The effect of public policies on the demand for higher education

A binomial logit model is fitted to the college attendance behavior of 27,046 male high-school juniors in 1960, divided into 20 subgroups defined by student ability and family income. Tuition, high admissions standards, travel costs, and room and board costs all have significant negative effects on attendance. The highest elasticities of demand are found to occur in the low-income strata and lower-middle ability quartile, suggesting that an efficient subsidy program should focus on these groups. Coefficients of forgone earnings and measures of the local payoff to college attendance are small but generally have the expected sign.