Two-way outsourcing, international migration, and wage inequality.

• Author: Yomogida, Morihiro

1. Introduction

   The rapid growth of international trade in business services is a topical issue among the general public as well as economists. Freund and Weinhold (2002) report that, from 1995 to 1999, the expansion of trade in business services has far outpaced the growth of overall goods and services. One driving force is the recent development of communication technology. The Internet makes it possible to access low-cost skills for many types of business services, such as data processing, software consulting and the reading of medical records such as X-ray films, ultrasound, and CT records. (1) As a result, the skilled-labor markets that used to be segmented are being integrated internationally.

   The media focus on the fact that high-tech companies are increasingly outsourcing their business services overseas to cut costs and alarms that such offshore outsourcing brings adverse effects to native workers (Moran 2003; Roberts 2005). Outsourcing also continues to attract the attention of economists (see, for instance, Bhagwati, Panagariya, and Srinivasan 2004; Blinder 2006). Mankiw and Swagel (2006, p. x) state, "As technology develops and global economic integration deepens, more jobs and people will be affected by actual or potential offshore outsourcing.... Further development of theoretical models will foster better understanding of the associated welfare impacts."

   Meanwhile, international movement of workers is another driving force that integrates labor markets across countries. The United Nations estimates that more than 175 million people in the world lived outside their countries of birth in 2002, and almost 1 in 10 people living in the more developed regions is a migrant. Especially in the United States, immigrants are everywhere, no matter the industry. Roberts (2005) reports that Wal-Mart is a leading business lobby in Washington, DC, for a more relaxed and tolerant approach to legal immigration, and the IT industry lobbies hard to relax visa restrictions on skilled immigrants from India.

   Obviously, the intense debates on outsourcing and migration stem from their effects on income distribution and especially on the wage inequalities across sectors and the wage gaps across countries. By casual observation, it might seem that outsourcing of production/services and migration of workers are two sides of the same coin. Thus in this article, we ask whether the labor market effects of offshore outsourcing are similar to those of international migration, and if so, in what aspects?

   Specifically, we adopt a general equilibrium model with a vertical production structure. There are two countries, each producing a final good with constant-returns-to-scale technology. The production of the final good consists of two stages. In the first stage, differentiated business services are combined and transformed into an aggregate service, which, together with unskilled labor, is used to produce the final good in the second stage. Each differentiated individual business service in turn requires skilled labor as input and is produced with increasing returns to scale technology in monopolistically competitive markets.

   We define two types of outsourcing. First, service outsourcing can be regarded as the arm's length trade of individual business services, and it arises because of product differentiation and scale economies. We find that service outsourcing is a two-way phenomenon that consists of both insourcing and outsourcing; that is, countries export and import business services simultaneously. This is in fact an important feature in the U.S. economy, but such a reality is often lost in the outsourcing debate. (2) Second, outsourcing in unskilled-labor-intensive processing occurs because of factor endowment differences. We examine how each type of offshore outsourcing and international migration affects the wage income distribution among skilled and unskilled workers, at both the national and international levels. (3)

   We identify two wage effects: the productivity effect and the factor abundance effect. In a closed economy, an increase in the number of skilled workers affects wages through two channels. The first is an increase of service input, reducing the skill premium of wages. This can be called the factor abundance effect. The second is the expansion of service varieties, leading to higher productivity in final goods production, which in turn raises both the skilled and unskilled wages. We call it the productivity effect. When the economy is open and outsourcing is allowed in both business services and labor processing, we find that skilled workers gain as long as the productivity effect outweighs the factor abundance effect. When we further allow international migration of workers, multiple equilibria arise. Which equilibrium emerges depends on the relative magnitude of the above two effects. Nonetheless, the wage impacts of outsourcing and migration can be ranked, regardless of which effect is bigger.

   We then use the model to examine whether offshore outsourcing and international migration exhibit complementarity or substitutability. On the one hand, outsourcing and migration are substitutes because migration erodes the incentive for outsourcing in unskilled-labor--intensive processing. On the other hand, they are complements because skilled-labor migration promotes service outsourcing. These results suggest that the effect of migration on outsourcing depends on which type of workers migrate across country borders.

   The literature on the relationship between trade and wages is extensive. Several empirical studies find that growing imports of labor-intensive goods from less developed countries help cause the wage gap between the more educated and less educated workers in developed countries (see Wood 1995), whereas other studies say that this effect is not important (e.g., Lawrence and Slaughter 1993). Our model is closest and complements Feenstra and Hanson (1996, 1999), who argue that trade expansion in intermediate goods plays an important role in explaining the observed demand increase for skilled workers. In their model, intermediate goods differ in skill intensities, and trade expansion causes a shift of intermediate good production across countries, leading to a rise in the average skill intensity. We show that trade in skilled business services raises service variety and hence the demand for skilled workers in both developed and developing countries because of an increase in their productivity, which is another channel through which trade raises the demand for skilled workers. (4)

   Like Ethier (1982), Helpman and Krugman (1985, ch. 11), Markusen (1989), and van Marrewijk et al. (1997) among others, we adapt the Dixit and Stiglitz (1977) model of monopolistic competition to international trade in differentiated intermediate services. (5) However, we adopt a production structure based on a Ricardo-Viner setting, in which each primary factor is specific to a sector; that is, skilled workers are used for business services and unskilled workers for final good production. The rationale is that workers with technical skills may be more mobile across countries than across sectors (e.g., Neary 1995), especially in this age of globalization and regional integration. Our approach can be used to analyze in a tractable way the various effects of both outsourcing and migration on income distribution and industrial structure, and these effects can be ranked.

   Harris (1998) points out that the Internet can cause "virtual mobility" of workers by reducing trade barriers in services. Whereas he focuses on such mobility across identical regions in a small open economy, our interest is in comparing the virtual mobility and the actual migration of workers between developed and developing countries, by explicitly considering two countries with skilled and unskilled labor, which enables us to examine the labor market effects of migration and outsourcing. (6)

   There is voluminous theoretical literature on international trade and migration, such as Ethier (1985), Brecher and Choudhri (1987), Bond and Chen (1987), and Ishikawa (1994) among others. However, none of these articles focus on outsourcing. Markusen (1988) and Francois (1994) examine migration issues with diversity of intermediate inputs and increasing returns in production. In Francois, differentiated inputs are assembled into a final good without using other primary factors. In our model, final good production requires unskilled labor as well as differentiated services, which enables us to examine the relationship between migration of workers and outsourcing of processing. Markusen (1988) explicitly considers the skill formation of workers, but the skilled wage is always equal to the unskilled wage in the diversified trade equilibrium, and hence income inequality cannot be analyzed. In our model, the skill premium is flexible so that we can analyze the effects of migration on the wage inequality. More importantly, the present article complements the existing literature by clearly demonstrating how the migration equilibrium is similar to or different from the outsourcing equilibrium in their effects on the wage distribution among skilled and unskilled workers, at both the national and international levels.

2. The Two-Sector Closed Economy

   We start with the simplest case of one country with two sectors, final good and business service. Business services ([Z.sub.1], [Z.sub.2] ..., [Z.sub.n]) are produced under variety-specific increasing returns to scale, and the market for business services is characterized by monopolistic competition. The final good (y) is produced with constant returns to scale technology under perfect competition, using unskilled labor L and a number of differentiated business services [Z.sub.1], ..., [Z.sub.n], in the following form:

   y = [L.sup.[beta]][S.sup.1 - [beta]], 0 < [beta] <1, (1)

   where S is the index of business services,

   S =...