Selected Research Papers
1. "Trade Cost and Export Diversification: Evidence from Chinese Firms" (Job market paper, joint with Yifan Li)
Abstract: We investigate the relationship between the number of varieties a firm decides to export (its export scope) and the characteristics of the destination country. Using Chinese firm-level customs data for 2001 and 2006, we document that Chinese exporters adjust their export scope to different characteristics of destination countries. We show that firms export fewer varieties to countries that display higher exchange rate volatility, that are farther away from China, or that impose higher import-tariff rate. Also, we find that the response to the tariff reduction process due to China’s entry into the WTO in 2001 is heterogeneous across firms: high productivity firms (the total factor productivity is measured through the Olley-Pakes method) expanded their export scope, while low productivity firms reduced it. With this evidence at hand, we develop a flexible and tractable theoretical model to rationalize our empirical findings. Our framework considers heterogeneous firms’ optimization decisions involving both production and export varieties and their interplay with the exchange rate volatility, the distance to the destination country, and the tariff rate. Our model predicts that the export scope decreases in the level of exchange rate volatility, distance, and tariff rate of a destination country: firms can reduce the export scope if the destination countries suffer negative demand shocks, but cannot expand the export scope if positive shocks occur, due to insufficient pre-investment in production capacity. Also, our model predicts that high-productivity firms have an advantage in producing higher quality products, and in response to a tariff reduction the demand for high quality products increases more than that for low quality products: thus, high productivity firms react by expanding their export scope, but low productivity firms may reduce their export scope due to the increase in market competition.
Abstract: We investigate the relationship between the number of varieties a firm decides to export (its export scope) and the characteristics of the destination country. Using Chinese firm-level customs data for 2001 and 2006, we document that Chinese exporters adjust their export scope to different characteristics of destination countries. We show that firms export fewer varieties to countries that display higher exchange rate volatility, that are farther away from China, or that impose higher import-tariff rate. Also, we find that the response to the tariff reduction process due to China’s entry into the WTO in 2001 is heterogeneous across firms: high productivity firms (the total factor productivity is measured through the Olley-Pakes method) expanded their export scope, while low productivity firms reduced it. With this evidence at hand, we develop a flexible and tractable theoretical model to rationalize our empirical findings. Our framework considers heterogeneous firms’ optimization decisions involving both production and export varieties and their interplay with the exchange rate volatility, the distance to the destination country, and the tariff rate. Our model predicts that the export scope decreases in the level of exchange rate volatility, distance, and tariff rate of a destination country: firms can reduce the export scope if the destination countries suffer negative demand shocks, but cannot expand the export scope if positive shocks occur, due to insufficient pre-investment in production capacity. Also, our model predicts that high-productivity firms have an advantage in producing higher quality products, and in response to a tariff reduction the demand for high quality products increases more than that for low quality products: thus, high productivity firms react by expanding their export scope, but low productivity firms may reduce their export scope due to the increase in market competition.
2. "Multiple-Quality Oligopoly and the Role of Market Size" (Joint with Ngo Van Long)
Abstract: We model an oligopoly where firms can choose the quality level of their products by incurring set-up costs that generally depend on quality level. If the set-up cost is independent of product quality, firms may choose to supply both types of quality. We focus on the long run equilibrium where free entry and exit ensure that the profit for each type of firm is zero. Using this framework, we study the implications of an increase in the market size. We show that for the existence of an equilibrium where some firms specialize in the low quality product it is necessary that the set-up cost for the lower quality product, adjusted for quality level, is lower than that for the higher quality product. In the case where the unit variable costs are zero, or they are proportional to quality level (so that unit variable costs, adjusted for quality, are the same), we show that an increase in the market size leads to (i) an increase in the fraction of firms that specialize in the high quality products, (ii) the market shares (both in value terms and in terms of volume of output) of high quality producers increases, and (iii) the prices of both types of product decrease. In the case where higher quality requires higher set-up cost (per unit of quality) but lower unit variable cost (per unit of quality), subject to certain bounds on the difference in unit variable costs, we obtain the result that an increase in the market size decreases the number of low quality firms, increases the number of high quality firms, and decreases the prices of both products. In the special case where the set up cost is independent of quality level, we find that all firms will produce both type of quality levels. In this case, an increase in the market size will reduce the value shares of low quality products, but will leave their volume share unchanged; and the market expansion induces a fall in the relative price of the low quality product, and in the prices of both products in terms of the numeraire good. We carry out an empirical test of a version of the model, where set-up costs now refer to set-up costs to establish an export market, and they vary according to the quality of product that the firm exports to that market. We show that the data supported the hypothesis that the average qualities of the product are higher for bigger export markets.
Abstract: We model an oligopoly where firms can choose the quality level of their products by incurring set-up costs that generally depend on quality level. If the set-up cost is independent of product quality, firms may choose to supply both types of quality. We focus on the long run equilibrium where free entry and exit ensure that the profit for each type of firm is zero. Using this framework, we study the implications of an increase in the market size. We show that for the existence of an equilibrium where some firms specialize in the low quality product it is necessary that the set-up cost for the lower quality product, adjusted for quality level, is lower than that for the higher quality product. In the case where the unit variable costs are zero, or they are proportional to quality level (so that unit variable costs, adjusted for quality, are the same), we show that an increase in the market size leads to (i) an increase in the fraction of firms that specialize in the high quality products, (ii) the market shares (both in value terms and in terms of volume of output) of high quality producers increases, and (iii) the prices of both types of product decrease. In the case where higher quality requires higher set-up cost (per unit of quality) but lower unit variable cost (per unit of quality), subject to certain bounds on the difference in unit variable costs, we obtain the result that an increase in the market size decreases the number of low quality firms, increases the number of high quality firms, and decreases the prices of both products. In the special case where the set up cost is independent of quality level, we find that all firms will produce both type of quality levels. In this case, an increase in the market size will reduce the value shares of low quality products, but will leave their volume share unchanged; and the market expansion induces a fall in the relative price of the low quality product, and in the prices of both products in terms of the numeraire good. We carry out an empirical test of a version of the model, where set-up costs now refer to set-up costs to establish an export market, and they vary according to the quality of product that the firm exports to that market. We show that the data supported the hypothesis that the average qualities of the product are higher for bigger export markets.
3. "Income Distribution, Vertical Differentiation, and the Quantity Competition"
Abstract: The paper analyzes the effects of the change of the income distribution on the equilibrium outcomes in the duopoly-quality model with quantity competition. The analysis results show that with zero quality-cost and an income inequality not too high, then both firms always choose the highest quality level. If the quality-cost is convex, then the average quality level will decrease and the vertical differentiation level will increase in the income inequality. These results are different from the Yurko (2011), who made a similar analysis under the quality-price competition model. Another contribution of the paper is that it gives the sufficient conditions for the single firm to choose multiple levels of the quality, i.e. the quality-cost function is convex, vertical differentiation is large enough, and the marginal cost is not too high.
Abstract: The paper analyzes the effects of the change of the income distribution on the equilibrium outcomes in the duopoly-quality model with quantity competition. The analysis results show that with zero quality-cost and an income inequality not too high, then both firms always choose the highest quality level. If the quality-cost is convex, then the average quality level will decrease and the vertical differentiation level will increase in the income inequality. These results are different from the Yurko (2011), who made a similar analysis under the quality-price competition model. Another contribution of the paper is that it gives the sufficient conditions for the single firm to choose multiple levels of the quality, i.e. the quality-cost function is convex, vertical differentiation is large enough, and the marginal cost is not too high.