A Simultaneous Model Approach to the Determinants of Male Earnings Differentials in Turkey for 1968

Abstract

Income distribution is a topical issue in all countries regardless of their economic and social systems or levels of development. In recent years, relationship between income distribution and process of development has come under increasing scrutiny. Parallel to this development, increasing attention has been given by economists and statisticians to empirical analysis of distribution of income. The purpose of this paper is to examine causes of annual earnings differentials among male household heads outside agriculture in Turkey for 1968.1,2 The primary data source is a survey conducted by Department of Population Studies of Hacettepe University of Turkey in 1968.3 In most of studies that deal with causes of inequality, key explanatory variables have been education and experience (Becker and Chiswick, 1966; Mincer, 1974). Others have introduced geographic location, ability, occupation, and physical capital factors in addition to human capital factors (Fishlow, 1974; Griliches and Mason, 1972). As Richard D. Morgenstern notes, however, the notion that socio-economic background is important in determining an individual's income has been played down in literature concurrent with rise in popularity of human capital explanation of earnings differences.4 Only recently, socioeconomic background factors and intergenerational transfers of inequality have been gaining their deserved importance in income distribution literature.5 The second shortcoming of single equation models of income distribution has been their lack of attention to interactions between explanatory variables. Ribich and Murphy point out that while single equation models might suffice as a reasohable equation for predicting earnings, relative importance of independent variables would be obscured by statistical tests that focused on this equation alone.6 In this paper, distribution of earnings among male heads of households is examined within frameworks of recursive and simultaneous equation models. The models developed have three endogenous variables, namely, education, occupation and earnings. The exogenous variables include availability and quality of schooling, experience, geographical location and socio-economic background factors. The causal model of income determination employed in this study is given in figure 1.