مقایسه بین المللی رشد بهره وری: نقش فن آوری اطلاعات و شیوه های نظارتی

While information technology (IT) is credited with the recent acceleration in productivity in the United States, many other industrial countries have not experienced a pickup in productivity growth. To explain this productivity divergence, we use panel data from 1992 to 1999 for 13 industrial countries and find that this divergence is driven in part by differences in both the production and adoption of information technologies. Based on this finding, we proceed to investigate what factors might play a role in explaining differences in IT adoption. Our results support the view that burdensome regulatory environments and, in particular, regulations affecting labour market practices have impeded the adoption of information technologies and have slowed productivity growth in a number of industrial countries.

In recent years, the United States has experienced an impressive pickup in productivity growth that, for the most part, has not occurred abroad.1 What is remarkable about this development is that the chief explanation for the acceleration in US productivity—the proliferation of computer and information technology (IT)—is available worldwide.2 Indeed, this divergence in productivity is puzzling given that many industrial countries have many of the same structural and institutional characteristics as the United States such as well-educated workforces, an openness to trade, and well-developed legal frameworks. What, then, explains the failure of productivity growth to pick up in many of these other industrial countries? This paper provides evidence for a view that can account for this recent productivity divergence between the United States and other industrial countries by emphasizing the role of regulatory practices in influencing the diffusion of information technologies. According to this view, burdensome regulations in various countries, but particularly in a number of European countries, have impinged on firms' incentive to adopt new technologies, slowing their rate of adoption. The delay in the adoption of new technologies in these countries then translates into slower productivity growth vis-à-vis the United States. With a wide variety of regulatory differences across countries, this view has been expressed in a number of ways. Some proponents focus mainly on product market regulations, specifically those inhibiting competition in goods markets. One variant in this genre, for instance, focuses on regulations raising the cost to entry of new firms. Others focus more specifically on labour market regulations, specifically those that undermine the ability of firms to adjust their workforce in a flexible manner. One proponent of this view is Greenspan (2000), who argues that in order to reap the high returns associated with information technologies, a firm must be able to reorganize its workforce, and employment protection legislation (EPL) interferes with a firm's ability to do so: Europe and Japan have participated in this recent wave of invention and innovation and have full access to the newer technologies. However, they arguably have been slower to apply them. The relatively inflexible and, hence, more costly labour markets of these economies appear to be an important factor. Another proponent advocating the importance of labour market regulations is Feldstein (2001): In Europe, fundamental changes in employment practices, labour markets, and management incentives are necessary to encourage rapid adoption of new technology that can raise productivity while increasing employment. Without such changes, the gap between US and European incomes will continue to widen. Fig. 1 and Fig. 2 show that the Greenspan–Feldstein argument is not lacking in empirical support. Specifically, Fig. 1 plots the change in labour productivity growth between 1991–1995 and 1996–2000 against the change in the ratio of IT expenditures to GDP between 1992 and 1999. As shown there, countries whose IT expenditures rose sharply in the 1990s also experienced a pickup in productivity growth. In contrast, countries where spending on information technologies fell or only picked up marginally did not experience an acceleration in productivity. Fig. 2 plots the ratio of IT expenditures to GDP in 1999 against the index of the level of employment protection in 1998. This index, constructed by the Organization for Economic Cooperation and Development (OECD), attempts to measure how much regulations impede a firm's ability to adjust its workforce. As shown there, countries whose level of employment protection is high are the ones with the smallest GDP share of IT expenditures. In contrast, countries with the smallest level of employment protection are those with the highest GDP share of IT expenditures.Thus, when taken together, the bivariate correlations of Fig. 1 and Fig. 2 clearly support the explanation posited by Greenspan and Feldstein: countries with the greatest employment protection have the lowest adoption of IT investment and the lowest productivity growth. What is not clear, however, is whether those correlations vanish if one controls for other factors that affect productivity performance and IT investment—a question we address in this paper. To do this, we postulate two linear regressions and estimate their parameters with a panel of industrial countries using data over the 1990s. The first relation explains productivity growth in terms of IT expenditures and production relative to GDP controlling for a variety of factors including changes in labour force participation rate, R&D expenditures, and cyclical influences. The second relation explains the ratio of IT expenditures to GDP using indices on regulations affecting employment and startup costs for new firms, while also controlling for factors such as the level of education, the size of the service sector, and the availability of venture capital. Our results are supportive of the view that burdensome regulations can impede the adoption of information technologies and slow productivity growth. In addition, our results corroborate the Greenspan–Feldstein view that labour market practices can hinder the adoption of information technologies. Our findings can also be viewed as providing support for papers such as Caballero and Hammour (1996), Saint-Paul (2002b), and Krusell and Rios-Rull (1996), which demonstrate how labour market regulations impede the adoption of new technologies.3 In these papers, technology is embodied in capital and workers are linked to specific technologies. As capital is replaced due to technological obsolescence, the associated labour is replaced as well by new employees with higher marginal products. In this environment, a higher tax on firing workers can delay the adoption of new technology, resulting in an older capital stock, lower wages, and lower productivity. The rest of this paper proceeds as follows. Section 2 discusses the literature related to our empirical analysis. Section 3 then examines the data on productivity growth, IT production and spending, and the regulatory variables of interest. We also describe the empirical results from our regression analysis. Section 4 then concludes.

Overall, our results support the view that technological adoption and productivity growth are generally impaired in countries with restrictive regulatory environments. Although such regulatory practices may insulate workers from the process of job creation and job destruction (and existing firms from the entry of startups), they also appear to have slowed the rate of adoption of information technologies in some countries. This in turn has played a role in explaining differences in productivity growth across countries in the 1990s. An important remaining question is how long these differences in productivity growth will be sustained. Although regulations may impede the diffusion of technology, productivity growth does not necessarily have to remain persistently higher in the United States than in other countries. In theoretical models such as Caballero and Hammour (1996) and Saint-Paul (2002b), countries with a more burdensome tax on firing workers eventually adopt these new technologies and in the end experience the same rate of productivity growth. However, these models also imply that these countries will have fewer skilled labourers and an older stock of capital on average than countries with a less burdensome tax on firing workers, leading to lower levels of productivity. Thus, labour market reform may be necessary in some countries to reap the full benefits of information technology.