that accompanied dam construction (Cernea 1988, 1999). From the mid-1980s onward, however, and especially from the 1990s, the World Bank’s poverty assessments were required to have a “qualitative” component that went beyond, and complemented, the standard distributional analysis from a representative household survey (Carvalho and White 1997). This requirement was replicated in other donors’ reports as well, meaning that methods particular to anthropology and sociology were now part of reports that were formerly entirely economics oriented, and still continued to be primarily economics oriented.
The push toward multidimensionality has moved alongside, and is related to, the push toward greater interdisciplinarity, or at least multidisciplinarity, in the discourse on distribution and poverty. In the mid-1980s the dominant approach in the World Bank and elsewhere in international agencies was the quantitative-economics one. A broadening has undoubtedly occurred since then. The juxtaposition of the methods of different disciplines, of household survey analysis combined with econometric analysis on the one hand, with participatory poverty analysis, unstructured interviews, and discourse analysis on the other, has led to a creative tension in these agencies and in the wider development studies community. This is an interaction that was simply not present before the mid-1980s.
Kanbur 2003b characterizes the “qualitative-quantitative” spectrum as being composed of five dimensions: (1) type of information, nonnumerical to numerical; (2) type of population coverage, specific to general; (3) type of population involvement, active to passive; (4) type of inference methodology, inductive to deductive; and (5) type of disciplinary framework, broad social science to neoclassical economics. In a subsequent paper (Kanbur and Shaffer 2007a), the divide is characterized in more fundamental epistemological terms, between the empiricist-positivist tradition and the tradition of critical hermeneutics. The latter suggests that tensions will always remain when different disciplines, and their methods, are brought to bear on the common problem of understanding poverty and distributional outcomes.
However, the greater interplay of disciplines and their methods has undoubtedly brought dividends, as shown, for example, in the compilation of studies in Kanbur and Shaffer 2007b, and in a number of other studies, all of which date from the early years of the 2000s. One example is an assessment of conditional cash transfers (discussed in the next section) in Nicaragua and Turkey. The donors required that both quantitative and qualitative methods be used. The results of the evaluation are presented and assessed by Adato (2008). The quantitative part of the assessment followed the best practice of randomized evaluation, discussed in the previous section. In Nicaragua, for example, “Out of 42 comarcas, 21 were randomly selected into the program, and 21 into the control group. Household and individual level data was collected in 2000, before the intervention began, in both control and treatment localities. Data on the same variables was then collected in the same households in 2002” (Adato 2008, 7). The impact of the program was then estimated through the double-difference method, to control for unobserved differences across communities and over time. For the qualitative component, however, “field researchers, with B.A. or M.A. degrees in sociology or anthropology, conducted research in two communities each (for a total of six communities in each study) over a period of 4–5 months, moving between them at different intervals. The field researchers resided with families in the communities” (Adato 2008, 9).
The quantitative analysis established that the programs were having the effects intended. However, the qualitative analysis revealed a number of features that could threaten the sustainability of the program in the long term. The targeting criteria, while performing well on their own terms, did not resonate with the population on the ground, creating tensions in the community among those receiving the transfer and those who could not understand why they were excluded. Also, having understood that transfers depended on their child falling below a certain nutritional assessment, the qualitative analysis indicated that some mothers were deliberately underfeeding to satisfy program requirements—fortunately, this led to a program redesign.
Perhaps the most radical departure in economics since the mid-1980s is the “behavioral revolution,” which incorporates insights from psychology into development. Daniel Kahneman was awarded the 2002 Nobel prize in economics for his work on this topic, and the essence of the approach is captured in Kahneman (2011). This behavioral perspective has now begun to seep into development economics and into the analysis of poverty and income distribution (see, for example, Datta and Mullainathan 2014; Jäntti, Kanbur, and Pirttilä 2014a, 2014b). The inroads being made into conventional development economics are further highlighted by the fact that the World Bank’s 2015 World Development Report is on this topic (World Bank 2014). One example of the value added by the behavioral perspective is a reassessment of interventions to encourage savings by introducing a range of commitment devices (Karlan 2014).
There are thus a growing number of examples of the benefits of interdisciplinarity in the analysis of poverty and distribution. These examples were simply not there when Kanbur 1987 was written. Despite the tensions, this trend is set to continue.
A lively debate had begun 30 years ago among philosophers on “legitimate” inequality. In particular, the issue was what role personal responsibility had in inequality. Dworkin (2000) is recognized as a prime mover in this debate, but other names include Cohen (2008), Arneson (2013), and Sen (2001). Roemer (1998) brought this debate squarely into the mainstream of economic analysis, and it entered the development economics discourse with the World Bank’s 2006 World Development Report (World Bank 2005). There is now a vibrant literature on theory (Fleurbaey and Maniquet 2012) and an ever-growing set of papers with empirical applications (for a recent survey, see Brunori, Ferreira, and Peragine ). A fruitful interdisciplinary interaction between economics and philosophy continues to occur.
Roemer’s (1998) formulation is now central to the literature. He considers an outcome variable of interest (income, for instance) and distinguishes conceptually between two sets of variables to which can be attributed variation in this outcome variable across the population. The first set are “circumstances,” over which the individual has no control (gender, race, parental wealth, and so forth), while the second set are labeled “effort,” being variables the individual can control. At the conceptual level, variation due to circumstances is a legitimate target of redistribution—it is inequality of opportunity. How big is inequality of opportunity relative to opportunity of income? One way to quantify this is to specify the circumstance variables and, through nonparametric inequality decomposition or parametric regression analysis, attribute a portion of the variation to circumstances. The ratio of this amount to the total variation is then a measure of the inequality of opportunity. This is the method followed by Paes de Barros and others (2009), using a set of methods that are growing in use in policy circles (Brunori, Ferreira, and Peragine 2013).
These developments in the inequality literature have been criticized recently by Kanbur and Wagstaff (2014). Treating the residual variation beyond that explained by the specified circumstance variables as reflecting effort and only effort surely understates inequality of opportunity. The residual includes luck, some of which may have been taken on willingly, but other parts of ill fortune may not have been of the individual’s doing. There are also unmeasured circumstance variables like innate talents, which are not the result of effort by the individual, or the individual’s environment including peer effects. Finally, some of the circumstances may themselves influence effort, and this effect should also be incorporated into inequality of opportunity. Some empirical approaches to address these issues have been developed, but they are under debate. It looks as though the inequality-of-opportunity discourse will continue to be present in a way that it was not 30 years ago.
Policies and Interventions
The case can be made that none of the examples considered above, for theory and empirics, point to anything new since the mid-1980s. All of them—intrahousehold inequality, evolution of inequality, interdisciplinarity, and so on—have antecedents in the previous period. This is certainly true, and this discussion is only highlighting a tendency and a pattern, rather than an absolute break in the mid-1980s. The difficulty of delineating a sharp break is most acute in the case of policy debates and interventions, since most debates of this type are generic and eternal in nature, and most interventions have been discussed if not actually implemented at some time in the past. Trade policy, redistribution and transfers, expenditure on health and education, and others are all the subject of intense debate now, but they have surely been discussed before. However, the four issues highlighted in this section do deserve their characterization as being somewhat new to the scene, at least in comparison to the vantage point of the mid-1980s.
Conditional Cash Transfers
Conditional cash transfers (CCTs) are no exception to the maxim that there is nothing new under the sun. They existed before the mid-1980s—public works schemes for famine relief are to be found in ancient times. However, there has been an explosion of these schemes, and they have risen to prominence in the policy and analytical debate only from the 1990s onward, and especially in the 2000s. The Employment Guarantee Scheme of Maharashtra State in India dates from the early 1970s, but in recent years programs that make transfers conditional on school attendance of children and other requirements have been introduced in Bangladesh, Brazil, Mexico, Nicaragua, Turkey, and a host of African countries as well (Kakwani, Soares, and Son 2005; Das, Do, and Ozler 2004; Levy 2006).
This proliferation of CCTs is remarkable. What explains the sudden interest in them, and what lessons have been learned from their operation so far? The answer to the first part of the question is twofold—growing inequality and positive evaluation results. CCTs can be seen as a response to the rising inequality that has accompanied growth in many countries. Even where growth has reduced poverty despite rising inequality, as discussed in a previous section distributional concerns have persisted and grown. One reason may be that the increasing inequality may be picking up ground-level realities that are missed by official statistics. For example, in many countries poverty measures that emphasize the depth of poverty are falling more slowly or even rising. In many more countries the overall poverty decline is an aggregation of significant numbers of winners and losers. Although the winners outnumber the losers, and their climb out of poverty is to be celebrated, the plight of those who have become poor or whose poverty has increased cannot be ignored.15 This suggests there is a role for redistributive policy to target the losers from economic reform and economic crisis. Furthermore, conditioning the transfer on behavior can help induce changes such as keeping children in school or increasing visits to health centers. Many governments have looked to CCTs to address the issue of increasing inequality in the context of economic reform and liberalization.
In addition to the felt need is the simple fact that evaluations of these programs have shown positive results. The evaluation of Mexico’s Progresa-Oportunidades program played a key role. The designers of the program incorporated evaluation using appropriate control groups right from the start. To quote Levy (2006, 37), “Between October 1997 and November 1999, a total of 24,000 families in 506 localities were interviewed regularly. Of those localities, 320 that were incorporated in the program as of October 1997 were in the initial treatment group and 186 were in the control group, until they were incorporated in the program in late 1999.” It was argued that the assignment to treatment and control was effectively random, so that randomized evaluation techniques could be implemented. The evaluation continued in similar fashion as the program was scaled up. The positive results that emerged from the evaluation were accepted by the technical community because of the methods used,16 helping greatly in the spread of the CCT message from country to country.
Thus, CCTs are now being implemented in a large number of countries, and the evaluations are mostly positive (Fiszbein and Schady 2009). What are the lessons being learned? The previous section discussed the sometimes surprising findings of qualitative as compared with quantitative analysis. Design details matters, and for long-term sustainability these qualitative issues have to be considered. Institutional details and the monitoring and accountability of officials are always important. Generic issues that arise with achieving the objectives of any program include low participation and fungibility (Das, Do, and Ozler 2004). Trade-offs between coverage and leakage, and between the specific objectives of the program and what people would like to spend the cash on, are ever present, must be handled in the design, and cannot be wished away.
Governance and Institutions
A constellation of related issues has come to the fore since the mid-1990s in a way that was not present before the mid-1980s. These issues broadly have to do with the role of institutions in the development process. Although applicable to the development process in general, they resonate particularly with poverty and distributional outcomes. If growth itself depends on the quality of institutions, then so does poverty reduction. If, furthermore, the distribution of gains from growth also depends on the nature of institutions, the importance of these factors is additionally magnified. These perspectives are well captured by Acemoglu and Robinson (2012). The literature has had both macro and micro strands.
The analysis of these factors was boosted by the publication of comparable cross-country governance indicators beginning in the mid- and late 1990s,17 which fed into and played a central part in the blossoming of the cross-country growth regressions literature discussed in the previous section. It remains one-half of the “institutions versus geography” debate—is growth better explained by the favorable (or otherwise) geographical location of the country, or by the institutions that regulate economic activity, including institutions of economic policymaking? From the policy perspective, however, the important question is—even if a relationship with the quality of institutions in a general sense is found, what specifically can be done about it? As Rodrik (2008, 100) says,
Desirable institutions provide security of property rights, enforce contracts, stimulate entrepreneur-ship, foster integration in the world economy, maintain macroeconomic stability, manage risk-taking by financial intermediaries, supply social insurance and safety nets, and enhance voice and accountability. But as the variety of institutional forms that prevail in the advanced countries themselves suggests…, each one of these ends can be achieved in a large number of different ways.
Therein lies the difficulty in generalizing, and finding “best practices” to transfer across countries. Country and context specificity is the constant pull away from generalizations based on cross-country regressions using the sorts of indicators developed by Kaufmann, Kraay, and Mastruzzi (2008). In any event, changing deep institutions is a hard task, not easily done by simply “transferring best practice.”
The more micro-oriented strand in the literature is more interdisciplinary and, not surprisingly, also emphasizes context specificity. It is perhaps best exemplified by the World Bank’s World Development Report 2004: Making Services Work for Poor People (World Bank 2003). The report focuses on “the relationships of accountability—between policymakers, service providers, and citizens” (World Bank 2003, xv) and proposes that this can be done by “(i)ncreasing poor clients’ choice and participation in service delivery [so] they can monitor and discipline providers… (r)aising poor citizens’ voice, through the ballot box and making information widely available… [and] by rewarding the effective and penalizing the ineffective delivery of services to poor people” (World Bank 2003, 1). But once again, the question might be asked—how exactly are these three things to be done?
The past 20 years have seen a move away from policy prescriptions disembodied from the institutional context, and toward a more explicit recognition of the importance of institutional structure. This is a major difference from the mid-1980s. Indeed, Kanbur 1987 reflects the somewhat technocratic tenor of the times. However, with the recognition of the importance of institutions on the policy front, and after an initial phase of attempting to implement uniform “best practices,” there is now considerable uncertainty encapsulated in the importance of context specificity of appropriate institutional design.
Macroeconomic Crises and Safety Nets
As noted earlier, most macroeconomic policy debates have a perennial feel to them. Issues of exchange rate management, fiscal balance, monetary policy, and so on were very much present in the 1980s and before. In this sense, therefore, far less is new about the macroeconomic perspective on poverty.
However, one macroeconomic issue was not as prominent as it is now. Perhaps the most striking global macroeconomic phenomena of the past three decades were the financial crises of 1997 and 2008. The first was centered on East Asia but with repercussions around the world. The implications for growth and poverty reduction were dramatic. By some estimates, the gains of a decade were wiped out overnight in countries such as Indonesia.18 The crisis of 2008 was even more severe, leading the world into a recession that came close to matching the Great Depression of the 1930s. Except through natural disasters, such rapid reversals, such vulnerability, were not really part of the mental makeup of the mid-1980s.
The global macroeconomic crises have influenced the policy discourse. Research of the past few years, including from the IMF, has been cautious about the benefits of capital market integration, and has emphasized the need to manage the risks and to carefully sequence the opening up of capital markets (for example, Kose and others 2006). Asian countries built up significant reserves to protect themselves against outflows, and perhaps against being forced to have recourse to the IMF. The crisis of 2008 similarly caused a rethinking of financial sector deregulation on the one hand, and on austerity policies during recessions on the other.
From the point of view of poverty and distribution, however, the crises of the late 1990s and the late 2000s brought to the fore like never before the issue of safety nets and cast it in a newer, sharper, light. The discourse of the 1980s had developed a negative view of transfer schemes such as food subsidies, and broader social security schemes like pensions. It was argued, quite rightly in many cases, that these schemes represented major fiscal exposures while at the same time being poorly targeted toward the poor. A leading illustration in the 1970s, for example, was the Sri Lanka rice subsidy scheme, which was dismantled in the late 1970s and early 1980s (Anand and Kanbur 1991). However, the macroeconomic crises, together with possible negative effects on some poor of policy reform, led in the 1990s to a revisiting of transfer schemes as a response to crisis. The popularity of public works schemes to counter short-term downturns in economic activity grew. India’s National Rural Employment Guarantee Act of 2004 is one of the largest in the world. True, the discourse on the rationale for these schemes mixes up the micro and the macro, but the sharp macroeconomic downturns experienced by many countries in the late 1990s and the late 2000s, and the expectation that such downturns were more likely in a globalized world economy, is a significant explanation. The consensus against safety nets has turned and, fully cognizant of the lessons of the earlier period on targeting and on implementation, and in concert with other rationales for transfers (such as keeping children in school), there is now far greater acceptance of them as policy instruments to manage distributional risk from macroeconomic and microeconomic shocks.19 At the same time, there is a far greater appreciation of external assistance in helping countries address the poverty and inequality consequences of sharp macroeconomic downturns.20
Global Public Goods and Distribution
The financial crises of the 1990s highlighted the linkages in the world economy and the vulnerabilities of the poor in one economy to events in other countries, perhaps very far away. Since the 1990s the policy discourse has become attuned to global public goods, or more generally cross-border public goods and the role of international public policy in providing them. The issues covered under this umbrella include greenhouse gas emissions, global forest cover, migration and refugees, financial contagion, water basin and riparian rights management, and the spread of disease.
There is nothing particularly novel at the conceptual level about these phenomena. The public economics techniques of the 1980s and earlier can be easily adapted to frame these issues. When there are cross-border externalities, there is a gain from coordination, and thus a gain to development, if some or all of the countries involved are developing countries. But the coordination mechanism itself is a public good, and a tendency to underinvest in such mechanisms will occur for the usual reasons. If the costs of this coordination can be borne by richer countries, then to the extent that the cross-border public good benefits poor countries, these resource inputs can legitimately be claimed to be development assistance.21
All of the above was understood 20 or more years ago. What is new is the veritable explosion of discussion, debate, and action on the design and financing of these coordination mechanisms. Particularly in the environmental arena, from the Montreal Protocol to the Kyoto agreement to the Clean Development Mechanism and carbon trading, there has been continuous concern and response to the concern during the past two decades. Indeed, the continued concern about climate change and how it and the attempts to mitigate it will affect poor countries and the poor in poor countries—the “climate justice” issue—will ensure that this policy issue stays on the agenda in the coming decades (Kanbur 2014a). Yet it was almost totally absent from the discourse 20 years ago.
This chapter adopts the “Rip Van Winkle” stratagem of asking what differences would be noticed, in the domain of poverty and distribution, by someone who fell asleep in 1985 and woke up in 2015. In one sense there is tremendous continuity—the discourse on poverty lines, poverty measurement, the inequality-growth relationship, fiscal balance, costs of inflation, public works schemes, food and energy subsidies, and so on would be very familiar to a visitor from three decades ago. But there are also discernible differences. This chapter highlights 10 such differences under three broad headings. Under facts and empirics the chapter emphasizes (1) the tremendous increase in household survey information for developing countries, particularly for Africa; (2) the sharp trend toward increasing inequality in many countries over the past 30 years; and (3) the dominance of macroeconomic cross-country regressions and the rise of microeconomic randomized evaluations, in contrast to the demise of computable general equilibrium models and project evaluation. Under concepts and theory the chapter highlights (1) poverty dynamics and risk, (2) gender and intrahousehold inequality, and (3) multidisciplinarity as distinguishing features of the past two decades. Finally, under policies and interventions the chapter picks out (1) conditional cash transfers, (2) governance and institutions, (3) macroeconomic crises, and (4) global public goods as being part of the discourse on poverty and distributional policy in a way that they were not when Kanbur 1987 was written three decades ago.
2012. Why Nations Fail: The Origins of Power, Prosperity and Poverty. New York: Crown Business.
2008. “Integrating Survey and Ethnographic Methods to Evaluate Conditional Cash Transfer Programs.” IFPRI Discussion Paper No. 00810, International Food Policy Research Institute, Washington.
1974. Economic Growth and Social Equity in Developing Countries. London: Oxford University Press.
1976. “Inequality, Poverty and Development.” Journal of Development Economics3 (4): 307–42.
1978. “The Economics of ‘Tagging’ as Applied to the Optimal Income Tax, Welfare Programs and Manpower Planning.” American Economic Review68 (1): 8–19.
1995. “Unitary Versus Collective Models of the Household: Is It Time to Shift the Burden of Proof?” World Bank Research Observer10 (1): 1–19.
2011. “Counting and Multidimensional Poverty Measurement.” Journal of Public Economics95 (7–8): 476–87.
1991. “Public Policy and Basic Needs Provision in Sri Lanka.” In The Political Economy of Hunger, Volume III: Endemic Hunger, edited by J. Dreze and A.Sen.Oxford, U.K.: Clarendon Press.
1993a. “Inequality and Development: A Critique.” Journal of Development Economics41 (1): 19–43.
1993b. “The Kuznets Process and the Inequality-Development Relationship.” Journal of Development Economics40 (1): 25–52.
2013. “Equality of Opportunity: Derivative Not Fundamental.” Journal of Social Philosophy44 (4): 316–30.
1973. “Some Ordinalist-Utilitarian Notes on Rawls’s Theory of Justice.” Journal of Philosophy70 (9): 245–63.
1970. “On the Measurement of Inequality.” Journal of Economic Theory2 (3): 244–63.
1980. Lectures in Public Economics. Maidenhead, U.K.: McGraw-Hill.
2014. “The Miracle of Microfinance? Evidence from a Randomized Evaluation.” MIT Department of Economics, Cambridge, Massachusetts. http://economics.mit.edu/files/5993.
2003. “The World Bank of the Future.” American Economic Review, Papers and Proceedings93 (2): 39–44.
2011. “Household Panel Data Sets in Developing and Transition Countries (Version 2, March 2011).” Chronic Poverty Research Centre. http://www.chronicpoverty.org/uploads/publication_files/Annotated_Listing_of_Panel_Datasets_in_Developing_and_Transitional_Countries.pdf.
1999. Randomness in the Experimental Samples of PROGRESA (Education, Health and Nutrition Program). Washington: International Food Policy Research Institute.
1988. “Food Subsidies and Poverty Alleviation.” Economic Journal98 (392): 701–19.
1992. “Collective Models of Household Behavior: An Introduction.” European Economic Review36 (2/3): 355–64.
1998. “Efficient Intra-Household Allocation: A General Characterization and Empirical Tests.” Econometrica66 (6): 1241–78.
2013. “Inequality of Opportunity, Income Inequality and Economic Mobility: Some International Comparisons.” ECINEQ Working Paper No. 2013–284, Society for the Study of Economic Inequality, Verona, Italy.
2009. “Chronic Poverty and All That: The Measurement of Poverty over Time.” In Poverty Dynamics: Interdisciplinary Perspectives, edited by A.Addison, D.Hulme, and R.Kanbur.Oxford, U.K.: Oxford University Press.
1997. “Combining the Quantitative and Qualitative Approaches to Poverty Measurement and Analysis.” World Bank Technical Paper No. 366, World Bank, Washington.
1988. Involuntary Resettlement in Development Projects. Policy Guidelines for World Bank Financed Projects. Washington: World Bank.
1999. “The Need for Economic Analysis of Resettlement: A Sociologist’s View.” In The Economics of Involuntary Resettlement: Questions and Challenges, edited by M. M.Cernea.Washington: World Bank.
2007. “China’s (Uneven) Progress against Poverty. Journal of Development Economics82 (1): 1–42.
1975. Patterns of Development, 1950–1970. Oxford: Oxford University Press.
Chronic Poverty Research Centre. 2005. The Chronic Poverty Report, 2004–05. Manchester, U.K.: Institute for Development Policy and Management, University of Manchester. http://www.chronicpoverty.org/uploads/publication_files/CPR1_ReportFull.pdf.
2007. “Evaluating Alternative Approaches to Poverty Alleviation in Madagascar: Rice Tariffs versus Targeted Transfers.” Fiscal Affairs Department, International Monetary Fund, Washington.
2008. Rescuing Justice and Equality. Cambridge, Massachusetts: Harvard University Press.
2004. “Conditional Cash Transfers and the Equity-Efficiency Debate.” World Bank, Washington.
2014. “Behavioral Design: A New Approach to Development Policy.” Review of Income and Wealth60 (1): 7–35.
2009. “Instruments of Development: Randomization in the Tropics, and the Search for the Elusive Keys to Economic Development.” Proceedings of the British Academy, 2008 Lectures162: 123–60.
2005. “Data and Dogma: The Great Indian Poverty Debate.” World Bank Research Observer20 (2): 177–99.eds.
1996. “A New Data Set Measuring Income Inequality.” World Bank Economic Review10 (3): 565–91.
1998. “New Ways of Looking at Old Issues: Inequality and Growth.” Journal of Development Economics57 (2): 259–88.
2004. “Growth and Shocks: Evidence from Rural Ethiopia.” Journal of Development Economics74 (2): 309–29.
2002. “Growth Is Good for the Poor.” Journal of Economic Growth7 (3): 195–225.
2007. “Using Randomization in Development Economics Research: A Toolkit.” In Handbook of Development Economics, Vol. 4, edited by T.Paul Schults and JohnStrauss,3895–62. North Holland: Elsevier Science Ltd.
2000. Sovereign Virtue: The Theory and Practice of Equality. Cambridge, Massachusetts: Harvard University Press.
1999. “The Effect of Household Wealth on Educational Attainment: Evidence from 35 Countries.” Population and Development Review25 (1): 85–120.
2009. Conditional Cash Transfers: Reducing Present and Future Poverty. World Bank Policy Research Report. Washington: World Bank.
2012. Equality of Opportunity: The Economics of Responsibility. Singapore: World Scientific Publishing Company.
2009. “A Class of Chronic Poverty Measures.” In Poverty Dynamics: Interdisciplinary Perspectives, edited by A.Addison, D.Hulme, and R.Kanbur.Oxford, U.K.: Oxford University Press.
1984. “A Class of Decomposable Poverty Measures.” Econometrica52 (3): 761–76.
2008. “Male Wages and Female Welfare: Private Markets, Public Goods, and Intrahousehold Inequality.” Oxford Economic Papers60 (1): 42–56.
1990. “How Serious Is the Neglect of Intrahousehold Inequality?” Economic Journal100 (402): 866–81.
International Monetary Fund (IMF). 2014. “Fiscal Policy and Income Inequality.” IMF Policy Paper, Washington. http://www.imf.org/external/np/pp/eng/2014/012314.pdf.
2000. “Is Transient Poverty Different? Evidence for Rural China.” Journal of Development Studies36 (6): 82–99.
2014a. “Poverty, Development, and Behavioral Economics.” Review of Income and Wealth60 (1): 1–6.
2014b. “Poverty and Welfare Measurement on the Basis of Prospect Theory.” Review of Income and Wealth60 (1): 182–205.
2014. “World Income Inequality Databases: An Assessment of WIID and SWIID.” ISER Working Paper No. 2014–31, Institute for Social and Economic Research, University of Essex. https://www.iser.essex.ac.uk/research/publications/working-papers/iser/2014-31.pdf.
2011. Thinking, Fast and Slow. New York: Farrar, Strauss and Giroux.
2005. “Conditional Cash Transfers in African Countries.” Working Paper No. 9, International Poverty Centre, United Nations Development Programme, Brasilia, Brazil. http://www.ipc-undp.org/publication/27379.
1987. “Measurement and Alleviation of Poverty: With an Application to the Effects of Macroeconomic Adjustment.” IMF Staff Papers34 (1): 60–85.
1990a. “Poverty and Development: The Human Development Report and The World Development Report, 1990.” In Poverty Monitoring: An International Concern, edited by Rolph vander Hoeven and RichardAnker.New York: St. Martin’s Press.
1990b. Poverty and the Social Dimensions of Structural Adjustment in Côte d’Ivoire. SDA Working Paper Series. Washington: World Bank.
2003a. “IFI’s and IPG’s: Operational Implications for the World Bank.” In Challenges to the IMF and the World Bank, edited by ArielBura.New York: Anthem Press.
2003b. Q-Squared: Qualitative and Quantitative Methods of Poverty Appraisal. New Delhi: Permanent Black.ed.
2004. “Cross-Border Externalities, International Public Goods, and Their Implications for Aid Agencies.” In Global Tensions: Challenges and Opportunities in the World Economy, edited by L.Beneria and S.Bisnath.London: Routledge.
2005. “Regional Versus International Financial Institutions.” In Regional Public Goods: From Theory to Practice, edited by A.Estevadeordal, B.Frantz, and T. R.Nguyen.Inter-American Development Bank: Washington.
2006. “New Directions in Development Economics: Theory or Empirics?” Economic and Political Weekly40 (40): 4328–46.ed.
2010a. “Globalization, Growth and Distribution: Framing the Questions.” In Equity in a Globalizing World, edited by RaviKanbur and A.Michael Spence,41–70. Washington: World Bank for the Commission on Growth and Development.
2010b. “Macro Crises and Targeting Transfers to the Poor.” Journal of Globalization and Development1 (1): Article 9.
2012. “Stress Testing for the Poverty Impacts of the Next Crisis.” In Knowing, When You Do Not Know, edited by AmbarNarayan and CarolinaSánchez-Páramo,50–55. Washington: World Bank.
2014a. “Education for Climate Justice.” The Climate Justice Dialogue. http://www.mrfcj.org/pdf/faces-of-climate-justice/Education-for-Climate-Justice.pdf.
2014b. “Globalization and Inequality.” In Handbook of Income Distribution, Vol. 2B, edited by A. B.Atkinson and F.Bourguignon.Amsterdam: Elsevier.
2000. “Why Is Inequality Back on the Agenda?” Annual World Bank Conference on Development Economics, 1999. Washington: World Bank.
2007. “Premature Mortality and Poverty Measurement.” Bulletin of Economic Research59 (4): 339–59.
2007a. “Epistemology, Normative Theory and Poverty Analysis.” World Development35 (2): 183–96.
2007b. Experiences of Combining Qualitative and Quantitative Approaches in Poverty Analysis. Special issue of World Development35 (2).eds.
2014. “How Useful Is Inequality of Opportunity as a Policy Construct?” Policy Research Working Paper No. 6980, World Bank, Washington.
2005. “Fifty Years of Regional Inequality in China: A Journey through Revolution, Reform and Openness.” Review of Development Economics9 (1): 87–106.
2012. “Confronting Rising Inequality in Asia.” In Asian Development Outlook 2012. Washington: Asian Development Bank.
2014. “Savings by and for the Poor: A Research Review.” Journal of Income and Wealth60 (1): 36–78.
2008. “Governance Matters VII: Aggregate and Individual Indicators, 1996–2007.” World Bank Policy Research Paper No. 4654, World Bank, Washington. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1148386.
2006. “Financial Globalization: A Reappraisal.” IMF Staff Papers56 (1): 8–62.
1955. “Economic Growth and Income Inequality.” American Economic Review45 (1): 1–28.
2006. Progress against Poverty: Sustaining Mexico’s Progresa-Oportunidades Program. Washington: Brookings Institution Press.
2003. “Measuring Vulnerability.” Economic Journal113 (486): C95–C102.
1991. “Project Appraisal and Planning Twenty Years On.” Proceedings of the World Bank Annual Conference on Development Economics, 1990. Washington: World Bank.
1980. “Marriage and Household Decisionmaking: A Bargaining Analysis.” International Economic Review21 (1): 31–44.
1987. The Theory of Taxation for Developing Countries. Oxford: Oxford University Press.eds.
2014. “Redistribution, Inequality and Growth.” IMF Staff Discussion Note No. 14/02, International Monetary Fund, Washington. http://www.imf.org/external/pubs/ft/sdn/2014/sdn1402.pdf.
2009. Measuring Inequality of Opportunities in Latin America and the Caribbean. Washington: World Bank.
2014. Capital in the Twenty-First Century. Cambridge, Massachusetts: Harvard University Press.
2003. “Resources at Marriage and Intrahousehold Allocation: Evidence from Bangladesh, Ethiopia, Indonesia, and South Africa.” Oxford Bulletin of Economics and Statistics65 (3): 283–328.
2008. “Inequality Is Bad for the Poor.” In Inequality and Poverty Reexamined, edited by JohnMicklewright and StevenJenkins.Oxford: Oxford University Press.
1971. A Theory of Justice. Cambridge, Massachusetts: Harvard University Press.
2008. “Second-Best Institutions.” American Economic Review98 (2): 100–4.
2009. “The New Development Economics: We Shall Experiment, But How Shall We Learn?” In What Works in Development? Thinking Big and Thinking Small, edited by J.Cohen and W.Easterly.Washington: Brookings Institution Press.
2012. “Why We Learn Nothing from Regressing Economic Growth on Policies.” Seoul Journal of Economics25 (2): 137–51.
1998. Equality of Opportunity. Cambridge, Massachusetts: Harvard University Press.
1976. “Poverty: An Ordinal Approach to Measurement.” Econometrica44 (2): 219–31.
1983. “Economics and the Family.” Asian Development Review1 (2): 14–26.
2001. Development as Freedom, second edition. New York: Oxford University Press.
2013. The Price of Inequality: How Today’s Divided Society Endangers Our Future. New York: W.W. Norton and Company.
2004. Indonesian Living Standards before and after the Crisis: Evidence from the Indonesia Family Life Survey. Santa Monica, California: Rand Corporation. http://www.rand.org/pubs/monographs/MG137.html.
United Nations. 2014. “Introduction to the Open Working Group Proposal for Sustainable Development Goals,” Sustainable Development Knowledge Platform, UN Department of Economic and Social Affairs, New York. http://sustainabledevelopment.un.org/sdgsproposal.
United Nations Development Programme (UNDP). 1990. Human Development Report. New York: Oxford University Press.
World Bank. 2000. World Development Report 2000/2001: Attacking Poverty. New York: Oxford University Press.
World Bank. 2003. World Development Report 2004: Making Services Work for Poor People, New York: Oxford University Press.
World Bank. 2005. World Development Report 2006: Equity and Development. Washington: World Bank.
World Bank. 2011. World Development Report 2012: Gender Equality and Development. Washington: World Bank.
World Bank. 2014. World Development Report 2015: Mind, Society and Behavior. Washington: World Bank.
An earlier version of this chapter, with a suitably different title, was written while the author was a Visiting Scholar at the Fiscal Affairs Department of the International Monetary Fund during March–April 2007. This version provides a significant update to the review taking into account developments in the last decade.
There are many compilations of data sources; see, for example, the United Nations World Income Inequality Database (http://www.wider.unu.edu/research/Database/en_GB/database/).
DHS Program (http://dhsprogram.com/Data/).
PovcalNet web page (http://iresearch.worldbank.org/PovcalNet/).
In their review, Kanbur and Lustig (2000) conclude that “the jury is still out.” The World Bank’s overview in World Bank (2005, 103) was also agnostic. However, a body of work later emanating from the IMF has provided support for the hypothesis that inequality impedes growth; see, for example, Ostry, Berg, and Tsangarides 2014.
As Little and Mirrlees (1991, p 359) note in their own 20-year retrospective, “A battle raged in the World Bank during the 1970s about whether social prices should be used. Formally, the ‘social price brigade’ won, in that their guidelines on the use of distributional weights were actually incorporated in the Operational Manual in 1980. In practice, we believe, they were hardly ever used except in an experimental manner.”
Kanbur 1990b is one attempt at this type of exercise. The sheer complexity of these models was perhaps the primary reason for their failure in the analytical and the policy arenas. It was very difficult to explain the reasoning behind the outcomes, which lay somewhere in the interaction of the sometimes hundreds of equations and the crucial handful of “closure rules,” and the results could be sensitive to the large number of assumed parameter values.
As Rodrik (2012, 141) argues, “Consider an illustration from trade policy. The estimated coefficient on import tariffs in growth regressions run for the contemporary period is typically negative (albeit insignificantly so) and rarely positive. One frequently hears the argument that we can at least draw the conclusion from this fact that import protection cannot be beneficial to growth. But once again this and similar inferences are invalid. A negative partial correlation between growth and import tariffs is not only consistent with protection being growth-enhancing, it is actually an equilibrium consequence of trade protection being used in a socially optimal fashion.”
For example, Dercon 2004 for Ethiopia.
This perspective is advanced by Chronic Poverty Research Centre 2005.
For an early critique, see Kanbur 1990a.
The move to many dimensions raises the conceptual question of how poverty is to be measured when there is deprivation along many dimensions. The past decade has seen a burgeoning of a literature on this topic—see, for example, Alkire and Foster 2011.
The World Bank’s World Development Report 2000/2001: Attacking Poverty (World Bank 2000) introduced empowerment as a key dimension of the poverty discourse.
These and other dimensions of the disconnect between official statistics and perceptions and reality on the ground are discussed in Kanbur 2010a.
See the World Bank website http://info.worldbank.org/governance/wgi/index.aspx#home, and Kaufmann, Kraay, and Mastruzzi 2008.
For a detailed assessment of the impact of the crisis on income and non-income dimensions of the standard of living in Indonesia, see Strauss and others 2004.
Proposals for special windows of crisis assistance have been accepted by the World Bank’s soft loan window, the International Development Association (Kanbur 2012).