Non-Technical Summary:
The measurement of inequality of opportunity is a growing topic in economics. In recent years many theoretical papers have been published and the number of empirical contributions to this literature has exploded in the last two decades. The increasing quality and availability of survey data will likely encourage further empirical research. Not surprisingly, journal articles and volume chapters have recently reviewed the main approaches to measuring inequality of opportunity and discussed the existing evidence. The aim of this paper is complementary to those contributions: it discusses the many practical issues that typically arise when measuring inequality of opportunity with survey data.
The paper originates from lecture notes for the workshop “Measuring inequality of opportunity” held at ISSR, The University of Queensland in February 2016. The workshop aimed at covering the main issues of the applied literature on inequality of opportunity measurement. The material is presented in an attempt to attract the interest of a multidisciplinary audience of social scientists and therefore does not get into the details of some more technical aspects. For example, the discussion of inference (generally base on bootstrap) is absent. The comprehensive list of references provided may accommodate the needs of most demanding readers.
The paper is organised in five sections. The first section introduces the ideal of equality of opportunity as it developed in the economic literature in the last decades.
The second section discusses two approaches to measure inequality of opportunity. The first proposed by John Roemer and the second introduced by Marc Fleurbaey and Francois Maniquet. These two approaches quantify inequality of opportunity following a similar three-step method: i) define the properties that a distribution of valuable outcome –such as income or health– must have to satisfy the principle of equality of opportunity; ii) obtain a counterfactual distribution of which reflects the violations of those properties in the actual distribution; iii) measure inequality in the counterfactual distribution of unfair inequality.
The third section discusses alternative methods to measure inequality of opportunity. The literature has proposed two main approaches –a parametric and a non-parametric approach– both methods have advantages and shortcomings that are discussed in the section.
The fourth section suggests two inequality measures that can be used to measure inequality in the estimated counterfactual distribution and discusses the effects of different choices.
The last section is devoted to two more advanced topics. The first is a method to decompose total inequality of opportunity by sources, that is it introduce a method to identify the share of inequality of opportunity associated with a specific characteristics (gender, race, socioeconomic origin for example). The second is an index of economic development sensitive to inequality of opportunity –the Human Opportunity Index– which has been proposed and popularized by the World Bank.