China’s rise, asymmetric trade shocks and exchange rate regimes

• Published date: 01 February 2019
• Date: 01 February 2019
• DOI: http://doi.org/10.1111/roie.12353
• Author: Francesca G. Caselli

Rev Int Ecom. 2019;27:1–35. wileyonlinelibrary.com/journal/roie

|

1

© 2018 John Wiley & Sons Ltd

1

|

INTRODUCTION

The impact of China’s growth on advanced economies has been at the center of the economic debate

over the past several decades. China’s economic integration into world markets has been unprece-

dented, with the country’s real exports increasing 12 times between 1990 and 2007 (Di Giovanni,

Zhang, & Levchenko, 2014). This paper explores the effect of China’s rise on the competitiveness

of OECD countries, controlling for countries’ specific exposure to Chinese competition. The paper

argues that China’s surge as a trade power constitutes an interesting experiment to study the effect of

an asymmetric trade shock on advanced economies real effective exchange rates (REERs). OECD

countries are subject to different shocks according to their industry specialization and exposure to

trade competition coming from China.

The economic debate has devoted considerable attention to the impact of China on Western econ-

omies’ welfare and unemployment (Autor, Dorn, & Hanson, 2013; Di Giovanni et al., 2014), but

less on its consequences on the dynamics of relative prices, namely competitiveness. This issue has

Received: 25 September 2016

|

Revised: 5 January 2018

|

Accepted: 11 April 2018

DOI: 10.1111/roie.12353

ORIGINAL ARTICLE

China’s rise, asymmetric trade shocks and exchange

rate regimes

Francesca G. Caselli

International Monetary Fund, Washington,

District of Columbia

Correspondence

Francesca G. Caselli, International Monetary

Fund, 700 19th Street, N.W., Washington,

DC 20431.

Email: fcaselli@imf.org

Abstract

China’s rapid growth provides a natural experiment to

study the effects of asymmetric trade shocks on the com-

petitiveness of OECD countries. The different levels of

exposure to Chinese trade competition, as measured using

an index of export similarity, triggered asymmetric shocks

as China’s trade surged. Motivated by a Ricardian frame-

work, this paper finds that countries with exports similar

to those of China experience a loss in competitiveness

compared with countries with a different trade structure.

Once an additional layer of distinction is introduced be-

tween fixed and flexible exchange rate regimes, I find that

countries with a fixed exchange rate and with relatively

high similarity to China experience a real appreciation.

2

|

CASELLI

become even more important in light of the divergent competitiveness dynamics between countries in

the core and periphery of the European Monetary Union (EMU). What has been the role of China in

determining the loss of competitiveness in southern EMU countries, given that they tend to be more

exposed to Chinese competition because of their trade specialization? This paper will attempt to an-

swer this question. When the debt crisis unfolded, it became clear that the euro area has been accumu-

lating substantial imbalances, not consistent with healthy convergence patterns (Giavazzi & Spaventa,

2011). Moreover, it became evident that competitiveness divergences in the periphery vis‐à‐vis the

core were endogenous to other shocks (Wyplosz, 2013). Consistently with this view, Chen, Milesi‐

Ferretti, and Tressel (2013) look at the emergence of China as a potential driver of the European

imbalances and find that exports of several Southern European countries were negatively affected by

Chinese competition.

Using a simple Ricardian framework, developed by Dornbusch, Fischer, and Samuelson (1977), I

investigate whether countries with greater similarity to China tend to experience a worsening of their

competitiveness. To study the effect of the rise of China on advanced economies’ relative prices, I

augment a real effective exchange rate regression with the share of China’s GDP in world GDP. I

use unit labor costs (ULC) deflated real effective exchange rates and, as controls, the variables from

Ricci, Milesi‐Ferretti, and Lee (2013). The regression identifies the asymmetric dimension of the

China Shock by exploiting differences in trade specializations. More specifically, I build the Finger

and Kreinin (1979) index of export similarity that captures the overlap between each country’s bundle

of exports and Chinese exports. This measure accounts for the degree of exposure to China’s compe-

tition since it incorporates information about market share and product penetration (Schott, 2008). A

well‐known drawback of relative unit labor costs is that, in the presence of quality differences, they are

a poor proxy for competitiveness. Quality, especially at the macro level, is very difficult to measure. I

tackle this problem by building the Hausmann, Hwang, and Rodrik (2007) index of export sophistica-

tion. This index is a weighted average of the productivity level associated with a country’s exports and

is intended to capture the exports’ overall quality level. Since Friedman (1953) and Mundell (1961),

it has been well known that countries that peg their exchange rate or enter a currency union lose the

possibility of counterbalancing an increase in relative prices by way of a nominal devaluation. To test

for this possibility, this paper compares the effect of the China Shock for countries with fixed and

flexible exchange rates.

The empirical results suggest that countries with higher export similarity to China experience a

loss in competitiveness compared with dissimilar countries. This indeed suggests that the shock com-

ing from China has been asymmetric and that increasing competition in specific sectors tends to neg-

atively affect countries with a similar specialization. Once an additional layer of distinction between

fixed and flexible exchange rate regimes is introduced, I find that countries with an export pattern

similar to China, and which cannot adjust their exchange rate, tend to experience real appreciation.

The empirical analysis also investigates whether there is a threshold of export similarity after

which the relationship between the China Shock and competitiveness changes. The paper finds that a

threshold indeed exists, and that the effect of the China Shock is stronger for European countries. The

results are robust to different measures of unit labor costs deflated REERs, to the potential endogene-

ity of the export similarity index and the exchange rate regime dummy, and to a dynamic specification

that takes into account the sluggish adjustment of prices.

This work contributes to the literature that explores the factors behind divergent competitiveness

dynamics in the EMU and the impact of asymmetric trade shocks on currency unions. Friedman

(1953) argued in favor of flexible exchange rates because they help in achieving external adjustment

following an external shock affecting a country’s competitiveness in the presence of price and wage

rigidities.1 This is a clearly relevant argument for countries joining a monetary union and thereby

|

3

CASELLI

losing monetary independence and the ability to devalue their currencies. Mundell (1961) stresses

that in a currency area, monetary independence may not be needed if shocks are symmetric, or when

shocks are asymmetric, labor mobility is required. Accordingly, an asymmetric shock to the euro area,

where monetary independence has been lost, poses important questions in terms of policy responses

that represent alternatives to nominal devaluations. One answer has been provided in a recent strand of

literature that advocates fiscal devaluations. Farhi, Gopinath, & Itskhoki (2014) show that a uniform

increase in value‐added taxes and a reduction in payroll taxes can replicate the real allocations reached

through a nominal exchange rate devaluation.

The paper is also linked to the literature on the effect of China’s rise on advanced economies’ trade

specialization and welfare. The most natural framework to study the effect of China’s trade integra-

tion on relative prices is a Ricardian framework, and more specifically the Dornbusch, Fischer, and

Samuelson (1977) (DFS) model. In an n‐goods setting, they show that a transfer of the least‐cost tech-

nology to the low‐wage country decreases real income in the high‐wage country. More recently, the

Ricardian trade model has been used to examine how the rise of China affects the welfare of advanced

economies. Samuelson (2004) argues that China’s growth and increasing similarity to western coun-

tries can generate welfare losses in the latter. Exploiting a multi‐country Eaton and Kortum (2002)

model, Di Giovanni et al. (2014) examine how changes in relative sectoral productivity in China affect

the welfare of the United States. In a two‐country setting, they find that welfare is lowest when relative

sectoral productivity is identical in two countries, confirming Samuelson’s result. However, with more

than two countries, welfare changes are also driven by the sectoral productivity of all the countries

serving the U.S. market. Autor et al. (2013) follow a related line of reasoning and look at the effect

of Chinese import competition on U.S. local labor markets. In fact, they allow for asymmetric trade

shock on U.S. local labor markets building a measure of import exposure to Chinese competition.

Finally, Schott (2008), analyzing very detailed trade data, finds that China’s export bundle overlaps

significantly with that of advanced economies.

The remainder of the paper is organized as follows. Section 2 discusses the theoretical channels

that might drive the empirical results. Section 3 describes the construction of the similarity and export

sophistication index. Section 4 presents the data, the empirical specification, and the results. Section

5 provides a battery of robustness checks and Section 6 concludes.

2

|

THEORETICAL FRAMEWORK

While not aiming to test a specific theory, this section underlines the channels that drive my empiri-

cal results relying on the DFS model. This theoretical framework is consistent with the definition

of competitiveness used in the empirical analysis, namely the unit labor costs deflated real effec-

tive exchange rate (Chinn, 2006; Golub, 1994; Manasse, 2013). Moreover, the DFS model frame-

work can be adapted to investigate China’s emergence as a trade shock and its effects on advanced

economies’ relative prices. This shock can be modeled as a harmonization of relative technologies

between developed countries and the world. For countries with a similar trade specialization in

labor‐intensive goods, this implies increased competition and adjustment in relative wages to re-

store the equilibrium. More specifically, harmonization means that unit labor requirements tend to

be equalized across countries. Dornbusch et al. (1977) show that such harmonization of technolo-

gies, while having a positive effect for innovating low‐wage countries, generates a reduction in real

income in high‐wage countries.The text that follows will provide more formal explanations of these

channels.