Risk aversion or simple entering? Considering relative risk aversion in inequality of educational attainment

Abstract

Relative risk aversion (RRA) hypothesis (Breen and Goldthorpe 1997) has been tested in various countries. However, a number of studies have rejected the hypothesis in Japan. In this paper, we propose a "simple entering model" as an alternative model of the RRA, and we test both models from macro and micro perspectives. In the simple entering model, it is assumed that individuals make their decisions about educational attainment based on the quota of students and their grades. We assume two types of the RRA model: Breen and Goldthorpe's model (the RRA model) and a "dual RRA model", in which individuals intend to minimize the risk of downward mobility also in academic status. In the macro perspective, we derive the college enrollment rate from respective model and test them using data JGSS-2000 and 2001. In the micro perspective, we derive propositions about individuals' decision-making. Subsequently, we test them with the data using logistic regression model. In order to identify the models, we apply the probit transformation for some variable. Our results show that models including RRA hypothesis do not fit empirical data, and the simple entering model has the most explanatory power; the RRA model is rejected at the macro perspective.