Statistical Inference and Lorenz Curves: An empirical example using Finnish Income Data

The main theme of this paper is to study statistical inference concerning empirical Lorenz curves and its generalisations. These curves are useful devices for examining welfare and inequality. Asymptotical sampling properties of the Lorenz curve are discussed. An improvement to a standard testing procedure of dominance relations is suggested. It is based on simulating the asymptotic distribution of the test statistic. The empirical part of this study is concerned with the Finnish distribution of disposable income in 1971, 1976, 1981, 1985 and 1990. It is found out that relative inequality has been highest on 1971, but the changes in relative inequality during period 1976-1990 have been almost negligible. Using a generalised Lorenz criterion it is also found out that welfare in income distribution has been rising in chronological order for the whole observation period.