Rethinking Inequality through a Multidimensional Perspective:
An Analysis of Amartya Sen’s Contribution to the Assessment of Social Disparities.
Claudia Cappa, Graduate Institute of Development Studies, Geneva
The purpose of this research paper is to develop a multidimensional framework for analysing and measuring the distribution of well-being[1]. The concept of ‘well-being’ refers to an abstract and unobservable phenomenon, which encompasses the social, economic and political opportunities enjoyed by individuals. According to this perspective, standard attributes such as income or consumption represent only a partial component of well-being. Health, life expectancy, education and access to basic services are some additional dimensions of human life that have to be included in the assessment of how well-being is distributed in a given society. Participation in decision-making process and political life represents another variable that needs to be taken into consideration while measuring disparities across individuals and social groups.
Questions concerning the distribution of well-being have challenged economists since the works of D. Ricardo (1817) and W. Pareto (1895), more than 100 years ago. From these pioneering researches onwards, many important strands of economic literature have focused on the dispersion of income, or monetary value of consumption, as measures of well-being distribution. More recently, however, the focus on distribution has been shifted from monetary measures to a more comprehensive way of conceptualising well-being and inequality. Indeed, while income is the most widely used and accepted indicator of well-being, it is at the same time the most criticized. The most common criticism concerns two major issues. The first issue is about it being an adequate and comprehensive index of economic well-being; the second issue concerns its one-dimensional nature. Indeed, income has proved to be intrinsically unable to capture non-economic dimensions of well-being: even if it is argued that income is a means, no systematic relations have been established so far between monetary measures and human development.
The need to move from an income-based perspective to a broader assessment of well-being has been widely advocated by sociological and economic literature during the last years. This tendency has manifested itself with a proliferation of works on multidimensional analysis of poverty and inequality. Several authors have moved beyond income, earnings and consumption and have focused on additional dimensions of well-being like education, health and capabilities. Contributors include P. Townsend (1979), E. Maasoumi (1986), R.A. Cummins (1996), F. Stewart (1996), M.C. Nussbaum (2000), D. Narayan (2000) and A. Sen (1973, 1987, 1992, 1999).
Within the cluster of multidimensional analysis of well-being and inequality, the capability approach developed by A. Sen focuses on the freedom of choice enjoyed by individuals. Well-being is achieved through a combination of resources, conditional on personal characteristics and social constraints. The combination of both these personal and social components determines the way in which well-being is distributed across individuals and social groups. While it is widely recognised that Sen’a theory represented an innovative conceptual contribution to the analysis of well-being, some important methodological choices have to be made in order to operationalize this approach and transform its theoretical framework into variables with potential empirical value[2]. This is mainly due to the extreme richness and complexity of this approach, which engenders strong informational and methodological requirements. The question that arises is therefore how to justify the necessity of this approach, given the difficulties linked to its operationalization, both in terms of methodology and data requirements. A further concern is to prove the usefulness of a capability-based approach in the policy making aimed at improving well-being and reducing inequality. This implies being able to demonstrate that a capability-based approach is not only pertinent and feasible, but also meaningful. Indeed, apart from the methodological difficulties linked to making this approach operational, there is still little empirical evidence about the policy guidance that an approach based on capability could offer.
Our research addresses these issues and aims at making two major contributions, one methodological and one empirical in nature.
First, we will try to develop a theoretical framework, built on earlier works by Sen, for a multidimensional assessment of inequality. We define the term ‘inequality’ as a measure of well-being distribution over an average value[3]. Two main assumptions will lead our analysis. The first one is that inequality is an institutional process. This means that inequality is generated and maintained in a given institutional milieu by the interaction of economic, social and political components (Tendulkar S.D., 1983). It is therefore impossible to fully understand any given form of inequality without assessing each of these correlated attributes. In this sense, inequality can be regarded as a social and multifaceted phenomenon, rather than a result of individual circumstances and behaviours. Our second assumption is that inequality is a cumulative phenomenon. In other words, inequality reproduces itself over time as a result of the interaction of different forces (Brittain J.A., 1977; Sen A.K., 1992). Inequality is not only the consequence of disparities in different fields (social, cultural, economic, political), but also the primary cause of adverse social and economic consequences, ranging from elitist biases in public policy-making to the reinforcement of other forms of inequality. In the lights of these preliminary assumptions, our research is based on the following hypothesis: a multidimensional assessment of well-being based on capabilities, rather than the one-dimensional perspectives regularly advocated by analysts and used by policy-makers, could lead to a significant broadening of the informational space on inequality and draw a more comprehensive picture of well-being than the one provided by standard approach and measures. More specifically, a multidimensional approach is needed to assess the different factors affecting the nature, the reproduction and the extent of inequality.
Once our theoretical framework is developed, our second contribution will consist in converting the above-mentioned framework into a methodological tool for application to a specific case study. The empirical application will serve to test the pertinence of a capability-based approach to well-being distribution, as well as to illustrate its added value with regard to standard approaches. This step is essential for exploring the practical foundation of this approach and understanding its practicability, meaning and usefulness in the design of public policies. Not only does a multidimensional assessment provide a more accurate picture of inequality, but it also reveals meaningful, in that it can serve as a functional instrument for understanding the nature of inequality and for defining policies that are geared toward reducing it. Indeed, for the purposes of evaluation and policy formulation, a capability-based approach can result in an important shift in the focus of development efforts, leading to a more effective way of improving the level of human well-being. In this sense, it will facilitate a better understanding of the distributional impacts of public policies in all sectors. Economic theory argues that public policies can influence inequality in three different ways. First, by affecting the redistribution of productive assets, including wealth. Secondly, through fiscal measures aimed at influencing the distribution of incomes. Thirdly, by directly affecting the levels of relative prices in individual markets (George K.D. and J. Shorey, 1978: 83-102). Whatever the approach adopted, the impact of these measures is likely to be immediate and limited to economic variables. There is, however, a certain level of circularity in this assertion. If we define inequality in term of wealth, income or consumption, public policies intended to reduce inequality will obviously focus on these attributes. On the contrary, in a broader perspective, the role and responsibility of public institutions in determining the nature and extent of inequality are profoundly renewed and expanded, in the sense that any policy choice may have an impact, either direct or indirect, on the distribution of well-being among individuals. As V. Donadoni points out, if the multidimensional nature of inequality is finally recognised and measured, public authorities may, for instance, start focusing their attention on the distribution of specific commodities in the population (e.g., health, housing, education, etc.), rather than on the distribution of current income per se (1995). A multidimensional measure seems therefore to be needed for the assessment of individual and social well-being, as well as for evaluating the impacts of economic and social policies.
Given the complexity of the issues discussed above, the objective of our research is two-fold. First, it aims at broadening the scope of distributional analysis, by proving the pertinence of a capability-based approach to conceptualising and assessing the distribution of well-being. In this perspective, we intend to discuss the concepts of well-being and inequality as interpreted and measured by mainstream economics, as well as the challenges and difficulties posed by the adoption of a multidimensional perspective. A critical analysis of these concepts and approaches is essential in order elaborate a new framework for analysing the nature of well-being and identifying the causal factors leading to its unequal distribution across individuals and social groups. Secondly, our research aims at testing the empirical feasibility of this framework, as well as its significance in the policy making. In this regard, we intend to operationalize our theoretical framework by developing a methodology for assessing the distribution of well-being in terms of capabilities. Such methodology will then be applied to a specific case study in order to examine the pertinence of our approach, as well as its added value with regard to standard measures. In this regard, the results collected from the application of our framework will be compared with the empirical evidence emerging from the use of standard approaches. This comparison promises to illustrate how a capability-based approach could be used for policy-oriented purposes. The empirical application will also serve to ascertain the characteristics and specificity of the context itself. Indeed, we can plausibly expect that a capability-based analysis will help understanding, with regard to the selected context, the nature of inequality by means of detecting the causal factors that can affect the distribution of well-being.
Compared to the breadth of our research objectives, the purpose of the present paper is by far more modest. Its aim is to present and discuss our conceptual framework, and to introduce the basic guidelines of a methodology intended to make this framework operational. In the light of these objectives, the paper is organised as follows. Paragraph 1.1 analyses the state of the current debate on well-being and inequality by briefly exposing how these concepts have evolved over the last three decades. Paragraph 1.2 discusses the concept of inequality as interpreted by mainstream economics and presents a short analysis of standard indicators in order to provide a critical perspective on the issues being discussed. Indeed, unless it is not proved that income or consumption cannot really capture all the other relevant dimensions of inequality, standard indicators will continue to be used. In particular, we will try to evaluate the consequences, both in terms of scientific and policy implications, of a monetary-approach to well-being distribution. A brief review of some reasons for giving a central role to a multidimensional assessment of inequality will follow. Paragraph 1.3 starts by analysing the multidimensional approach as conceptualised by different contributors. In line with the purposes of our research, the attention will be focused on the conceptual foundations of the multidimensional approach, rather than on the analytical methods and techniques that have been developed by different authors to construct multivariate indicators. A particular attention will be given to the analysis of the capability approach as developed by Sen. Paragraph 1.4 is devoted to the assessment of the operational content of a capability-based approach. In particular, we will review some critical points of the capability approach and examine some strategies for its operationalization. With these preliminaries, the basic lines of our approach are proposed and analysed. Concluding comments are given in the final part.
1.1 Well-being and inequality. The state of the current debate.
Human well-being has constantly been the subject of extensive theoretical and empirical research, as well as an issue of primary concern. Since the publication of Adam Smith’s classic paper “Enquiry on the Nature and Causes of the Wealth of Nations” (1776), social scientists have systematically inquired about the origin of well-being, while trying to find the variables leading to its expansion and distribution.
As Ravi Kanbur suggests (2003), during the last three decades, the analysis of well-being has seen two main phases. The first period that goes approximately from the 1970’s through to the mid 1980’s, has been characterised by a great ferment, both at the conceptual and empirical level. At the conceptual level, questions began to be raised about the validity of Gross Domestic Product (GDP) or per capita consumption as measures of well-being. A strong case was put forward for using a much broader concept of “quality of life” which would include, among other variables, nutritional status, education and life expectancy[4]. Increased attention started to be given to political factors influencing the allocation of resources across individuals. The concept of “basic needs” was launched, offering a new path to conceptualising and operationalising value judgments on distributional issues[5]. This ferment stimulated serious investigations into theoretical aspects of concepts and measurements (see Andrews F. M. and S.B. Withey, 1976; Hicks N. and P. Streeten, 1979: 567-580). Fundamental advancements were made on how to assess well-being and on how to incorporate distributional questions in applied economic research (see Chenery H. et al., 1974; Streeten et al., 1984).
This phase of conceptual advancement has been followed, from the mid-1980’s onwards, by a period of consolidation and policy debate (Kanbur R., 2003). After fifteen years of theoretical controversies, the conceptualisation of well-being made way for empirical debates about the extent of poverty and inequality, mainly conceived in terms of income and consumption. This was the result of the implementation of an operational approach based on the systematic application of indices on household data sets, which were increasingly made available for both rich and poor countries. The findings of these extended empirical studies served as basis for the adoption of different macro-economic policy packages intended to increase welfare and reduce poverty and inequalities, mainly through the implementation of fiscal and budgetary measures.