21 there is strong empirical support for a positive

Info icon This preview shows pages 12–14. Sign up to view the full content.

21. There is strong empirical support for a positive effect of education on adult health, even after controlling for initial health using panel data (van Doorslaer, 1987; Wagstaff, 1993). This leaves the issue of whether child labour is indeed a substitute for education. In a simple model in which a child faces the option of either full-time education or full-time work, increased work activity is obviously at a substantial cost to education. However, the child may be able to divide its time more flexibly between work, school and play. In which case, the issue is whether marginal increases in work are at the expense of schooling, play or both. Where schooling choices are severely constrained by family resources, there is the possibility that child labour even has a positive effect on education through providing the resources necessary to pay for schooling. 22. The existence and the degree of any trade-off between child labour and education is an empirical question on which the evidence is mixed. There is growing evidence of a substantial proportion of kids in developing countries combining school and work (Patrinos and Psacharopoulos, 1995; Akabayashi and Psacharopoulos, 1999; Anker, 2000; Cigno and Rosati, 2001). There is, however, a great deal of heterogeneity across countries in the extent to which child work activity and schooling overlap (Anker, 2000; Heady, 2000). In general, the combination of work and schooling is most common in Latin America and least common in Asia, with some African countries lying in the middle. Testing the proposition that child work “crowds-out” schooling is complicated by the fact that child labour and schooling decisions are taken simultaneously and so are potentially influenced by common unobservable factors, which bias the estimated relationship. In order to circumvent this endogeneity problem, the existence of a trade-off has been tested indirectly by examining whether factors that encourage child work activity also tend to discourage school attendance. The weight of the evidence is in support of a trade-off. Analyses of data from Bolivia and Venezuela (Psacharopoulos, 1997), Cote d’Ivoire (Grootaert and Patrinos, 1998), India (Rosenzweig and Evenson, 1977; Cigno and Rosati, 2001) and Zambia (Nielsen, 1998) all support the crowding-out hypothesis. On the other hand, no support is given from another analysis of data from India (Skoufias, 1994) and in data from Peru (Patrinos and Psacharopoulos, 1997). 10 23. A limitation of all the research quoted above is that it concentrates on the trade- off between work activity and time spent in school whereas the central concern is 10 Ravallion and Wodon (2000) take a different approach to circumventing the endogeneity problem, examining the impact of a subsidy for school attendance in Bangladesh on time spent in work and school, instrumenting participation in the subsidy programme. The subsidy is successful is raising school attendance, with most of the extra study time coming at the expense of leisure, rather than work.
Image of page 12

Info icon This preview has intentionally blurred sections. Sign up to view the full version.