The Real Effects of Banking the Poor: Evidence from Brazil
We study how financial development affects economic development and wage inequality. We use a large expansion of government-owned banks into Brazilian cities with low bank branch coverage and combine it with data on the universe of employees from 2000-2014. We find that higher financial development fosters firm growth, higher labor demand, and higher average wages, especially for cities initially in banking deserts. However, these gains are not shared equally. Instead, they increase with workers’ productivity, implying a substantial increase in wage inequality. The changes to inequality are concentrated in cities where the initial supply of skilled workers is low, indicating that talent scarcity can drive how financial development affects inequality. Our results are inconsistent with alternative explanations such as differential exposure to Brazil's economic boom, an overall increase in government lending, and other government or social welfare programs. These results motivate embedding skill heterogeneity into macro-finance development models in order to capture these distributional consequences.
We thank Victor Duarte and Chenzi Xu for support and numerous discussions. We also thank Shawn Cole (discussant), Tatyana Deryugina, Pascaline Dupas, Melanie Morten, Chad Jones, Ben Moll, Jacopo Ponticelli, Rebecca de Simone (discussant), Yongseok Shin, and seminar participants at Stanford Economics, Stanford GSB, the Finance-Organization-and-Markets Conference at Dartmouth, WashU Olin, Boston College, Queen Mary University, University of Pittsburgh, FGV-EPGE, Insper, University of Sao Paulo, HEC-Paris, the Brazilian Econometric Society Seminar, Cheung Kong University GSB, the Bank of Lithuania, UGA (Georgia), University of Arizona, and Imperial College for helpful comments. Filipe Correia, Thomas Gleizer Feibert and Peilin Yang provided excellent research assistance. Matray gratefully acknowledges financial support from the Griswold Center for Economic Policy Studies and the Louis A. Simpson Center for the Study of Macroeconomics. The views expressed herein are those of the authors and do not necessarily reflect the views of the National Bureau of Economic Research.