on education in order to reduce poverty and to speed up development appears to be justified. Our empirical findings indicate that improving the quality of education rather than merely expanding access to education should play a crucial role in development strategies.
Several issues for future research are immediately apparent from our analysis. First, the direction of causality between inequality and human capital accumulation somehow remains an open question. Notwithstanding our results in Table 5.2 , more empirical research based on alternative instrumental variables is probably necessary to support the interpretation given in our chapter. Second, while our findings provide an encouraging impetus for the use of education policies as part of anti-poverty programmes, a rigorous theoretical framework supporting such a claim is still missing.
Third, and most importantly, highlighting the importance of education policy, as we do, should be accompanied by a more precise identification of effective education policies that would actually generate the expected effects. This is an important caveat because recent empirical evidence for OECD countries and for selected East Asian countries tends to suggest that additional schooling resources do not automatically guarantee improved schooling outcomes (Gundlach, Wößmann, and Gmelin 2001 ; Gundlach and Wößmann 2001). The international empirical evidence presented in Wößmann (2001) indeed reveals that schooling outcomes depend more on schooling institutions than on schooling resources. Hence, creating efficient schooling systems is probably more important for improving the stock of human capital than increasing schooling expenditure.