Estimation of the Treatment Effect's Dynamics of Higher Education on the Individual's Health by Age Cohorts: Multivariate Probit Model and Matching

The question of the existence of a causal relationship between education and health is still the subject of scientific debate. In the conditions of a non-randomized experiment, the problem of non-random selection (self-selection) arises. The simplest regression methods do not take into account the existence of this problem, so there is a need to use more complex methods to solve the problem. The issue of determining the dynamics of the impact of education on the health of individuals in terms of age cohorts is also relevant. In this study with the use of multivariate probit model and matching methods the treatment effect of higher education on various health indicators in the context of age cohorts is estimated. The presence of multidirectional treatment effects of education is revealed, however, the results are not statistically significant for all cohorts and health indicators. Based on the results obtained, it is concluded that it is necessary to study the treatment effect of education separately for each health indicator, separately for men and women and certain age groups of the population. The results of the work can be useful both from a methodological and applied point of view for future researchers in the field of the relationship between education and health.