VoxEU Column Labour Markets Poverty and Income Inequality

Wage effects of trade reform with labour-market frictions and endogenous worker mobility

What are the effects of globalisation on wages and jobs in international and domestic firms? This column finds that data on employers and employees in Brazil tell a more nuanced story than the typical findings from firm-level data.

It is a well-established finding among economists that globally engaged firms and domestic firms differ in their productivity, their size, their employment composition, and their wages. These findings go on to suggest that further trade liberalisation will therefore lead to ‘inter-firm’ reallocations within an industry. In other words, the more productive, exporting firms will expand and the least productive firms will shrink or exit the industry (Melitz 2003). This may be a good thing for the overall economy, but in an economy with ingrained labour-market frictions, this within-industry reallocation can be tough for those workers stuck in the wrong company.

On the theoretical side, the international trade literature has discussed a variety of channels through which a decline in trade protection could result in differential changes in average wages at exporting firms relative to firms selling only to the domestic market, especially in the presence of labour-market frictions. For instance, if firms engage in some form of rent-sharing with their workers, the wages of workers employed in exporting firms, which experience a relative improvement in profits or market share after a decline in protection, will correspondingly be higher compared to workers employed in firms serving only the domestic market. Furthermore, exporters wishing to improve their product quality for foreign markets could respond to a decline in protection by paying (higher) wages in order to induce increased effort from otherwise identical workers. Moreover, if the labour allocation process is subject to search and matching frictions, exporting firms may screen workers more intensively, employ workers with higher match-specific ability (worker productivity that arises in the context of employment in that specific firm), and pay higher wages relative to non-exporting firms (Helpman et al. 2010). In this context, an opening of the economy to trade increases the wage gap between exporting and non-exporting firms, and increases wage inequality between otherwise identical workers in heterogeneous firms (within-group inequality).1,2

Does trade liberalisation, in fact, affect differently workers employed in exporting firms and non-exporting firms? Are labour-market frictions relevant to our understanding of these effects? In recent research (Krishna et al. 2011), we have explored these issues empirically using a detailed matched employer-employee dataset from Brazil for the years 1990-1998. The dataset traces individually identifiable workers across employers over time and contains detailed information on worker characteristics such as age, gender, education, occupation, and tenure at the firm, which allows us to suitably account for the role of both observable and unobservable worker, firm, and match characteristics in determining wages.

In contrast to most previous empirical analyses of the link between trade and wages, and consistent with recent theoretical models emphasising search dynamics and other frictions as important determinants of the relationship between trade liberalisation and wages (such as Helpman et al. 2010 and Davidson et al. 2008), we start our analysis by exploring whether the assignment of workers to firms is random. We find decisive support for non-random assignment of workers to firms (i.e. endogenous worker mobility across firms) in the data. This is illustrated in Figure 1 which plots, for the set of workers who switch firms, the transition probabilities of moving from firms in a productivity quintile in an initial period (psi_lag) to another quintile in a subsequent period (psi). Clearly, this distribution is not uniform – implying a rejection of the assumption of exogenous worker mobility.

Figure 1. Transition probabilities for workers who switch jobs from firms in an initial (psi_lag) productivity quintile to firms in the subsequent (psi) quintile.

The finding of endogenous worker mobility is important – an analysis of wages conducted without taking this into account will yield biased results on the relative wage effect of liberalisation on exporters. Using detailed information on worker and firm characteristics to control for compositional effects of trade liberalisation and firm-worker match-specific effects to account for the endogenous mobility of workers, we find an insignificant differential effect of trade openness on wages at exporting firms relative to domestic firms. We also find that the composition of the workforce improves systematically in exporting firms in terms of innate (time-invariant) worker ability and in terms of the quality of their worker-firm matches formed in these firms post-liberalisation. This finding also serves to underscore the relevance of analysing the effects of trade reforms on wages at the worker level rather than the level of the firm. If average (innate or match-specific) worker ability improves systematically in exporting firms following trade liberalisation, and this change is not taken into account, it will appear that trade liberalisation leads to a differential wage improvement for workers at exporting firms, even if this is not the case.

Specifically, our findings imply that, following trade liberalisation, a given worker (with fixed innate ability) who continues to be employed at a given exporting firm (with a fixed worker-match effect) will not experience any differential effect on his/her wage relative to another worker who continues to be employed at a non-exporting firm. All things being equal, workers who switch to firms with which they are better-matched will, however, earn higher wages because of their higher productivity there. Exporting firms will pay a differentially higher average wage post-liberalisation because of the improvement in the composition of their workforce in terms of innate ability and worker-firm match quality.

Our findings using matched employer-employee data suggest a quite different picture of the links between trade liberalisation and wages than those obtained by analysing the data at a more aggregate (firm) level, and underscore the importance of labour-market frictions and the allocation mechanisms in determining the effects of trade policy changes on wages.


Abowd, John, Kevin McKinney, and Ian M. Schmutte (2010), “How Important is Endogenous Mobility for Measuring Employer and Employee Heterogeneity?”, mimeo.

Davidson, Carl, Steven Matusz, and Andrei Schevchenko (2008), “Globalization and Firm Level Adjustment with Imperfect Labor Markets”, Journal of International Economics, July, 295-309.

Helpman, Elhanan, Oleg Itskhoki, and Stephen Redding (2010), “Inequality and Unemployment in a Global Economy”, Econometrica, 78(4):1239-1283

Krishna, Pravin, Jennifer Poole, and Mine Z Senses (2011), “Wage Effects of Trade Reform with Endogenous Worker Mobility”, NBER Working Paper No. 17256.

Melitz, Marc J (2003), “The Impact of Trade on Intra-Industry Reallocations and Aggregate Industry Productivity”, Econometrica, 71(6):1695-1725.

Yeaple, Stephen Ross (2005), “A Simple Model of Firm Heterogeneity, International Trade, and Wages”, Journal of International Economics, 65:1-20.

1 In general, the various theoretical channels linking trade protection to wages could also be associated with changes in between-group inequality (inequality between workers with different characteristics, such as, levels of education) and possibly even no change in inequality at all. For example, if trade liberalisation changes the returns to worker characteristics, this will result in a change in between-group inequality. Alternatively, if workers are paid competitive wages and the only impact of trade liberalisation is to reallocate workers across firms without any changes in the returns to worker characteristics, there will be no change in either within-group or between-group wage inequality. See Krishna et al. (2011) for a detailed discussion.

2 Of course, a rising wage gap between (heterogeneous) workers in exporting and non-exporting firms may be obtained in other contexts. For instance, Yeaple (2005) considers ex-ante identical firms which choose (high or low) technologies and the type of worker (high skill or low skill) to employ and finds that a reduction in trade frictions will result in an increase in the wage premium earned by high skill workers.


525 Reads