Dispersion in Borrowing Costs and its Aggregate Implications: Evidence from India (Job Market Paper)
Abstract: How does dispersion in borrowing costs reflect resource misallocation in developing economies? Using Indian data, I show that in response to identified positive foreign demand shocks, establishments with higher borrowing rates expand more and experience larger reductions in borrowing rates subsequently. To interpret these patterns, I develop an establishment-dynamic model with endogenous borrowing spreads that replicates the empirical relationship between borrowing rates and establishments' characteristics as well as the heterogeneous responses to demand shocks. The model features a Pareto jump process for idiosyncratic demand. Unlike the standard AR(1) Gaussian process, in this environment, many establishments remain under-capitalized and those with high borrowing costs are risky establishments with high growth potential. Dispersion in borrowing costs thus reflects a large extent of misallocation. Finally, I show that investment subsidy is effective in reducing misallocation, whereas directly subsidizing borrowing costs is not due to establishments' endogenous responses on increasing leverage.
Climate Change and Structural Transformation: A Global Perspectives (with Yueling Huang)
STEG PhD Research Grant
Abstract: We explore how emissions interact with multiple sectors (i.e.: agriculture, manufacturing, services) and study the implications of such interactions for structural transformation, aggregate productivity, and cross-country differences. We document three facts. First, agriculture and manufacturing have higher emissions relative to their sectoral shares than services. Second, the effect on agriculture is negative (positive) in hot (cold) countries, but insignificant in non-agriculture. Third, emissions per capita decrease with agriculture share, but increase with services share. We quantify a multi-sector model of structural transformation with three novel ingredients: energy production sectors, the carbon cycle, and two-way interaction between emissions and sectoral economic activities. Our model highlights that emissions hinder the reallocation from less productive agriculture to more productive manufacturing and services, thereby lowering aggregate productivity. The model with emissions yields approximately 18% higher cross-country productivity differences, compared to the standard model without emissions.
Transformation in the Retail Sector: Demand vs. Supply Factors