« Back to Results

Social Identity, Mobility, and Networks

Paper Session

Sunday, Jan. 3, 2021 3:45 PM - 5:45 PM (EST)

Hosted By: Econometric Society
  • Chair: Emily Breza, Harvard University

The Strength of Weak Ties: Education Externalities and Intergenerational Mobility in Africa

Leonard Wantchekon
,
Princeton University

Abstract

We use historical micro-level data from the first regional schools in colonial Benin to estimate the effect of education on social mobility over three generations. Since school location and student cohorts were selected quasi-randomly, the effect of education can be estimated by comparing the treated to the untreated living in the same village as well as those from villages with no school (Wantchekon et al. [2015]). We find positive treatment effects of education on social mobility across three generations. Surprisingly, the effect is strongest for descendants of grandparents who were exposed to education only through their social networks (the untreated living in villages with a school). We interpret this result as evidence of “the strength of weak ties” (Grannoveter [1977]). Finally, exploring the underlying mechanism of our results, we find that mobility from the first to the second generation is driven by parental aspiration, which is sustained by the risk attitudes and mobility of the third generation."

Does Identity Affect Labor Supply?

Suanna Oh
,
Paris School of Economics

Abstract

Does identity—one's concept of self—influence economic behavior in the labor market? I investigate this question in rural India, focusing on the effect of caste identity on labor supply. In a field experiment, casual laborers belonging to different castes choose whether to take up various real job offers. All offers involve working on a default manufacturing task and an additional task. The additional task changes across offers, is performed in private, and differs in its association with specific castes. Workers' average take-up rate of offers is 23 percentage points lower if offers involve working on tasks that are associated with castes other than their own. This gap increases to 47 pp if the castes associated with the relevant offers rank lower than workers' own in the caste hierarchy. Responses to job offers are invariant to whether or not workers' choices are publicized, suggesting that the role of identity itself—rather than social image—is paramount. Using a supplementary experiment, I show that 43% of workers refuse to spend ten minutes working on tasks associated with other castes, even when offered ten times their daily wage. This paper's findings indicate that identity may be an important constraint on labor supply, contributing to misallocation of talent in the economy.

Affirmative Action, Attitudes and Social Networks: Evidence from Caste-Based Reservation in India

Emily Breza
,
Harvard University
Arun Chandrasekhar
,
Stanford University
M.R. Sharan
,
Harvard University

Abstract

Abstract: Affirmative action is an oft-used tool in order to ensure representation and equal opportunities in a variety of contexts such as education and employment. In India, quotas (reservations) are also used in local governance. We study whether and how reservation affects social and economic network structure, beliefs about members of other communities such as trust and intelligence, and norms over cross-groups social interactions. To do this, we make use of a discontinuity design in the algorithm by which rural communities are selected to have reservations for the office of president of the local government. We document changes in network structure coupled with a deterioration of beliefs toward the historically disadvantaged group and an increase in untouchability practices in the village. We also show that these changes may have downstream effects on social learning using an information intervention around COVID-19. Knowledge of symptoms, precautions, and health seeking procedures vary considerably by historical governmental reservation.

A Division of Laborers: Identity and Efficiency in India

Guilhem Cassan
,
University of Namur
Daniel Keniston
,
Louisiana State University
Tatjana Kleineberg
,
World Bank

Abstract

Workers’ social identity affects their choice of occupation, and therefore the structure and prosperity of the aggregate economy. We study this phenomenon in a setting where work and identity are particularly intertwined: the Indian caste system. Using a new dataset that combines information on caste, occupation, wages, and historical evidence of subcastes’ traditional occupations, we show that caste members are still greatly overrepresented in their traditional occupations. To quantify the effects of caste-level distortions on aggregate and distributional outcomes, we develop a general equilibrium Roy model of occupational choice. We structurally estimate the model and evaluate counterfactuals in which we remove castes’ ties to their traditional occupations: both through their direct preferences, and also via their parental occupations and social networks. We find that the share of workers employed in their traditional occupation decreases substantially. However, effects on aggregate output and productivity are very small– and in some counterfactuals even negative–because gains from a more efficient selection based on individuals’ comparative advantage are offset by productivity losses due to weaker caste networks and reduced learning across generations. Our findings emphasize the importance of caste identity in coordinating workers into occupational networks, thus enabling productivity spillovers.
JEL Classifications
  • J6 - Mobility, Unemployment, Vacancies, and Immigrant Workers