Zusammenfassung:
The paper contrasts the pattern of returns to human capital in different economic sectors. As job mobility, especially across sectors, is limited, it is argued that coefficients of experience in earnings regressions may capture or be interpreted as the growth rate – net of depreciation – of earnings ability propitiated by schooling when years of education are also included in the right hand-side of the equation. As a consequence, under long-term contracts, labor market equilibrium is compatible with different “gross” rates of return to schooling, provided initial earnings levels allow for the same accumulated present value. That implies a special relation between the intercept and experience coefficient of earnings regressions performed for different sectors. Additionally, implications of (log-stable) nonstationary environments for rate of return inference from log-earnings regressions – appropriate for pooled (or panel) estimation and nominal earnings information - are also investigated. Then, the trend coefficient measures the (steady-state) nominal productivity growth; the experience coefficients approximate individuals’ earnings profiles growth rates net of the human capital depreciation rate; schooling’s, the nominal rate of return in the economy net of the nominal productivity growth rate. Tests of the hypothesises are provided, along with the inspection of the determinants – including financial ratios and productive organization indicators, calculated from aggregate balance sheet information - of the observed differences across industries. A study of the estimated variances of rate of return estimates was also conducted, as an attempt to capture features of financial risk in human capital investment.