Labor productivity and trade liberalization: Evidence from China

Wang Bin

 December 27, 2016

Published by Investing Across Borders LLC, in conjunction with Université Paris 1 – Panthéon Sorbonne

Advising Professor Sandra Poncet, Paris School of Economics & Université Paris 1 - Panthéon Sorbonne



International trade is a very important part of China’s economy. And one of the most important issues is the trade liberalization on labor productivity. This paper is based on the theory of Melitz and Ottaviano (2008). We revise the empirical model developed by Chen et al. (2009) so that it will be more consistent with the theoretical model. We separate labor productivity into two dimensions: extensive margin labor productivity and intensive margin labor productivity. For the theoretical model, which takes intensive labor productivity more heavily into consideration, we take value added per unit per worker as the input labor productivity in the empirical analysis. The model is in the sector level. All the sector level databases are from CEPII and GDP data is from World Bank. We take China as the domestic country and 180 countries are included in the sample from 1996 to 2006. We make pairwise of China – foreign country – sector to estimate the short run effect by using OLS, fixed effect model and first order difference model, and to estimate the long run effect by using error correction model.

The results show that domestic trade liberalization has a positive effect on labor productivity, while foreign trade liberalization has a negative effect on labor productivity.


From 1978, after the trade reform and opening-up policy, China has achieved great success in international trade: at the beginning of the trade reform, the total amount of imports and exports was only US $20.6 billion, reaching $100 billion for the first time in 1988. The growth rate of the total amount of imports and exports sped up after 2002. During 2002-2007, the total growth amount of imports and exports exceeded 20% per year. Although the financial crisis in 2008 put a damper on trade, with imports and exports falling, thereafter both imports and exports increased rapidly. Until 2011, the total amount of imports and exports was $3.641 trillion. In the past 30 years, imports and exports of China increased by an average of 18% per year. China has been the second largest trade country in the world, behind the United States.

In fact, in order to speed up economic reform and evolve into a multilateral trading system, China has implemented a number of policies for eliminating tariffs and non-tariff barriers since the 1990s. From 2001, after becoming a member of WTO, China implemented the trade liberalization policy step by step. As the trade enivornment gradually opens up to China, how will trade liberalization influence the China’s economy? In this paper, we mainly use the theoretical model developed by Melitz and Ottaviano (2008) for theoretical support, and revise the empirical model developed by Chen et al. (2009) to analyze the effect of trade liberalization on labor productivity in China.

Literary review

In the past several decades, the total amount of global trade increased rapidly, with the total imports and exports increasing from US $4.113 trillion to $33.248 trillion in 2015 (Data from WTO statistics database: This phenomenon arose from an intense discussion about trade liberalization in academia. One of the topics that drew most attention was the trade openness on productivity.

Bernard et al. (2003) and Melitz (2003) built a model of international trade with firm heterogeneity. The model built a bridge between sector productivity and firm heterogeneity, making productivity as a function of trade openness at the sector level.
Bernard et al. (2003) extended Ricardo’s model, emphasizing the importance of the difference in firm-level efficiency to emerging firm heterogeneity. When low-productivity firms exit due to high competitiveness, labor force moves to the export-switching firms with higher productivity, thus sector productivity improves. In the Melitz (2003) model, trade liberalization intensifies the factor competition, increasing the productivity threshold for firms’ survival, and impels the exit of low-productivity firms so that the sector average productivity increases.  Meanwhile, exports improves sector productivity by expanding high-productivity firms and contracting low-productivity firms. Major theoretical breakthroughs associated with Melitz (2003), and Bernard, Eaton et al. (2003) among others have resulted in new ways of thinking about firmheterogeneity and participation in international markets (Greenway & Kneller, 2007).

Some scholars make an empirical analysis of sector productivity based on the theories above. Trefler (2004) states that labor productivity could increase by 14% in the sectors where tariffs of final products decreases most. Schor (2004) anaylzes tariffs by using sector level data in Brazil. He states that tariffs of intermediate goods have negative effects on productivity. Fernandes (2007) tests the impact of trade liberalization on productivity in Colombian manufacturing industries and finds a strong positive impact of tariff liberalization on plant productivity.

Melitz and Ottaviano (2008) build up the theoretical model of trade with firm heterogeneity from the aspect of trade affecting the toughness of competition, which feeds back into the selection of heterogeneous producers and exporters in that market. This results in the exit of low-productivity firms and the improvement of average productivity in the sector. This effect is similar to the analysis in Melitz (2003), while the mechanisms are different. Melitz (2003) shows that the exposure to trade will induce only the more productive firms to enter the export market so that real wage raises and competition in factor market becomes more intense, and will simultaneously force the least productive firms to exit. In this model, the firms’ mark-ups and demand price elasticity is given and exogenous. But keep in mind that according to Melitz and Ottaviano(2008), the firms’ mark-ups and demand price elasticity is endogenous, lower trade costs means more foreign firms compete in the domesticmarket which leads to more fierce competion in domestic market. The variety increases, the demand elasticity increases in the given demand level, and the price and mark-ups decrease. These effects force the least productitive firms to exit. Melitz and Ottaviano (2008) also consider the short run and long run effects of trade liberalization on sector productivity seperately: in the short run, trading intensifies the competition and increases the average productivity in the sector significantly; in the long run, the effect becomes less significant and even negative.

Chen et al. (2009) transfer the theoretical model into empirical model. They use data in ten manufacturing sectors from seven EU countries (Belgium, Germany, Denmark, Spain, France, Italy and the Netherlands) in the period from 1989 to 1999 to analyze the relationship between trade openness and sector-level competitive effect. Chen et al. (2009) develop an estimation equation, in which international differences in trade openness at the sector level reflect international differences in the competitive structure of markets.They find that in the short run, trade openness exerts competitive effect, but in the long run, the effects are more ambiguous and may even be anti-competitive.

During the last three decades the volume of international trade has increased enormously, among developed countries since the 80’s and extending to developing countries since the 90s (Navas & Licandro, 2011). Take Asian countries as an example, the share of China’s exports in the world increased from only 1.5% in 1990 to 11.13% in 2012, and the share of China’s imports in the world increased from 1.5% in 1990 to 9.78% in 2012; The total amount of imports and exports in India increased from $41.55 billion in 1990 to $783.83 billion in 2012; Malaysia’s import and export increases from $58.71 billion in 1990 to $424 billion. Getting rid of the financial crisis years (1998 and 2009), the annual growth rates of India and Malaysia’s imports and exports reached 17.60% and 13.65%, respectively. Meanwhile, Thailand, Vietnam, and Laos also have increased their development in involving the Asian market and global trade system.

In developing countries, improving productivity is crucial for trade and for the economy. When more and more developing countries participate in international markets, systematic analysis of the effects of trade openness on labor productivity becomes nessecary. We adopt the models ofMelitz and Ottaviano (2008) and Chen et al. (2009), and apply them to the analysis of the effects of trade openness on labor productivity in sector levels in China.

Theoretical model

Empirical Model

Research Methods and Data Description

Trade openness

Labor productivity



Empirical Analysis

Firstly, we only focus on the short-run effect of trade openness on productivity in sector level without error correction term. Secondly, error correction term is included and we use the direct method, that is, opening the brackets directly, to estimate the long-run and short-run effects together by using OLS.

Short-run effect

In this part, we use an OLS, fixed-effect model and first-order difference to estimate the impact of trade liberalization to labor productivity in the short run.

From Table 3, the empirical evidence shows that along with the level of trade liberalization, the labor productivity increases as well. But the level of trade liberalization in the foreign countries has a negative effect on labor productivity in China. Meanwhile, the increase of relative level of trade liberalization in China has a positive effect on domestic labor productivity. These empirical results are consistent with theoretical models. Trade liberalization will promote the labor productivity increasing.

However, the increase in the number of active firms in domestic and foreign markets will lead to a different conclusion. The possible reason is that the number of active firms is a substitute variable for N, and therefore the substitution may not be perfect. So in some circumstances, the mechanism inside the substitution will be more complex. In the situation of China, we can get a different empirical result. If the number of active firms increases in China, the labor productivity will decrease. While the number of active firms decreases in foreign countries, labor productivity will increase. The mechanisms are, in the short run, if the firms not producing (exporting) find that the price of the product is increasing, or the demand elasticity and product variety are decreasing, they will choose to produce (export) again. So increasing the number of firms in the market sometimes reflects that less productive firms enter into the market rather than productivity improving.

Long-run effect

Generally, there are two ways to run an error correction model. The first one is Engel and Granger 2-step approach, and the second is direct estimation by using OLS. Here, we use the direct estimation method.

In this part, we use the full estimation equation to run the regression. The relations between trade liberalization and labor productivity both in the long run and short run are showed below:

In Table 4, we know that the increase of trade liberalization in China leads to an increase of labor productivity, while the increase of trade liberalization in foreign countries will destroy domestic labor productivity. The relative trade liberalization effects domestic labor productivity positively, which is consistent with the short run effect. And in the long run, both the number of domestic active firms and the number of foreign active firms will have a positive effect on productivity because the mechanism which is described above in the short sun may not exist for a very long time. As the number of firms increases again, the demand elasticity and product variety will increase again. It leads to an improvement of labor productivity again.

Because of the existence of entry and exit costs, the long-run effect of trade liberalization is always very ambiguous. In the case of China, the long-run effects of trade liberalization are very similar to the short-run effects while the increase of the number of active firms both in the domestic and foreign market has a positive effect on labor productivity.

When considering market scale and wages, the effects are not so significant in China.


Melitz and Ottaviano (2008) developed a firms' heterogeneity trading model. They stated that trading openness makes more foreign firms compete in the domestic market. When the demand is given, price and mark-ups will decrease, so that less productive firms will exit. The total productivity in the sector will be higher. Chen et al. (2009) transform the Melitz and Ottaviano (2008) model into an empirical model. They use international differences of trade liberalization to explain the international differences of sector level productivity. In our paper, we refer to the theoretical model of Melitz and Ottaviano (2008) and the empirical model of Chen et al. (2009), and revise the labor productivity input variable to make sure that the empirical model is more consistent to the theoretical model.

During the analysis, we take China as the domestic country, using sector database from CEPII and GDP data from the World Bank. In the sample, 180 countries are included in the empirical analysis, with import penetration rate as a measurement of trade liberalization, and value added per unit per worker to represent labor productivity. After making the pairwise between China and other foreign countries, we estimate the country pair – sector combination by using OLS, fixed-effect model and first-order difference model.

The results show that, with the increase of the level of trade liberalization in China, labor productivity will increase, while the increase of the level of trade liberalization in foreign countries has a negative effect on China’s labor productivity. It is consistent with the theoretical model. WhenChina becomes more open and increases its trade liberalization level, more foreign firms enter the domestic market and compete. It increases the product variety and promotes competition in domestic market, so that the average sector productivity increases. To conclude, the tradeliberalization in China has a competitive effect, which has a positive impact on relative productivity in China. The result is that, the productivity gap between China and other developed countries will be smaller, and the international competitiveness of China will be higher.

(1) ChenNatalie, ImbsJean, & ScottAndrew. (2009). The dynamics of trade and competition. Journal of International Economics, 77.
(2) Fernandes, A. M. (2007). Trade policy, trade volumes and plant-level productivity in Columbian manufacturing industries. Journal of International Economics, 71, pp. 52-71.
(3) Greenway, D., & Kneller, R. (2007, 2). FIRM HETEROGENEITY, EXPORTING AND FOREIGN. The Economic Journal, 117, pp. F134-F161.
(4) Melitz, M. J., & Ottaviano, G. I. (2008). Market Size, Trade, and Productivity. Review of Economic Studies, 75, pp. 295-316.
(5) NavasAntonio, & LicandroOmar. (2011). Trade Liberalization, competition and growth. The B.E. Journal of Macroeconomics, 11, pp 1935-1690.
(6) SchorAdriana. (2004). Heterogeneous Productivity Response To Tariff Reduction: Evidence From Brazilian Manufacturing Firms. Journal of Development Economics, 75, pp 373-396.
(7) Trefler, D. (2004, 9). The Long and Short of the Canada-U.S. Free Trade Agreement. The American Economic Review, 94, pp. 870-895.

© Copyright 2016 Investing Across Borders, LLC. All rights reserved.