Trade Costs and Current Accounts

• Date: 01 October 2016
• Published date: 01 October 2016
• DOI: http://doi.org/10.1111/twec.12318
• Author: Clément Nedoncelle

Trade Costs and Current Accounts

Cl

ement Nedoncelle

Universit

e de Lille, CNRS, LEM UMR 9221, Lille, France

1. INTRODUCTION

THE main issue tackled in this paper is whether global current account imbalances are a

consequence of trade liberalisation. These growing imbalances have become a subject of

interest and concern as recalled in Obstfeld (2012). Current accounts are around four times

more dispersed in 2010 than they were in 1990, which implies diverging current account bal-

ances. Over the same time period, trade costs have decreased allowing larger trade flows, at

both the intensive and the extensive margin (Arvis et al., 2013). The present paper asks

whether these trade cost reductions are a plausible determinant of these increasing current

account imbalances.

An extensive literature has investigated the key forces driving current account dynamics.

The empirical literature has listed many potential candidates including openness, financial

integration and demographic trends, among others. Chinn and Prasad (2003) provide a survey

of the empirically relevant determinants of current account balance variations across countries

and over time. Surprisingly, the role of trade liberalisation has received less attention. How

much of the current account imbalances is linked to regional integration? This paper provides

an explanation for these growing imbalances which relies on the nature of the regional inte-

gration process. More specifically, it supports the idea that national industrial structure s are

affected by the changing trade costs faced by firms when they export goods or services. Trade

cost variations affect national trade structures, the level of trade and its nature in terms of fac-

tor service. They are likely to affect current account balances insofar as these changes in the

industrial structure translate into changes in the national investment level.

I build on the recent approach by Jin (2012) by emphasising the role of trade costs in shap-

ing industrial structures and the current account balance. Two predictions are derived from

the model and brought to the data. The model predicts (i) that trade cost variations interact

with changes in the industrial structure to determine the effects of regional integration on cur-

rent accounts and (ii) that the depth of regional agreements shapes the response of current

accounts to changes in trade costs and in specialisation patterns: the joint effect of trade costs

and of the national capital intensity of production on current accounts is exacerbated for coun-

tries with highly integrated institutions.

The current account ratio to GDP is regressed on the capital content of exports, its interac-

tion with trade costs for the 1988–2005 period. A full set of factors is included in the specifi-

cations, factors which could potentially influence current account balances by controlling for

savings, openness and other country-specific variables. The endogenous determination of the

capital content of exports is corrected for by implementing an instrumental variable strategy.

The World Governance Indicators from Kaufmann et al.’s (2010) data on institutional quality

are used as instruments for the capital intensity of production at the country level. Following

I am fully indebted to J

er^

ome H

ericourt for his continuous support and his comments on previous

versions of this work. This paper has also benefited from discussion at the 2013 ETSG annual confer-

ence (Birmingham) and IEG seminar in Louvain-la-neuve.

©2015 John Wiley & Sons Ltd 1653

The World Economy (2016)

doi: 10.1111/twec.12318

The World Economy

Acemoglu et al. (2001, 2005), institutional quality is a strong determinant of growth, and a

significant part of the effect is channelled through investment.

The estimates report a positive joint effect of capital intensity and trade costs on current

accounts: ceteris paribus, when trade costs decrease, countries where production is oriented

towards capital-intensive activities have larger current account deficits, consistently with the

existence of a higher demand for capital in these countries. Estimations controlling for endo-

geneity confirm this result. The results are confirmed by robustness checks which deal with

unobserved heterogeneity and potential lagged effects.

I infer from these results that, aside from the direct effect generally put forward in standard

macro analysis, changes in production patterns may be an additional channel of impact of

trade liberalisation on current accounts.

The remainder of the paper is composed of five sections. The next section provides a liter-

ature review. The theoretical model, emphasising the joint effect of trade costs and changes

on the industrial structure on current accounts, is presented in Section 3, where the two testa-

ble predictions are derived. Section 4 translates the theoretical predictions int o an empirical

strategy and defines data sources. Section 5 presents empirical evidence of the impact of capi-

tal intensity and trade cost variations on current accounts and investigates how institutional

integration may shape the relation between current accounts and capital intensity. Section 6

shows some robustness check results, including an investigation of lagged effects and unob-

served heterogeneity. The last section provides concluding comments.

2. LITERATURE REVIEW

This paper contributes to two strands of the literature. First, it is related to the literature

that rationalises the link between regional integration and global imbalances. Focusing on cur-

rent accounts, Blanchard and Giavazzi (2002) advocate that integration is likely to increase

current account imbalances, as borrowing and lending conditions are implicitly relaxed in an

integrated area. Borrowers are led to borrow more in an integrated area, implying increasing

current accounts. Schmitz (2011) provide evidence of growing financial integration in the euro

area as a source of increasing current accounts in member countries, where the single cur-

rency plays a substantial part. Interestingly, the introduction of the euro in 1999 is mainly

interpreted as deepening financial openness in the euro zone, while the euro may also be a

source of a decrease in exchange rate volatility and thus in trade costs (Berthou and Fontagn

e,

2013). I explicitly focus on this latter hypothesis and provide insights on its magnitude in cur-

rent account dynamics.

Second, a recent interest in the relationship between trade and capital flows has emerged.

While in standard HOS theory trade and capital flows are perfect substitutes, a major recent

improvement has been to depart from the standard HOS framework, thus allowing both for

trade in goods and in factors. Using Obstfeld and Rogoff’s (1996) terminology, this literature

departs from the standard inter-temporal analysis by combining both intra-temporal trade and

inter-temporal trade. In this field, Jin (2012) provides a seminal contribution and a tractable

theoretical framework which can replicate the empirical features of the direction of capital

flows, as mentioned by Gourinchas and Jeanne (2013) who provide evidence that capital tends

to flow towards countries with high level of investment and low growth. Jin (2012) provides

a theoretical framework that allows for the endogenisation of current accounts by trade struc-

ture. She shows that the capital intensity of the industrial structure is a quantitatively impor-

tant determinant of current accounts.

©2015 John Wiley & Sons Ltd

1654 C. NEDONCELLE