Copyright © 2006 Board of Trustees of the Leland Stanford Junior University
All right reserved.

ISBN: 978-0-8047-4842-1

Chapter One

Introduction The Conceptual Foundations of Poverty and Inequality Measurement

David B. Grusky and Ravi Kanbur

There is a growing consensus among academics, policy makers, and even politicians that poverty and inequality should no longer be treated as soft social issues that can safely be subordinated to more important and fundamental interests in maximizing total economic output. This newfound concern with poverty and inequality, which dates back to at least the early 1990s (see Atkinson 1997), may be attributed to such factors as (1) the dramatic increase in economic inequality in many countries over the last quarter century, (2) the rise of a "global village" in which spectacular regional disparities in the standard of living have become more widely visible and hence increasingly difficult to ignore, (3) a growing commitment to a conception of human entitlements that includes the right to seek or secure employment and thereby be spared extreme deprivation, (4) an emerging concern that poverty and inequality may have negative macro-level effects on terrorism (cf. Krueger and Maleckov�� 2003), total economic production (e.g., Bertola 2000), and ethnic unrest (e.g., Olzak forthcoming), and (5) a growing awareness of the negative individual-level effects of poverty on health, political participation, and a host of other life conditions. Although the growth of anti-inequality sentiment thus rests in part on an increased awareness of just how unequal and poverty-stricken the world is, it may also be attributed to an ever-evolving and accreting list of human rights (i.e., the "normative" account) as well as a growing appreciation of the negative externalities of inequality and poverty (i.e., a "consequentialist" account).

In what ways has the newfound concern with poverty and inequality manifested itself? This concern is, we would argue, principally revealed in the form of grassroots political mobilization on various anti-poverty platforms as well as a growing acceptance of and commitment to anti-poverty and anti-inequality initiatives among elite opinion leaders and their organizations (e.g., Millenium Development Goals). By contrast, academic research on issues of inequality and poverty has not flourished to quite the same extent, and the modest takeoff in such scholarship that is under way has focused disproportionately on matters of description and the methodological intricacies of measurement rather than more fundamental conceptual issues that, as we see it, must now be taken on (see Srinivasan 2004). The present book therefore provides an unabashedly academic approach to poverty and inequality reduction that proceeds from the radical assumption that more in the way of careful reflection and conceptual ground clearing might serve us well.

We emphasize conceptual issues not out of some intrinsic fascination with theory (although we confess to that as well) but because we think that pressing problems of policy cannot be adequately addressed without first making conceptual advances. The need for new conceptual work is especially apparent, we think, on three distinct but related fronts (see Kakwani 2004; Reddy 2004):

Defining the dimensions: Simple though it may seem, an important starting point is to develop and justify a list of valued resources that define the "inequality space," a list that presumably goes well beyond income (or wealth) alone. As many of our contributors note, there is a growing consensus that the income distribution cannot by itself satisfactorily capture the structure of poverty and inequality, yet much work remains in developing an elaborated list of endowments (e.g., schooling), investments (e.g., work experience), and living conditions (e.g., neighborhood attributes) that does suffice to describe this structure. Characterizing multidimensional space: Secondarily, new methods must be developed to measure inequality and poverty within the context of this multidimensional space, a task that is complicated because a great many parameters may be required to adequately characterize such a space. Moreover, given that various social groups (e.g., classes, ethnic groups, genders) exist within this space and constrain patterns of interaction, researchers must develop models that recognize that these groups can give rise to distinctive preferences (e.g., a "culture of poverty") that in turn affect how individuals react to poverty and inequality. Remediation in a multidimensional world: The third and final conceptual challenge is that of devising new approaches to remediation that remain viable under this more expansive definition of poverty and inequality. This task, which is arguably even more daunting than the foregoing two, requires targeting those aspects of inequality and poverty (e.g., residential segregation) that are causal with respect to many outcomes and hence likely to bring about cascades of change. On each of these three fronts, important advances have recently been made, and obviously we do not intend to minimize such advances. We wish merely to identify these three research fronts as especially deserving of continuing attention. In hopes of spurring such a commitment, we have invited six leading scholars of inequality and poverty to lay out these three conceptual challenges in more detail, to identify other challenges that should be the focus of future scholarship, and to outline possible solutions to them.

The chapters of our book address to varying degrees the three themes identified above. The first contribution, authored by Amartya Sen, lays out the case for a multidimensionalist understanding of inequality and poverty, while the following chapter by Martha Nussbaum renders the multidimensionalist approach more concrete by developing an explicit list of "fundamental entitlements." In the next chapter, Fran��ois Bourguigon likewise argues that the old income paradigm is unduly limiting, but he goes on to emphasize that a new multidimensionalist approach must solve major problems in measurement and modeling to achieve a "level of operationality" comparable to that currently enjoyed by the old paradigm. The final three chapters all attend to the rise of social groups within multidimensionalist space and the importance of developing a measurement approach that captures this social lumpiness. The contributions of William Wilson and Douglas Massey address, in particular, the spatial concentration of poverty and the associated rise of an "underclass" and racially segregated neighborhoods, while the final contribution by Martha Fineman examines the peculiar institution of gender in which males and females are assembled together into families on spectacularly unequal terms.

As the foregoing makes evident, we have developed a book that is unapologetically conceptual in its approach, but hopefully not one that, by virtue of this emphasis, appeals only to ivory-tower types. We devote the rest of this introductory chapter to an intellectual history of inequality and poverty scholarship that situates our contributors in their disciplinary context and thus prepares the uninitiated to nonetheless profit from these contributions. We concentrate on the disciplines of economics and sociology because of our own backgrounds and because the most sustained commitment to understanding poverty and inequality is perhaps found within these disciplines.

This intellectual history reveals rather stark disciplinary differences in how the study of inequality and poverty have been pursued. These differences emerge, in part, because both disciplines have long-standing research traditions on poverty and inequality, traditions that are so well developed that their distinctive approaches have crystallized and differences have become magnified. This is not to suggest that the disciplines have developed independently of one another. Indeed, each discipline has worked with stylized and outdated understandings of the orientation of the other discipline, a state of affairs that this book seeks to begin to rectify. The purpose, then, of our introductory chapter is to rehearse the main conceptual tools with which the disciplines of economics and sociology have historically sought to organize and make sense of inequality and poverty. We do so separately for each of the two disciplines and then, at the close of the chapter, outline the conceptual issues that both disciplines should begin to address in analyzing poverty, inequality, and distributional questions more generally. This exercise will, we think, set an agenda in this area for the social sciences and cognate disciplines, an agenda to which the disciplines could contribute in their own particular way, singly or in concert.


We begin with a characterization, perhaps controversial, of the last thirty years of research on distributional questions in economics, especially development economics. Somewhat arbitrarily, consider the period beginning with Atkinson's classic 1970 paper "On the Measurement of Inequality" (Atkinson 1970), and ending with Atkinson and Bourguignon's state-of-the-art edited volume, Handbook of Income Distribution (2000). These thirty years may be divided, very roughly, into a first phase stretching from the 1970s to the mid-1980s and a second phase stretching from the mid-1980s to the end of last century. The first phase was one of great conceptual ferment and was exciting for that reason, whereas the second phase was focused on consolidation, application, and fierce policy debates, especially on the distributional consequences of macroeconomic policies in developing and transitional economies. We review each of these two phases below.

The First Phase: Conceptual Ferment

The first phase, covering the 1970s and early 1980s, was one of conceptual ferment around four broad questions:

How should inequality and poverty be measured? Should policy recommendations on issues of poverty reduction and equalization rest on simple utilitarian premises? Are households best treated as unitary entities? Can the complicating effects of social interaction be readily incorporated into analyses of poverty and inequality?

We consider in turn each of these conceptual questions and the conceptual ferment that they engendered.

The debate over how inequality and poverty are best measured has a long and distinguished history. In the 1970s, three key contributions defined this ongoing debate: Atkinson's 1970 paper "On the Measurement of Inequality" (Atkinson 1970), Sen's 1973 book On Economic Inequality (Sen 1973), and Sen's 1976 paper "Poverty: An Ordinal Approach to Measurement" (Sen 1976). These contributions provided a way into conceptualizing and operationalizing value judgments on distributional issues, serving as an antidote to a natural instinct among economists to avoid distributional questions, an instinct that goes back to debates in the 1930s launched by Robbins' The Nature and Significance of Economic Science (Robbins 1932). The appearance of these papers in the 1970s sparked a prolonged discussion of how to incorporate distributional value judgments. In the literature on poverty measurement, the culmination of this process was undoubtedly the famous 1984 paper by Foster, Greer, and Thorbecke, "A Class of Decomposable Poverty Measures" (Foster, Greer, and Thorbecke 1984). The poverty measure developed in that paper has now become the workhorse of applied work on poverty the world over.

The debate on utilitarianism and its usefulness in policy prescription also emerged in the 1970s as philosophical discourse began to enter and enrich economic work on distributional issues. The influential paper by Mirrlees, "An Exploration in the Theory of Optimal Income Taxation" (Mirrlees 1971), is famous for many reasons (including winning a Nobel Prize), but perhaps especially for its application of thoroughgoing utilitarianism to the policy question of how progressive income taxation should be. The shortcomings of such utilitarian fundamentalism were, by contrast, highlighted by Sen in a number of well-known works, including his 1987 book, The Standard of Living (Sen 1987). In a related line of analysis, Arrow (1973) introduced Rawls (1971) to mainstream economists in terms they would understand, namely maxi-min strategies in the face of uncertainty (also Nozick 1974). As Rawls argued, when people are placed under a "veil of ignorance," they should rationally support a constitution that aims for the greatest good of the worst off because, "but for the grace of God," any one of them could be the worst off. As a sign of this new intercourse between economics and philosophy, such journals as Economics and Philosophy and Philosophy and Public Affairs were filled with contributions from both disciplines.

At the same time, economists were also completing the conceptual work necessary to represent processes of social interaction in a wide range of economic models, including those pertaining to poverty and inequality. Within the rational choice framework, Akerlof, Spence, and Stiglitz sought to incorporate issues of imperfect and asymmetric information into economic models, thereby launching a body of work that won them the Nobel Prize in 2001. This framework was used by Akerlof (1970) and Stiglitz (1973) to analyze the underclass in developed economies and to explain why the very poor in developing countries failed to invest much in education. It was argued that, in the presence of imperfect and asymmetric information, the market economy can produce multiple equilibria, some more efficient and more equitable than others, and that public action and intervention was necessary to move away from the "bad" equilibria.

As a final example, we move to debates about the proper unit of analysis for poverty research, debates that flourished during this period because of concerns that the usual household-based analyses ignored intra-household exchange and thus glossed over inequalities prevailing within households. The obvious starting point here is again the work of Sen. In the 1970s and early 1980s, a series of publications (some of which are reprinted in Sen 1984) brought home to economists that "unitary" models of the household, models that ignore intra-household inequalities, simply could not capture or explain the evidence on deprivation among females in developing countries. Although slow to develop, this line of inquiry ultimately blossomed, leading to much important applied and policy analysis.

The Second Phase: Consolidation, Application, and Policy Debate

In the mid-1980s, the foregoing conceptual ferment on distributional issues gradually died down, and the field moved into a new phase of consolidation, application, and policy debate. This second phase was neither less useful or less exciting, just conceptually less innovative. How did this second phase play out? We answer this question by returning to each of the four arenas discussed in the prior section and considering how the literature in those arenas developed.

With respect to matters of measurement, the various "index wars" of the 1970s gradually waned, and attention turned to applying existing indices to data sets in rich and poor countries alike. This second phase was characterized, in particular, by increased availability of household survey data sets for developing countries. In Africa, for example, Cote d'Ivoire fielded the first high-quality nationally representative household survey in 1985, and presently more than a dozen countries have at least one such survey. Indeed, half a dozen African countries now have panel studies (in which the same households are surveyed two years in a row), as do many other countries throughout the less developed world. This increase in data availability means that the measures developed in the 1970s and 1980s will have many applications in the years to come.

The literature on intra-household bargaining and gender issues has also progressed to consolidation and application. When a group of economists (Alderman et al. 1995) wrote a paper entitled "Unitary versus Collective Models of the Household: Is It Time to Shift the Burden of Proof?" they provided a strongly affirmative answer to their rhetorical question, an answer that would once have been controversial but is no longer. There is of course still resistance from adherents of the "unitary" model, but the debate is more on the details of particular empirical tests, not on whether factors such as intra-household bargaining between the genders in principle have a role to play.


Excerpted from POVERTY AND INEQUALITY Copyright © 2006 by Board of Trustees of the Leland Stanford Junior University. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.