Three essays on firm productivity in industrial organization and international trade

This dissertation focuses on the measurement of productivity at the firm level and the implication of firm-level heterogeneous productivity in industrial organization and international trade. Chapter 1 studies the interaction between firm productivity and importing decision and quantify the static and dynamic gains from importing intermediate inputs. Chapter 2 develops a new way to estimate the production function consistently when some of the input prices are missing. Chapter 1 and Chapter 2 are based on the Hicks-Neutral technology assumption. Chapter 3 studies the biased technology change over time and biased technology dispersion across firms.;Chapter 1: Static and Dynamic Gains from Importing Intermediate Inputs: Theory and Evidence: This paper constructs a dynamic structural model to characterize firms' decisions to buy imported inputs or rely exclusively on domestically-supplied inputs and quantifies their effects on firm value and productivity. The model provides a unified framework to analyze the determinants of firms' import decisions and to empirically decompose the gains from importing into a static effect and a dynamic effect. Empirical results using Colombian plant-level data show that more productive plants tend to import intermediate inputs and that the total gain from importing is large. The decomposition shows importing is important mostly because it dynamically generates higher future productivity growth.;Chapter 2: Production Function Estimation with Unobserved Input Price Dispersion: We propose a method to consistently estimate production functions when intermediate inputs are not observed in the presence of input price dispersion. The traditional approach to dealing with unobserved input quantities---using deflated expenditure as a proxy---requires strong assumptions for consistency. In particular, we show that the traditional approach tends to underestimate the elasticity of substitution and bias estimates of the distribution parameters. Our approach applies to a general class of production functions with a mild identification restriction. As a demonstration, we apply our approach to the CES production function. A Monte Carlo experiment illustrates that the omitted price bias is significant in the traditional approach, while our method consistently recovers the production function parameters. We apply our method to a firm-level data set from Colombian manufacturing industries. The empirical results are consistent with the predictions that the use of expenditure as a proxy for quantities biases the elasticity of substitution downward. Moreover, using our preferred method, we provide evidence of significant input price dispersion and even wider productivity dispersion than is estimated using traditional methods.;Chapter 3: Biased Technology and Contribution of Technological Change to Economic Growth: Firm-Level Evidence: The increasing mean wage-interest ratio and decreasing mean capital-labor ratio observed in some Chinese manufacturing industries suggest that technological change is factor-biased. In order to study the nature of technological change and its contribution to economic growth, this paper builds and estimates a structural model of firms' production decisions with biased technological change. This model allows me to identify and estimate the firm-time-specific factor-biased technology using micro data. The basic idea of the estimation is that the choice of inputs contains information about the unobserved productivities; therefore we can invert the inputs demand function to recover the unobserved productivities. I estimate the model from a firm-level data set of four Chinese Manufacturing industries. The empirical results provide firm-level evidence of biased technological change over time and biased technological dispersion across firms. The estimation results show that technological change contributes to the growth of gross output by 1.81%-3.10% annually and value added by 12.67%-21.16%, which is higher than the combined contribution of capital and labor. Capital efficiency grows much faster than labor efficiency in China, and the contribution of technological change to economic growth is mainly due to the change of capital efficiency. The results also show that large firms have a higher capital-labor efficiency ratio and that biased technological dispersion explains a large part of the dispersion of capital-labor ratio across firms.