Interpreting the Coefficient of Schooling in the Human Capital Earnings Function

Abstract

In estimating the rate of return from schooling, the coefficient of the schooling variable is often interpreted as the rate of return from schooling. This may be the correct interpretation, but in principle - in many circumstances - is not. The human capital earnings function (HCEF) has become a fundamental tool in research on earnings, wages, and incomes in industrial and developing economies. It is accepted procedure in litigation about earnings, such as cases involving the value of lost earnings due to injury, death, or discrimination. It is also often used to make educational policy decisions based on estimates of the rate of return from schooling. The HCEF relates the natural logarithm of earnings to investments in human capital measured in time, such as years of schooling and years of post-school work experience. Among its desirable features: ° It is not an ad hoc specification; it is derived from an identity. So the coefficients of the equation have economic interpretations. ° It uses data efficiently. ° It is flexible, allowing for easy incorporation of variables appropriate for a particular study. ° And the coefficients of the HCEF are devoid of units, facilitating comparisons across space (such as countries) or across time periods (such as decades). In estimating the rate of return from schooling, the coefficient of the schooling variable is often interpreted as the rate of return from schooling. This may be the correct interpretation but Chiswick shows that in principle - in many circumstances - is not. He also discusses the effects on the coefficient of schooling of the treatment of the labor supply (weeks worked and hours worked per week) and other measures of labor market outcomes, such as occupational status. This paper - a product of the Human Development Department - is part of a larger effort in the department to understand the economics of education.