What Do Twins Studies Tell Us about the Economic Returns to Education? A Comparison of Australian and US Findings

An analysis of the economic returns to education is undertaken using a rich and particularly large sample of twins (1170 pairs). The findings are placed in the context of certain existing American work, the study by Behrman et al (1977) and the recent innovative analysis by Ashenfelter and Krueger (1993). Following Ashenfelter and Krueger (1993) both fixed effects and selection effects models are estimated and correction is made for measurement error in self reported schooling levels. The basic findings are rather similar to those of Behrman et al (1977). However, they contrast sharply with those of Ashenfelter and Krueger (1993) who report that the conventional estimate of the rate of return almost doubles to 13 - 18 per cent when correction is made for measurement error and that there is no role for family effects in this relationship. The main finding is that the rate of return of schooling is in the range of 4 - 8 per cent and that there is a role for at least the genetic component of family effects in the relationship between schooling and income.