Human Capital, Heterogeneity and Estimated Degrees of Intergenerational Mobility

Abstract

We study the parental investment model of intergenerational mobility, where heterogeneity in 'innate' earnings ability and parents' altruism rate is explicitly modelled. We show that heterogeneity increases the difficulty of detecting the existence of borrowing constrained families. Conversely, the presence of heterogeneity means that economic and linear statistical models of inheritance generate similar intergenerational data on consumption and earnings. We also suggest that nay cross-country differences in intergenerational earnings mobility are more readily interpreted according to the heterogeneity of inherited ability, rather than optimal family responses to countiy-specific institutions for accumulating human capital. The relationship between a citizen's economic status and his family background is an important clue for understanding and evaluating a country's economic performance. Some (eg, Bowles (1972)) have suggested a close relationship between economic status and family background indicates some failure of labour markets or of institutions fostering investment in human capital. Others (eg, Altonji et al. (1992); Mulligan (1997, Table 4.1)) have offered an apparently contradictory suggestion, that 'efficient' allocations involve a very close relationship, on some measures, between the economic status of parents and children. Still others find family background effects to be the most offensive sources of inequality, regardless of whether those effects are consistent with 'efficiency'. Empirical studies of this relationship date back to the earliest days of statistical social science (eg, Galton's 1869 British study of the 'success' and 'eminence' of relatives), and many can be found outside the field of economics (see Goode's 1966 survey for a few examples), but the increasing availability of new data sets for exploring the relationship between earnings, income, wealth, and consumption of parents and, years later, of adult children has appropriately led to new and better econometric estimates of 'intergenerational mobility', or the intergenerational (auto)regression of various measures of economic status, and has even permitted some careful comparisons of those estimates across countries. However, fewer results are available on interpreting such estimates in terms of economic models of inheritance or of the labour market. Loury (1981), and Becker and Tomes (1986) are among the exceptions,' where they show how access to capital markets or other means of