Technology, Productivity, and International Trade
Paper Session
Saturday, Jan. 7, 2017 7:30 PM – 9:30 PM
Hyatt Regency Chicago, Plaza A
- Chair: Richard Pomfret, University of Adelaide
Capital Misallocation, Total Factor Productivity and the Size Distribution of Firms in South Africa
Abstract
Misallocation of labor and capital can greatly reduce aggregate productivity. In this study we examine the extent of such resource misallocation in the context of South Africa. While we find that productivity increased on average in the manufacturing sector between 2009 and 2014 we also find evidence the labor and capital are misallocated across firms within sectors. We estimate that if resources were allocated optimally across firms, productivity would have been around 50 per cent higher. We examine the firm specific factors related to misallocation focusing in particular on access to credit and the impact of government incentives that aim to facilitate a more efficient allocation of resources. We find that credit constraints, allowances for labor training and a depreciation allowance which favors the use of capital all potentially have distortionary effects that lead to a widening in the productivity distribution. We also find evidence that allowances for R&D expenditure are potentially productivity enhancing by leading to a more efficient allocation of capital. We find that the extent of misallocation is greatest among the smallest firms.Pareto Distributions in International Trade: Hard to Identify, Easy to Estimate
Abstract
We show that, in heterogeneous-firm international trade models, common forms of heterogeneity and uncertainty drive a (multiplicative random) wedge between the observable exports distribution and the latent distribution of firm productivity. Even if the latter is exactly Pareto distributed, this wedge, correlated with firm productivity, distorts the exports data, often making it look log-normally distributed. We show this distortion to be quantitatively relevant, meaning that empirical evidence of a log-normal exports distribution does not preclude an underlying Pareto distribution for productivity. Furthermore, this wedge renders common maximum-likelihood and quantile-based estimators misspecified, hence inconsistent. We provide general conditions in a broad class of international trade models under which, despite this misspecification issue, the tail of the exports distribution can be used to estimate the power exponent of the productivity distribution consistently---provided the sample size of left-truncated export data is large enough.Dissecting Gravity: From Customs Forms to Country-Level Trade Flows
Abstract
This paper studies the micro-foundations of the gravity equation of bilateral aggregate trade flows between any two countries. At the country level it states that the value of bilateral trade is inversely proportional to the distance between countries, typically measured between the capitals of those countries, and directly proportional to the gross national products of those countries. Using a unique and novel data set for Russian firms we study properties of firm-to-firm export flows from the most disaggregated level of individual customs forms to the firm-level, and then aggregate these flows to the country level. Our data set encodes all Russian exporters and foreign importers and the import and export transactions that are tied to specific suppliers in every country as well as buyers in every destination country. The data includes good's classification, quantity, weight in kilograms, value in both local currency and USD, exact locations of good origination and delivery, including ports of entry and exit, delivery method, and some contractual details regarding payment obligations. Our first result is on the role of exact distance in the gravity equation at every level instead of the capital-to-capital surrogate distance. We find that using exact distance at the country level increases by 32% the absolute value of the coefficient on the log of distance, which is both statistically and economically significant. Using either distance, however, yields the traditional negative relation between the trade value and distance at the country level. We then test the gravity relation at the customs form level, shipment level, and firm level and find a positive relation between the trade value and distance between trading partners at all these levels. We then study why the data aggregates at the country level to the classical gravity equation. We find that the most crucial property of the data leading to the correct aggregation is associated with the shipment level data. Specifically, we show that the average value per shipment should increase slower with distance than the decline in the average number of shipments with distance to yield the correct aggregation at the country level. We then demonstrate that as long as this condition is satisfied, the firm level aggregation may yield quite different results ranging from the traditional negative relation between value and distance, to no relation between them, to positive relation between them, thus making the firm-level evidence much less informative regarding the country-level gravity. We also provide support to theories emphasizing that higher quality and thus more expensive goods get shipped over longer distances.Trade, Input Sourcing and Misallocation
Abstract
Aggregate TFP losses resulting from the misallocation of production resources (across heterogeneous firms) due to domestic distortions have been variously analyzed and quantified in the economics literature. This paper develops a framework to study such misallocation in a context in which these firms undertake global sourcing decisions in the presence of additional distortions that affect supply prices of upstream firms (both domestic and international). Thus, this paper studies the role of input-output linkages in amplifying the costs of misallocation in the economy. Empirical evaluation, using highly disaggregated data on input sourcing from India and China, suggests economically large amplification effectsJEL Classifications
- F1 - Trade