Evidence from industrialized countries

185 M. Binder Economics of Education Review 18 1999 183–199 important. 3 A positive relationship between an economic downturn and schooling should not then be interpreted as a pure price effect, since it also may contain a positive response to the rate of return to schooling, especially in the changing economic environment of Mexico in the 1980s.

3. Evidence from industrialized countries

Ultimately, the effect of a recession on schooling decisions is an empirical question, which has not, appar- ently, been applied to developing countries. 4 Cross-sec- tional and time-series analyses for industrialized coun- tries generally find a positive income effect and negative price effects, where price is measured by both tuition for direct costs and the unemployment rate for fore- gone costs. 5 The elasticities for income are usually much larger than those for unemployment. This gen- eralization appears to hold for school-leaving in Great Britain Whitfield and Wilson, 1991; Rice, 1987; Pissar- ides, 1981, 6 as well as for high school and college enrollments and college entry Betts and McFarland, 1995; Kane, 1994; Manski and Wise, 1983; Mattila, 1982; Lehr and Newton, 1978 in the United States. In addition, the following patterns emerge: 1 high school and community college enrollments appear to be more responsive to economic variables than four-year college enrollments Mattila, 1982; Betts and McFarland, 1995, 2 blacks tend to have higher elasticities than whites Kane, 1994, and 3 the college enrollment rates of younger men are more elastic than those of older men Mattila, 1982. In each of these cases, the more respon- sive group likely has a higher proportion of marginal schooling decision makers. For blacks and community college students, high elasticities may arise from low income. Low income families are more likely to face binding liquidity constraints and may have a low valu- ation of consumption component of schooling relative to the investment component. 7 Similarly, the potential high school population will have a lower mean income than 3 Author’s survey conducted in Guadalajara, Arandas and Tijuana in 1993. 4 Despite an extensive search, I could find no such studies in either the English or Spanish language literatures. 5 Many studies control for returns to schooling, so that the unemployment rate can be unambiguously interpreted as a price effect. 6 One notable exception is Micklewright et al. 1990, who find a positive relationship between unemployment and school-leaving. 7 Manski and Wise 1983 find that parent schooling is more influential than parent income in the decision to enroll in col- lege. This may reflect a relatively larger consumption compo- nent in the decision to pursue schooling. the college-bound population Manski and Wise, 1983. The younger age category in 3 also contains more mar- ginal decision makers, since the decision to attend col- lege is typically made in the last year of high school. Men who begin to work rather than enroll in college face higher opportunity costs of college as they gain experi- ence in the labor market.

4. Overview of the Mexican schooling system