The Impact of Financial Structure on Poverty Reduction : A Study using Household Data from Peru , 1985 - 2000

This paper estimates/calibrates a growth model with endogenous financial structure using household data from Peru. The idea is to feature the role of financial infrastructure in generating GDP growth and increasing inequality as observed in Peru during the 1990’s through the lens of an economic model of households’ saving/investment decisions. The typical approach to estimate the contribution of financial sector expansion to GDP growth is to run simple cross-country regression based on stationary relationship among variables. This paper takes an alternative approach that has two main contributions. First, actual data is thought of as a transitional phenomenon – whole history is considered as one sample – which allows studying whether the simulated economy with specific shocks can generate the same time paths of actual aggregate data. The analysis provides implications for the evolution of income inequality and its interaction with financial development. Second, the paper provides a benchmark to investigate whether observed levels and rates of financial expansion have been adequate. Provided that government and private sector resources are limited, quick expansion may not be possible or desirable. The approach in the paper explicitly solves the household’s saving/investment problem and, by employing historical data, offers a benchmark for financial expansion consistent with observed time path of GDP growth. In the artificial, simplified model, households accumulate savings to pay a fixed cost to participate in the formal financial sector and, thus, to benefit from higher return and insurance against idiosyncratic shocks. Investment technology parameters are estimated from household data and the aggregate paths of the economy are calibrated to actual GDP growth. When I compare “actual” and “model-predicted” time paths I find that the benchmark economy largely overestimates actual financial participation in Peru during the period 1985-2000, 19% vs. 7%. To a great extent, this over prediction seems explained by the significant reduction in the number of households saving in formal financial institutions observed in the household data (26% to 10%). Observed increasing inequality is also predicted in the calibrated economy, but, in the model, the large welfare costs from “restricted” financial participation affect more middle-wealth groups rather than the poor. The findings in the paper suggest that there are large welfare gains from expanding (or recovering past) formal financial participation levels. * An Application to the Case of Peru of “Financial Deepening, Inequality and Growth: A Model Driven Quantitative Evaluation” by Robert M. Townsend and Kenichi Ueda (2001). † The bulk of the analysis in my paper was carried out in the last year of my graduate studies at the University of Chicago, with the guidance and collaboration of Professor Robert M. Townsend. I am very much in debt to Natalia Ramondo and Kenichi Ueda, former colleagues at Chicago, for their patient help in running the programming codes.