The Role of Credit Constraints in Educational Choices: Evidence from NCDS and BCS70. CEE DP 48

In this analysis we seek to shed light on the extent to which credit constraints may affect individuals' choices to stay in full-time education past the age of 16 and to complete higher education (HE) qualifications in the United Kingdom, and on how this has varied between individuals born in 1958 and in 1970. The observed correlation between family income and educational outcomes can be interpreted as arising from two quite distinct sources: (1) short-run credit constraints, whereby a limited access to credit markets means that the costs of funds are higher for children of low-income families; (2) long-run family background and environmental effects, which produce both cognitive and non-cognitive ability and also mould children's expectations and tastes for education, all of which crucially affects schooling choices and outcomes. The aim of the analysis is to separate out the pure effect on schooling of short-term income constraints from longer term family income and background influences. We apply to the UK the procedure of Heckman and Carneiro (2003), aimed at estimating the share of individuals who are not staying on or who are not attaining an HE qualification because of short-term "credit constraints". The share of the population being "credit constrained" in their educational choices is defined as any residual gap that remains in the educational participation rates of individuals whose families are in the lower three quartiles of the parental income distribution compared to individuals in the top quartile (who are, by definition, not credit constrained), once having controlled for a number of observed measures of early family and environmental influences. Note that to the extent to which we were unable to proxy all those early and background long-run factors related to growing up in a family in the top income quartile that also affect educational outcomes, our estimates represent an "upper bound" on the share of individuals truly facing short-run credit constraints. Especially in the light of this caveat, overall our results do not seem to point to particularly large fragments of the population being credit constrained in their educational choices, although we found evidence that the importance of short-term credit constraints on the staying-on decision has increased for the younger cohort. As to the decision of whether to stay in full-time education past the age of 16, we found no evidence of credit constraints for individuals born in 1958. By the time the 1970 cohort was making this decision some 12 years later in 1986, however, this situation had changed, particularly for individuals in the top and bottom ability tertiles. In particular, around a 7 percentage point gap in staying-on rates remains between individuals coming from families in the top income quartile and those from the bottom three quartiles. As to the importance of credit constraints in terms of HE attainment decisions, we find that for males, such constraints seem to have surfaced between the two cohorts, though the share of the younger cohort being affected remains fairly minor (2-3%). By contrast, females of both cohorts seem to have been affected by credit constraints in roughly equal measure (3-6%). Subject to the proviso that we have no way to gauge if and how these patterns have evolved from the mid 1980s to today, our findings seem to suggest that policies aimed at reducing the possible impact of short-run credit constraints on education decisions should target individuals at the age of 16 (or possibly earlier) when they are making decisions about whether or not to continue in full-time education (e.g. like the Education Maintenance Allowances programme), rather than at 18 when individuals are making Higher Education decisions. Appended are: (1) BCS Credit Constraint Regression Results; and (2) NCDS: Proportion staying on and attaining HE by parental income quartiles. (Contains 1 box, 8 figures, 6 tables and 11 footnotes.)