On the measurement of unfairness An application to high school attendance in Argentina

This paper presents a framework to measure unfairness in certain outcomes, like education attendance or basic health services consumption. The determinants of an individual outcome are divided into socially acceptable and unacceptable sources of differences in that outcome. To detect an unfair situation, comparisons are restricted to those individuals who share the same value of the vector of acceptable factors. The relevant argument to compare is the expectation of the outcome conditional on the vector of unacceptable variables. Unfairness is related to inequality in the distribution of those conditional expectations across individuals. An illustration of the framework is presented for the case of high school attendance in the Greater Buenos Aires area and other Argentine cities. Received: 25 August 2000/Accepted: 6 August 2001