The Macroeconomics of Micro Distortions in Developing Countries.

摘要

This dissertation consists of three main chapters.;In the chapter titled "Trade Liberalization with Idiosyncratic Distortions: Theory and Evidence from India," we study the effects of a trade liberalization on trade patterns, the skill premium, and welfare in a multi-sector economy with heterogeneous firms. Firms are potentially subject to idiosyncratic distortions, i.e. firm-specific distortions that constrain large/highly productive firms while favoring small ones. We argue that this type of distortions reduces the perceived degree of productivity heterogeneity across firms. We formally show that within-sector heterogeneity is a determinant of comparative advantage, and thus contributes to shaping the pattern of specialization and relative factor prices. We then explore a quantitative application using India's 1991 trade reform episode. India's economy was opened up for trade without completely removing the numerous micro distortions in the manufacturing sector. A calibrated model with idiosyncratic distortions inferred from firm-level data is successful in reproducing several key facts of India's development path, namely reversal of endowment-based comparative advantage trade patterns, increased inequality after trade liberalization, and a structural change process that is uniquely driven by services. Finally, our quantitative exercise shows that the welfare costs of idiosyncratic distortions are magnified by trade openness.;Hsieh and Klenow (2009) present evidence of substantial mis-allocation of labor and capital in India's manufacturing sector, consequently leading to large aggregate TFP losses. The chapter "Labor Market Policies and Mis-Allocation in India" proposes and tests a theory of that mis-allocation based on labor market distortions. We start by documenting India's restrictive labor market policies and its disproportionate emphasis on skill-intensive manufacturing despite being relatively abundant in unskilled labor. We then develop a two-sector general equilibrium model of firm dynamics in the spirit of Hopenhayn and Rogerson (1993), in which distortions to the efficient allocation of labor generated by the labor laws interact with sector-specific volatility to produce the observed shifts in sectoral composition. A quantitative exercise shows that labor market regulations result in substantial mis-allocation, as measured by dispersion of the distribution of revenue-based TFP, and consequently a loss of aggregate manufacturing TFP on the order of five to eight percent, and potentially higher.;The final chapter is titled "The Implications of Non-Convex Technology for Mis-Allocation Inference." Recent literature aiming to infer the extent of idiosyncratic distortions from micro data often examine data through the lens of a standard model in which plants have Cobb-Douglas production technology. This class of models predict that the undistorted within-sector distribution of labor productivity is degenerate---a counterfactual prediction given the observed large within-sector dispersion in labor productivity in the United States, which is a relatively undistorted economy. We develop a simple model that incorporates non-convex technology in the form of a fixed labor cost and use it to infer idiosyncratic distortions from India's plant-level data as in Hsieh and Klenow (2009). We allow for sector heterogeneity in fixed cost and show how to plausibly derive the distribution of fixed costs from observed average establishment size. Then we show that this minor modification to the production function changes our inference of the extent of mis-allocation from micro data under certain conditions. The model also uncovers interesting cross-sector pattern of mis-allocation, with larger scale industries suffering from more severe idiosyncratic distortions.