Economic integration across Latin America: Evidence from labour markets, 1990–2013

• Author: Daniel Lederman,Raymond Robertson
• DOI: http://doi.org/10.1111/twec.12610
• Date: 01 May 2018
• Published date: 01 May 2018

ORIGINAL ARTICLE

Economic integration across Latin America:

Evidence from labour markets, 1990–2013

Daniel Lederman

1

|

Raymond Robertson

2

1

The World Bank, Washington, DC, USA

2

Texas A&M University and IZA, College Station, TX, USA

1

|

INTRODUCTION

Possibly motivated by empirical findings suggesting a strong positive link between trade and

growth (Frankel & Romer, 1999; Noguer & Siscart, 2005), governments have sought to deepen

economic integration. Trade and capital flows, two often-cited bellwethers of economic integration,

increased over the same period. While international trade and financial flows have increased, there

is considerable debate about the extent of actual international economic integration that has fol-

lowed. One dimension that is particularly debated is whether labour markets have become more

integrated across borders. Whereas there is a voluminous literature on the relationship between

globalisation and wages, relatively little attention has been directed towards understanding the

degree of international labour market integration.

Labour market integration matters for at least two reasons. First, the core of the trade and growth lit-

erature (e.g., Frankel & Romer, 1999) is based on the premise that increased economic integration leads

to higher GDP per capita and, by implication, higher wages. This conclusion is a small step from neo-

classical trade models (such as the Heckscher–Ohlin theorem) that predict that economic integration

leads to wage convergence between high-wage and low-wage countries.

1

The exchange of goods , and

the resulting convergence in output prices, should raise wages in low-wage countries and contribute to

wage convergence. This idea is commonly known as the factor-price equalisation theorem.

Empirical results about wage convergence across countries have been mixed. Ben-David (1993)

compares income differentials before and after the removal of trade barriers. He found that

incomes across signatories of European Economic Community began to converge after trade liber-

alisation. Research from other regions includes Bigsten, Mengistae, and Shimeles (2013) from

Ethiopia, Yi and Zhou (2016) from China and Morten and Oliveira (2016) from Brazil. Notably,

Bloom and Noor (1995) find the labour market integration increased sharply among East and

South-East Asian countries from 1980s to 1991. The integration was achieved mainly through the

sharp increase in trade, but the contributions of labour mobility and capital mobility were smal l.

1

Some papers suggest that reducing barriers to trade leads to Heckscher–Ohlin–Samuelson (HOS) effects. Michaels (2008)

finds evidence of HOS effects within the United States after trade costs were reduced by the US highway system. Other

international examples that compare output prices with changes in relative wages as predicted by the strict specification of

the Stolper–Samuelson theorem relatively strong support for the Stolper–Samuelson theorem, such as Beyer, Rojas, and Ver-

gara (1999) for Chile, Robertson (2004) for Mexico and Gonzaga, Menezes Filho, and Terra (2006) for Brazil.

DOI: 10.1111/twec.12610

World Econ. 2018;41:1269–1287. wileyonlinelibrary.com/journal/twec ©2018 John Wiley & Sons Ltd

|

1269

Other papers (Clemens, Montenegro, and Pritchett, 2009; Kiyota, 2012; Gandolfi, Halli day, and

Raymond, 2017; Maskus and Nishioka, 2009) suggest that integration does not reduce the “place

premium”characterising wage differences across countries. The lack of evidence supporting factor-

price equalisation, however, suggests that more international comparisons are needed.

The second reason why labour market integration matters is that the lack of labour market inte-

gration may suggest a lack of allocative efficiency (Caruana-Galizia, 2015; Chowdhury & Mukher -

jee, 2016). One of the goals of international economic integration is to take advantage of

differences in prices across borders. Indeed, price differences across borders are the source of gains

from trade. As people take advantage of price differences across borders, these price differences

would diminish and perhaps even gradually disappear. Persistent wage differences across borders

may suggest that economic integration is incomplete, and therefore, there are still signi ficant

opportunities to realise these efficiency gains.

Indicators of labour market integration are an important tool for understanding the consequences of

international integration. The first step is developing appropriate metrics and baseline statistics for labour

market integration. The goal of this paper was to examine three indicators of labour market integration

for Latin America. The first measure is long-run absolute wage convergence. The second is the co-

movement in wages across countries. The third is the speed of convergence. All three measures follow

from a sizable literature on both market integration g enerally and factor-market integration in particular.

Latin America is the focus of this paper for various reasons. First, Latin America experienced a

rapid expansion in the number of bilateral and multilateral trade agreements, and a commensurate

reduction of intraregional trade barriers, over the last 25 years, thus making it an inte resting region

to analyse from the viewpoint of international market integration.

2

Second, labour market integra-

tion in Latin America has been less studied than labour market integration in Asia. Third, newly

combined data sets from labour force and household surveys in Latin America offer the opportu-

nity to advance the literature generally on labour market integration.

The rest of the paper follows the usual structure. The next section briefly reviews the literature

on indicators of market integration (of which labour market integration is a special case) and then

presents theoretic considerations that guide our empirical measures of labour market integration.

The third section presents the results of the three indicators, and the fourth section concludes with

a discussion of the implications of these findings.

2

|

MEASURING FACTOR-MARKET INTEGRATION

The literature provides three alternative concepts that are useful for assessing the degree of integra-

tion of factor markets across borders. We first review such indicators and then provide a theoretical

framework with the objective of showing how these concepts are applied to labour markets.

2.1

|

Three indicators of market integration

Measures of market integration are found throughout the academic literature. Perhaps t he most com-

mon is price convergence.

3

Engel and Rogers (1996) and McCallum (1995) are two well-known

2

See Bown, Lederman, Pienknagura, and Robertson (2017) for a recent comprehensive assessment of regional economic

integration in Latin America and the Caribbean.

3

In particular, studies of agricultural markets are common. Examples include Dawson and Dey (2002), Ghosh (2003),

Mohanty and Langley (2003) and Mohanty, Peterson, and Smith (1996).

1270

|

LEDERMAN AND ROBERTSON