We study the political determination of the proportion of students attending university when access to higher education is rationed by admission tests. Parents differ in income and in the ability of their unique child. They vote over the minimum ability level required to attend public universities, which are tuition-free and financed by proportional income taxation. University graduates become high skilled, while the other children attend vocational school and become low skilled. Even though individual preferences are neither single-peaked nor single-crossing, we obtain a unique majority voting equilibrium, which can be either classical (with 50% of the population attending university) or “ends-against-the-middle”, with less than 50% attending university (and parents of low and high ability children favoring a smaller university system). The majority chosen university size is smaller than the Pareto efficient level in an ends-against-the-middle equilibrium. Higher income inequality decreases the majority chosen size of the university. A larger positive correlation between parents’ income and child’s ability leads to a larger university populated by a larger fraction of rich students, in line with the so-called participation gap. Our results are robust to the introduction of private schooling alternatives, financed with fees.
majority voting ends-against-the-middle non single-peaked preferences non single-crossing preferences higher education participation gap income ability correlation size of university