Centre for Development Economics
Department of Economics, Delhi School of Economics


Self-Confidence and Poverty Trap


Dyotona Das Gupta

CDE, Delhi School of Economics

Thursday, 29th August 2019 at 3:00 PM

Venue : Seminar Room (First Floor)
Department of Economics, Delhi School of Economics

All are cordially invited

Individuals often discount observations obtained from socially distant groups. This paper shows how this discounting can create over and underconfidence which in turn create poverty trap in the economy. In particular, we construct an overlapping generation model where an adult (parent) decides whether to invest in the education of her child or not. While costly education is necessary but not sufficient for getting a job in the skilled sector, the income of a skilled worker is always higher than that of an unskilled worker. Based on their education and job type, parents belong to different groups. Given that identity, parents discount observations obtained from groups which are distant while forming beliefs about return to investment in education. A parent invests in education only if that provides her higher expected income, given her beliefs.
We find that affluent (skilled) parents overestimate the probability with which a child from their community gets a skilled job whereas other parents underestimate that. These differences in beliefs yield four key results. First, we find that once uneducated, a dynasty never becomes educated again. Further, since education is necessary for becoming a skilled worker, no member of that family becomes skilled again. Second, when parents are patient the mass of uneducated workers increases over time, whereas when parents are impatient, they never invest in education. Third, in contrast to the benchmark case where everyone knows the return to education, the long run skill distribution has fewer skilled people in the economy — in the steady state, fewer rich people in the economy earn higher income. And finally, in comparison to the benchmark case, the economy converges to the steady state faster.

back to seminars…