استحکام بازار کار و رشد مبتنی بر R-D در اقتصاد جهانی

This paper constructs a dynamic model of trade between two countries, one with flexible wages (America) and another with a rigid wage for less-skilled workers (Europe). The model incorporates global R&D races that generate random switches in trade patterns. Furthermore, rigid wages in Europe give rise to unemployment in that region. Using the model, I study the role of international trade and labor market institutions in mediating local and global shocks. I also investigate the factors behind the stylized trends in Europe and the US: rising European unemployment, rising wage inequality (especially for the US), skill upgrading, and rising R&D intensities. I find that technology shocks coupled with institutional response from the European labor markets can generate results consistent with the trends.

Over the past three decades, there have been major changes in wage inequality and unemployment patterns in the advanced countries. In the US, between 1973 and 1989, the wage premium of college graduates over high-school graduates has increased by 20 percentage points (Freeman and Katz, 1995, p. 34). The widened wage gap persisted in the 1990s with the exception of a modest narrowing in the late 90s. In Europe, on average, there has been a mild increase in the skill premium with some cross-country heterogeneity (see Freeman and Katz, 1995; Nickell and Bell, 1995). However, all European countries have experienced a secular and substantial increase in the rate of unemployment, mostly concentrated among the less-skilled workers. In 1973, the average rate of unemployment in OECD Europe was 2.21 percent. By 1995 it reached 9.32 percent and remained high since then.1 In search of explanations, researchers have focused on two culprits: increased international trade and skill-biased technological change (henceforth SBTC) (see Bhagwati, 2002; Acemoglu, 2002 for recent overviews). Most economists have considered the issue using the perfect-competition-based Heckscher–Ohlin (H–O) model of trade. In this model, for increased trade to have an impact on relative wages, the Stolper–Samuelson mechanism must come into play. This mechanism implies that the prices of less-skilled labor-intensive goods must go down in order to generate a reduction in the relative wage of less-skilled labor. Moreover, according to the H–O model if trade is the main cause of the rising relative wage, then within each industry firms should substitute away from skilled labor towards less-skilled labor – the factor which becomes relatively cheaper. Nevertheless, there is no strong evidence that the prices of less-skilled intensive goods have been falling. Moreover, the relative employment of skilled workers within industries has risen in both the US and Europe.2 As a result, many scholars have rejected the trade explanation in favor of SBTC. However, there is an emerging literature that seriously challenges the above deduction by considering trade models based on monopolistic and oligopolistic market structures.3Dinopoulos and Segerstrom (1999a) and Şener (2001) consider two-country endogenous growth models in which firms can exercise temporary monopoly power if successful in their R&D efforts. These models propose a Schumpeterian version of the Stolper–Samuelson mechanism which establishes a positive link between the relative price of innovation and the relative wage of skilled labor. Neary, 2002a and Neary, 2002b constructs an oligopolistic general equilibrium model of trade in which higher levels of strategic investment by incumbent firms is associated with rising skill premium. In addition, Dinopoulos et al. (2001) construct a monopolistic competition model of intra-industry trade to study the effects of trade on relative wages. In these papers trade can cause rising wage inequality while conforming with empirical regularities. In all of the papers cited above, the trading countries exhibit the same institutional structures in their labor markets. Consequently, these models cannot account for the observed differences in wage and unemployment patterns between Europe and the US, which have often been attributed to institutional differences. The popular view is that in Europe strong unions, high minimum wages and generous benefits have propped up less-skilled wages at the expense of employment, whereas in the US the relatively flexible labor market has allowed less-skilled wages to fall and thereby created jobs.4 Recently, Krugman (1995) and Davis (1998) have incorporated institutional differences in H–O-type trade models by considering a structure in which one country (Europe) implements a rigid-wage policy and another country (America) has flexible wages. Using this asymmetric structure, a number of studies have examined the role of labor market institutions in mediating shocks through the web of trade. Nevertheless, these studies have exclusively used the static H–O setting and ignored the growth and R&D dynamics of the global economy.5 This is quite contrary to the recent findings of the empirical trade literature, which provides strong support for the role of innovative activity in explaining trade flows between advanced countries.6 It also contrasts with the recent findings of the empirical labor literature, which establishes a strong and robust relationship between skill upgrading and R&D intensity for high-income OECD countries.7Table 1 shows the trends in skill upgrading (measured either by the employment share or wage bill share of non-production workers) and R&D intensities (R&D spending divided by value added) for a select group of European countries and the US. The values indicate that between 1973 and 1988 skill upgrading has been a pervasive phenomenon in advanced countries that has coincided with sustained increases in R&D intensities.Motivated by these considerations, I construct a model of trade in a growth-theoretic context to investigate the recent labor market trends. In particular, I have three objectives in mind. The first is to use a dynamic trade model with endogenous R&D instead of a static perfect-competition-based H–O framework. The second is to capture the differences between Europe and the US in a unified framework where the two countries are integrated via trade and yet differ in their labor market institutions. The third is to analyze the role of asymmetric institutions in mediating local (i.e., affecting only one country) and global shocks (i.e., affecting the two countries simultaneously).8 The setting is a dynamic two-country model of trade. In each country, there are two types of labor: skilled and less-skilled, and two types of activities: R&D and final goods manufacturing. The countries differ in terms of their wage setting behavior. Following the terminology of Krugman (1995), and Davis (1998), I call the country with flexible wages America and the country with a rigid less-skilled wage Europe. In America, due to flexible wages, there is no unemployment whereas in Europe, the rigid less-skilled wage is binding and less-skilled workers are exposed to unemployment. In each country, individuals respond to increases in the skill premium by undertaking training. Hence, the ratio of skilled to less-skilled labor is determined endogenously. On the production side, the world economy consists of a continuum of industries. In each industry, entrepreneur firms of both countries participate in sequential R&D races to innovate higher-quality products. The winner of an R&D race sustains temporary monopoly power in the global market and sells its product to both America and Europe. Hence, successful innovation by an American (European) entrepreneur firm implies industry leadership for America (Europe). In equilibrium, each country commands leadership in a constant fraction of industries. The dynamic trade model differs from the standard H–O model in two major aspects. First, in each industry the pattern of trade is determined by whether a European or an American entrepreneur becomes a quality leader. Since the arrival of innovations is governed by a stochastic process, the model allows for random switches in trade patterns for each industry. Second, in my model, two countries participate in a zero-sum sequential competition to capture leadership in a fixed continuum of industries, i.e., one country's industry leadership comes at the expense of the other country. These ingredients capture the role of innovation in determining trade patterns (see footnote 6) and the nature of within-industry competition among the advanced countries. Using this analytical model, I investigate the factors behind the stylized long-run trends in Europe and the US: rising unemployment in Europe, rising relative wages for skilled workers (with a larger increase in America), skill-upgrading, and rising R&D intensities. As explanations, I consider local and global technology shocks, institutional adjustments and demographic shocks. I find that there is no single shock or institutional adjustment that can explain the stylized trends; instead, a combination of factors can be at work. In this vein, I identify two combinations. The first is global technological change in R&D (henceforth TCRD) coupled with an institutional response from Europe that props up the relative wages of less-skilled workers. This combination generates results that are consistent with the observed trends across the board. The second is global SBTC in manufacturing coupled with an institutional response from Europe again aimed at maintaining wage equality. This combination generates the same results except for one caveat: it leads to a reduction in R&D intensity in both regions. Hence, a positive shock to R&D technology appears to be a necessary condition to fully explain the recent trends. What about the role of trade in affecting labor market outcomes? I find that a movement from autarky to free trade raises European unemployment and reduces wage inequality in both regions. Thus, opening up of trade does not seem to fully capture the stylized trends nor does it appear to be a part of a reasonable shock–response combination.9 However, in the presence of asymmetric labor market institutions, economic integration plays a crucial role in mediating shocks. In an integrated world, the rigid European economy is subject to more volatility compared to the flexible American economy. In all instances, shocks originating from America exert a sizeable influence on the European economy. On the other hand, America is almost immune to shocks originating from Europe—except for shocks that directly affect European manufacturing costs and thus final goods prices.10 The paper also complements the recent literature which endogenizes the direction of technological change. In a series of papers Acemoglu, 1998 and Acemoglu, 2003a finds that increased trade or an exogenous increase in the supply of skilled labor can endogenously stimulate SBTC and thereby raise the skill premium. In a closed economy model with search unemployment, Acemoglu (2003b) shows that the wage compression efforts of institutions can discourage SBTC and thus explain the divergent wage and unemployment patterns between US and Europe. Using a two-country model with endogenous innovation, Thoenig and Verdier (2003) find that economic integration can increase the threat of leapfrogging and force more firms to pursue skilled-labor-intensive production technologies, leading to endogenous SBTC. In an alternative model, Beaudry and Green (2003) show that introduction of a new technology can lead to rising skill premium if there is under-accumulation of physical capital relative to human capital. Using data from Germany and the US, they find that differences in the evolution of capital intensities between countries can explain the divergent patterns of skill premia. In this paper, I depart from the above SBTC literature on two accounts. First, I consider asymmetric labor market institutions and emphasize the role of trade linkages. In the above-cited papers either closed economy models or multi-country models with symmetric labor market institutions are considered. Second, I model the skill intensity differentials between R&D and manufacturing activities and uncover implications of TCRD and SBTC for labor markets. I argue that TCRD can be an alternative to SBTC in explaining the stylized labor market and R&D intensity trends. The rest of the paper consists of four parts. Section 2 develops the model. Section 3 presents the comparative steady-state results. Section 4 presents the numerical simulations. Section 5 provides an overall discussion and Section 6 concludes the paper. Proofs of all propositions and other technical details are relegated to Appendices (available at http://www1.union.edu/~senerm/ and also upon request).

In this paper, I constructed a dynamic two-country model of trade with endogenous R&D. The model considered the institutional differences between Europe and the US in a unified framework. The paper has three main results. First, it proposes TCRD as an alternative explanation to SBTC in the trade/wages/unemployment debate. Second, the paper implies that an economy with a flexible labor market can more easily pass on the effects of shocks to an economy with a rigid labor market. Third, the model uncovers the role of international trade and the related mechanisms in transmitting local shocks. The model can help identify the forces that can generate the stylized trends; however, many extensions can be suggested to improve the match between observed trends and numerical simulations. Incorporating heterogeneities across Europe (as advocated by Nickell, 1997) and also adding a North–South dimension to the model (as recently done by Grieben, 2004) can be the first steps along these lines. In addition, one can incorporate search frictions into the asymmetric labor markets setting and study wage inequality and unemployment (see for instance Acemoglu, 2003b; Arnold, 2002; Şener, 2001 for models with search unemployment). Finally, considering indicators of R&D productivity and economic integration as explanatory variables in the empirical framework of Blanchard and Wolfers (2000) can be a fruitful avenue for empirical research.