Foreign competition and quality sorting: Overlaps in US and Chinese exports

• DOI: http://doi.org/10.1111/twec.12791
• Date: 01 August 2019
• Published date: 01 August 2019
• Author: Karsten Mau

2300

|

World Econ. 2019;42:2300–2325.

wileyonlinelibrary.com/journal/twec

Received: 15 April 2016

|

Revised: 16 February 2018

|

Accepted: 26 February 2019

DOI: 10.1111/twec.12791

ORIGINAL ARTICLE

Foreign competition and quality sorting: Overlaps

in US and Chinese exports

KarstenMau

SBE—Department of Economics,Maastricht University, Maastricht, The Netherlands

KEYWORDS

China, export competition, product heterogeneity, quality, United States

1

|

INTRODUCTION

Low- wage competition has been a major topic in international trade for more than a decade. Numerous

empirical studies analyse how firms in high- income countries adjust to increased import competition.

They document that affected firms downscale production and employment, upgrade their products

through research and innovation or switch into less- exposed product markets (e.g. Bernard, Jensen,

& Schott, 2006; Bernard, Redding, & Schott, 2011; Bloom, Draca, & van Reenen, 2016; Utar, 2014).

Yet, the most productive and dynamic firms—namely exporters—are likely to face such competition

also in foreign markets. A US exporter who ships goods to Germany not only competes with domestic

German suppliers, but may also be confronted with a Chinese firm that supplies a similar product. The

US exporter has a competitive disadvantage due to the lower wages paid in China.

One can assume that if production technologies are sufficiently different between the two coun-

tries, the resulting qualities of the good differ so that consumers in Germany would not substitute

the US for the Chinese variety (Schott, 2004, 2008). This is an optimistic view, however. Manova

and Zhang (2009) and Amiti and Freund (2010) document sizable contributions of processing trade

and foreign- invested firms' performance to China's recent export growth. Production offshoring has

become an own strand in the literature (Feenstra, 2010). In fact, it could be another US firm that out-

sourced production to China and ships its goods directly to the final destination. In that case, technol-

ogies would not differ so much, after all, and the quality advantage of the firm that operates from the

US would be reduced (or lost).

Despite the increasing integration of industrial organisation, labour markets and international

trade, low- wage competition in foreign markets is merely unstudied.1 An exception is Martin and

1This refers to export competition between high- and low- wage countries. “South–South” competition in exports is well docu-

mented. See, for instance, Greenaway, Mahabir, and Milner (2008); Mattoo, Mishra, and Subramanian (2017); Utar and Torres

Ruiz (2013).

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and

distribution in any medium, provided the original work is properly cited, the use is non- commercial and no modifications or adaptations are made.

© 2019 The Authors. The World Economy Published by John Wiley & Sons Ltd

|

2301

MAU

Méjean (2014) who analyse the evolution of French export quality and find a positive impact of low-

wage export competition. However, they neglect other adjustment strategies, which could be relevant

when the low- wage competitor already offers high quality, or if the product itself provides little scope

for differentiation (Johnson, 2012; Khandelwal, 2010). If that is the case, firms may be forced to cut

their prices, exit the market or switch to less- exposed products.

Motivated by these gaps in the existing literature, this paper investigates the evolution of US ex-

ports under exposure to Chinese competition in third markets. The US and China are chosen for two

reasons. The first reason is data availability. Publicly accessible data on US commodity trade feature

a high level of detail. This is particularly useful for the study of export unit values which can be used

as a proxy for product quality (Feenstra & Romalis, 2014). The other reason is that the two countries

belong to the world's largest exporters of manufacturing goods. This increases the number of cases

for which export overlaps can be observed; over time and within narrow product–destination pairs.

The analysis focuses on several outcome variables of US exports across an extensive set of prod-

ucts and destination markets. Starting with a pooled sample, it finds that US export values, quantities

and market shares decrease as China's market shares increase. Also, it is shown that a larger market

share of Chinese exports is associated with a higher probability of market exit of US exports. Evidence

of product switching cannot be observed in terms of increasing exports of non- exposed products

within an exposed industry.2 Yet, it appears that whenever the US exits one product market and enters

another, of the same industry, this is more likely to occur if the new product market is relatively less

exposed to Chinese competition than the old market. To establish confidence for a causal interpreta-

tion of these patterns, several robustness checks are carried out, including an instrumental variable

approach that exploits the event of Chinas WTO entry in December 2001. On average, the estimates

suggest that China's export expansion between 1996 and 2006 is associated with reduced US export

revenues and quantity of about 0.1%–0.5% and with a 2 percentage point reduction in market shares.

Average US export unit values appear to be insensitive to Chinese competition in the pooled sam-

ple, motivating further analyses along the respective cross sections of the panel. Carrying out the same

steps as before, but for individual destination countries, leads to similar findings, although some het-

erogeneity is revealed. Evidence and magnitude of statistically significant relationships appear to be

linked to the economic size of the destination, suggesting that economies of scale and product market

diversification might play a role. Turning to an analysis of individual products, a substantial amount

of heterogeneity is revealed. Statistically significant correlations between US exports and Chinese

market shares can be observed for only about 25% of the products that were included in the pooled

sample. This suggests that overall competition between Chinese and US exports is limited. However,

in individual markets it is likely to be much more intense than the results of the pooled sample suggest.

For US export quantity, for instance, large displacement effects are found in chemical industries, in-

cluding plastics and rubber. In most cases, the lower quantities of exports coincide with higher export

unit values, indicating aggregate increases in product quality. Notable decreases in US export unit

values are most prevalently found in the machinery and electronics industries.

Besides the empirical analysis, the paper also relates its results to existing theories of firms in

international trade. Findings conform to predictions of models featuring heterogeneous firms and het-

erogeneity across industries in their scope for quality differentiation (e.g., Baldwin & Harrigan, 2011;

Johnson, 2012). Yet, as such models also predict that changes in export unit values must coincide

with changes in exported quantities, cases where the latter is not observed are cannot be rationalised.

2Note that the term industry refers to a higher level of aggregation than a product; that is, several products are nested in one

industry. In this paper, a product refers to a 6- digit code of the Harmonized System nomenclature (HS), and an industry refers

to its corresponding upper- tier 4- digit code.