سرمایه خارجی، سرمایه انسانی، و بهره وری : تجزیه و تحلیل مرزی تصادفی برای کشورهای در حال توسعه

We use stochastic Frontier analysis to study which of the three channels of technology diffusion, foreign direct investment (FDI), imports of machinery and equipment, or imports of research and development (R&D) expenditures, affect the total factor productivity of developing countries. We also analyze whether a developing country’s openness to technology diffusion is affected by their existing levels of human capital. We find that FDI, imported capital goods, and imported R&D are all important channels for improving efficiency, as is human capital accumulation. However, the positive effect of FDI, imported capital goods, and imported R&D depends crucially on the level of accumulated human capital. In addition, we find that in the process of technology diffusion, the impact of formal education is more important for imported R&D than it is for imported capital and FDI, whereas the opposite is true for learning by doing, which is found to be more important for knowledge diffusion through FDI and imported capital.

The debate on whether economies that are open to the rest of the world see faster economic growth goes back to the time of Adam Smith and continues even today. One way that international trade can help in the process of economic growth is by transferring the benefits of technology across borders. Seminal papers by Grossman and Helpman, 1991 and Barro and Sala-i-Martin, 1995 argued that larger trade implies greater openness which helps in an economy’s adoption of more efficient techniques of production. This technology transfer or technology diffusion leads to faster growth of total factor productivity and, hence, faster growth in per capita income levels. Moreover, the “new” growth theory of Romer, 1990 and Barro, 1997 has human capital playing an important role in growth because human capital can help in explaining an economy’s capacity to absorb new technologies ( Abromovitz, 1986, Cohen and Levinthal, 1989, Kneller, 2005 and Kneller and Stevens, 2006). Recent empirical research by Eaton and Kortum (2001) showed that only a few research and development (R&D) intensive countries produce most of the world’s capital; the rest of the world just imports this equipment that embodies new technology. Consequently, a country’s productivity “depends on its access to capital goods from around the world and its willingness and ability to use them (p. 1196).” This is also collaborated by Caselli and Coleman (2001) who in their studies show that most countries acquire embodied technologies through capital imports from world technological leaders. In the context of developing countries this is especially important because growth in total factor productivity can be affected by the extent of their adoption and implementation of new technologies that are already in use in developed countries. In addition, how well they use these technologies can be affected by the level of human capital in these countries.1 For example, Benhabib and Spiegel’s (1994) empirical study suggests that human capital plays a role in economic growth by helping in the adoption of technology from abroad and in creating the appropriate domestic technology. In response to the question of how technology diffusion affects economic growth, there has been an emerging empirical literature examining the nexus between the channels of technology diffusion, openness, and human capital in promoting economic growth. The evidence on this issue is mixed as seen in the differing conclusions of Miller and Upadhyay, 2000, Miller and Upadhyay, 2002 and Olofsdotter, 1998.2 However, a consistent feature of all the empirical studies on this issue is the use of a cross-country regressions framework on a sample of developed and developing countries. Cross-country regressions cannot control for the unobservable heterogeneity which can arise, for example, from the different institutions in the various countries. Furthermore, Rodriguez (2006) argued that policy analysis within the growth-regression framework can carry considerable risks from the misspecification bias that come from using such a specification when it is not valid. In order to avoid the pitfalls of these cross-regression studies we use an alternative empirical methodology, stochastic Frontier analysis, to answer our research question of whether a developing country’s openness to technology diffusion is affected by their existing levels of human capital.3 In our empirical framework, technology diffusion can take place through either flows of foreign direct investment, through imports of foreign capital, or through imported R&D. We estimate a stochastic production Frontier for developing countries to determine the three channels through which greater openness diffuses technology—the foreign direct investment (FDI) channel, the imported capital channel, and the imported R&D channel. In addition, by using this empirical methodology we can also study the interaction between these three channels by which greater openness diffuses technology and the existing level of human capital in the country. There have been earlier studies which have looked at the impact of each of these three channels on production efficiency or some combination of these channels. However, to the best of our knowledge, there is no study which has investigated all three channels through which openness diffuses technology for developing countries simultaneously. By investigating the interaction between all three channels of openness (FDI, imported capital goods, and imported R&D) and human capital we can determine what has been a source of growth in total factor productivity in developing countries. We apply stochastic Frontier analysis in a macroeconomics context, where countries are producers of output (i.e., GDP) given inputs (e.g., capital, labor, and technology), to empirically examine the role of foreign capital in the process of technology diffusion in developing countries. This closely matches the concept of Frontier technology and the innovation of technology found in growth theory (Acemoglu et al., 2006, Aghion et al., 2001 and Aghion et al., 1997). In this context, countries can be thought of as operating either on or within the Frontier, with the distance from the Frontier reflecting inefficiency. Over time, a country can reduce its inefficiencies and reach the Frontier or the Frontier itself can shift outwards over time, indicating technical progress.4 Moreover, a country can move along the Frontier by changing inputs. Thus, output growth can be seen as being made up of three components: efficiency change, technical change, and input change with the first two components being the “productivity change” (Koop, Osiewalski, & Steel, 2000a). We find two important results. First, we find that FDI, imported capital goods, and imported R&D are important channels for improving efficiency, as is human capital accumulation. However, there is an important difference between the three channels in that knowledge diffused through human capital is more general, or disembodied, than knowledge diffused from FDI and imported R&D, which tend to be more embodied. Second, we find that the positive effects of FDI, imported capital goods, and imported R&D depend crucially on the level of accumulated human capital. Moreover, we find that in the process of technology diffusion the impact of human capital in the form of accumulated formal education is more important for imported R&D than it is for imported capital and FDI. However, human capital that is more specific, like learning by doing, is more important for knowledge diffusion through imported capital and FDI. The remainder of the paper is organized as follows: In Section 2, we discuss the role of technology diffusion, openness, and human capital in developing countries. We then construct the empirical model in Section 3 and present the estimation strategy in Section 4. In Section 5, we discuss the empirical results. We conclude in Section 6.

According to the literature, the main channels through which the foreign technology diffusion occurs are through foreign direct investment, imports of machinery and equipment and imported R&D. Our paper contributes to the literature on the diffusion and absorption of technology by utilizing stochastic Frontier analysis and investigating all four potential sources of inefficiency, differences in human capital, differences in FDI, differences in machinery and equipment, and differences in imported R&D, for 57 developing countries, over the period 1981–2000. More importantly, this study differs from previous studies on similar topics by investigating all the three main channels which affect technology-catch up toward the Frontier simultaneously. We find strong evidence that developing countries differ with respect to the efficiency with which they use Frontier technology. We find that human capital, imported machinery and equipment, and imported R&D play a significant and quantitatively important role in explaining these differences in efficiency. Moreover, there is clear evidence that human capital affects efficiency both directly and indirectly through its effect on R&D, imported machinery and equipment, and FDI. While the growth literature underlines the important influence of human capital externalities on productivity growth, the empirical evidence on this issue is still incomplete and mixed. This research adds to the previous evidence on the human capital externalities by explicitly using various measures of human capital to better disentangle the differential impacts of the different human capital indicators on productivity. By using the Barro–Lee education data we find that the effect of R&D becomes stronger with the increase of human capital accumulation at higher levels of education and that the effect of R&D becomes very insignificant at very low levels of human capital. Another important measure of human capital is learning by doing which is an alternate source of human capital accumulation since workers can gain experience and improve productivity through working (Lucas, 1988). We find that FDI and imported capital have increasingly positive impacts on efficiency in the presence of learning by doing. This supports the view of complementarities between disembodied knowledge of multinational firms and the absorptive capacity in host countries. Given that there is little agreement on whether developing countries would do well in inviting foreign direct investment in a non-discriminatory manner, whether they should reduce industrial tariffs in particular for capital goods, and what role intellectual property rights plays in development, this paper has some significant policy relevance. The results in the paper lead us to conclude that the positive effects of multinational firms on the productivity of developing countries are not direct but through human capital. The absorptive capacity of each country provides a useful explanation for the differences in productivity. The evidence here suggests that developing countries that pursue relatively closed trade policies will not gain the benefits from technology implicit in international trade. This supports Hoekman, Maskus, and Saggi (2005) policy recommendations that argue for liberal trade policies for all countries. Furthermore, the governments in developing countries should facilitate the process of international technology transfer by encouraging the establishment of the necessary infrastructure and by providing incentives to support the development of domestic innovative capabilities.