Abstract

This paper investigates the determinants of income differentials in the framework of the human capital model along with offering solutions to two comman problems faced in the analysis of cross-sectional data, namely heteroscedasticity and sample selectivity bias. Four provinces included in the study are selected on the basis of development index created by State Planning Organization. Selected provinces, İstanbul, Adana, Malatya and Diyarbakır represent the most developed provinces in their respective regions. Two models are estimated for each province separately. Model 1 includes classical variables of the human capital model. Model 2 includes both the human capital variables, and others such as sex, marital status, age, education and occupation. Results indicate important differences among the provinces in terms of determinants of income differentiation. Although the returns to education in general vary between 6% and 8% for the provinces, educational returns are higher and different for women than men who have similar returns across the provinces. Education brings higher returns for women in less developed provinces compared to more developed ones. Married earn more income than unmarried. Entrepreneurs, businessmen and administrators in İstanbul and Diyarbakır, and those who work in commerce in Malatya and Adana earn highest income compared to service workers.

Key words: na

Article Language: EnglishTurkish