تقارب فن آوری، تحقیق و توسعه، تجارت و رشد بهره وری
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|11369||2005||33 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : European Economic Review, Volume 49, Issue 3, April 2005, Pages 775–807
This paper analyses productivity growth in a panel of 14 United Kingdom manufacturing industries since 1970. Innovation and technology transfer provide two potential sources of productivity growth for a country behind the technological frontier. We examine the roles played by research and development (R&D), international trade, and human capital in stimulating each source of productivity growth. Technology transfer is statistically significant and quantitatively important. While R&D raises rates of innovation, international trade enhances the speed of technology transfer. Human capital primarily affects output through private rates of return (captured in our index of labour quality) rather than measured TFP.
‘It may be seriously argued that, historically, European receptivity to new technologies, and the capacity to assimilate them whatever their origin, has been as important as inventiveness itself.’1 A number of authors have emphasised the transfer of technology from leader to follower countries as an important source of economic growth. Rosenberg (1982) notes that three of the great European technical developments – printing, gunpowder and the compass – are the result of technology transfer. Less prosaic examples include the development of the crucible steel industry in early 19th Century France based on British technology, the diffusion of mass production techniques for motor car manufacture from the US to Europe during the early 20th Century, and the development of the Japanese semi-conductor industry.2 This paper evaluates the role of technology transfer in explaining productivity growth at the industry-level in the United Kingdom since 1970. We present an empirical framework in which innovation and technology transfer provide two sources of productivity growth for an economy behind the technological frontier. The difference in levels of total factor productivity (TFP) between the United Kingdom and a frontier country (the United States) is used as a direct measure of the potential for technology transfer.3 This approach allows for knowledge spillovers from both formal research and development (R&D) and the informal activities not captured in R&D statistics that a wide range of empirical evidence suggests are important for productivity growth.4 ‘Technology transfer’ is used in the paper to refer to convergence in technical efficiency within individual industries over time. The analysis controls for both observable and unobservable characteristics that determine whether and at what speed technology transfer occurs. We consider the roles played by R&D, international trade, and human capital. We examine whether each variable has a direct effect on rates of TFP growth (innovation) and whether the variable's effect on TFP growth depends on distance behind the technological frontier (technology transfer). The use of panel data on industries over time enables us to examine the disaggregated forces underlying country-level growth performance, while at the same time controlling for unobserved heterogeneity in the sources of productivity growth. Existing work on R&D knowledge spillovers often assumes that technology transfer occurs through a specific mechanism such as international trade. An advantage of our approach is that we explicitly test whether technology transfer occurs through international trade against the alternatives that its pace is determined by domestic ‘absorptive capacity’ (in the form of human capital and R&D investments) and that it proceeds autonomously (independently of the economic variables considered).5 In steady-state, the level of productivity in non-frontier countries lies an equilibrium distance behind the frontier, such that productivity growth from innovation and technology transfer exactly equals productivity growth from innovation alone in the frontier. The analysis thus sheds light on the existence and determinants of long-run differences in productivity across countries in individual industries. 6 Our main results are as follows. First, we find an important role for technology transfer as a source of productivity growth in UK manufacturing. This result is robust across a wide range of econometric specifications and to the use of a number of different measures of TFP. Other things equal, the greater the gap in levels of technical efficiency between the United States and the United Kingdom in a manufacturing industry, the faster the rate of UK productivity growth. Second, there is a positive direct effect of R&D on productivity growth through rates of innovation. This finding is consistent with both the endogenous growth literature and the micro-econometric literature on R&D and productivity. The result is again robust across a wide range of specifications, to the use of different measures of R&D and TFP, and with different lag structures. Third, we find that increased international trade raises rates of UK productivity growth through technology transfer but not innovation. Our measure of international trade is the ratio of a UK industry's imports from the whole world to gross output. A problem in the literature on trade and growth is the potential endogeneity of international trade. Another feature of our approach is the use of instruments that capture exogenous variation in the degree of international integration in individual industries over time. Estimating the model using instrumental variables strengthens the finding of trade-based technology transfer. The instruments are highly statistically significant in the first-stage regression, and we present evidence that the identifying assumptions underlying the instrumental variables estimation are satisfied. Fourth, our preferred measure of TFP controls for variation across countries and industries in the skill composition of the workforce. Numbers of skilled and unskilled workers are weighted by their respective shares of the wage bill. In so far as any increased productivity of skilled workers is reflected in their wages (a private rate of return), it will already be captured in our measure of TFP. We present evidence that, once one controls for the direct effect of human capital on output through private rates of return, there is no evidence of an additional effect through externalities. The paper relates to two main strands of existing literature. First, a body of empirical work has examined the relationship between R&D and productivity growth at the firm and industry-level. Classic references include Griliches (1980), Griliches and Lichtenberg (1984), and Mansfield (1980).7 The conventional approach regresses TFP growth on measures of R&D activity. Microeconomic foundations are supplied by the theoretical literature on endogenous innovation and growth,8 and a positive and statistically significant estimated coefficient provides evidence of R&D-based innovation. One branch of this first empirical literature has examined R&D knowledge spillovers across industries, countries, and regions.9 Foreign R&D knowledge is typically found to be an important source of productivity growth, although there remains much debate concerning the mechanisms through which this occurs. A number of studies assume that international trade flows are the conduit. However, this assumption has recently been questioned by Keller (1998). The use of the technology gap as a direct measure of the potential for technology transfer in our approach allows for knowledge spillovers from both formal R&D investments and informal sources of productivity growth. Rather than assuming a particular mechanism through which knowledge is transferred, we test the statistical significance of various economic mechanisms that have been proposed, while also allowing technology transfer to occur independently of international trade and the other economic variables considered. Second, the use of a direct measure of distance from the technological frontier means that the analysis relates to both the literature on the measurement of TFP across countries, industries and time and to work on productivity convergence.10 By combining data from the Census of Production in the United Kingdom and United States with that from a number of other sources, we obtain a rich source of industry-level information, with which we are able to make a number of adjustments to standard TFP measures. For example, our analysis controls for variation across countries and industries in hours worked, the skill composition of the workforce, capacity utilization, and manufacturing prices as captured in industry-specific purchasing power parities (PPPs). In contrast to a number of existing papers which focus on a Cobb–Douglas production technology, we employ extremely general TFP measures consistent with any constant returns to scale production technology. A large number of papers have examined convergence of income per capita and productivity at the country-level.11 Typically, convergence is only observed after controlling for a variety of determinants of long-run income per capita/productivity levels (‘conditional convergence’), and several country-level studies emphasize the idea that convergence is dependent on the promotion of ‘absorptive capacity.’ Thus, Benhabib and Spiegel (1994) argue for important effects from human capital, while Abramovitz (1986) stresses ‘social capability’. A companion literature has examined productivity convergence at the industry-level.12 Our empirical framework is consistent with conditional productivity convergence, which emerges as an implication of a long-run cointegrating relationship between TFP levels in the United Kingdom and United States. The econometric equation that we estimate is an equilibrium correction model (ECM) representation of this long-run cointegrating relationship. Our analysis also explicitly incorporates a role for R&D, human capital, and international trade in determining productivity growth. We test empirically whether each variable affects productivity growth through innovation and/or technology transfer. The paper is structured as follows. Section 2 introduces the theoretical framework from which our main econometric equation is derived. The theoretical discussion provides structure for our empirical work and yields implications for the determinants of long-run relative productivity levels. Section 3 analyses the evolution of relative levels of TFP in the manufacturing sectors of the United Kingdom and United States since 1970. Section 4 estimates the econometric relationship between TFP growth, distance from the technological frontier, R&D, human capital, and international trade. Section 5 concludes.
نتیجه گیری انگلیسی
This paper examines the determinants of productivity growth at the industry-level in the United Kingdom since 1970. We began by outlining innovation and technology transfer as two sources of productivity growth for a country behind the technological frontier. The difference in levels of total factor productivity (TFP) between the United Kingdom and a frontier country (the United States) is used as a direct measure of the potential for technology transfer, and we employ extremely general measures of TFP that control for a variety of sources of potential measurement error suggested in the literature on TFP measurement. Our framework allows for knowledge spillovers from both formal research and development (R&D) and the informal activities not captured in R&D statistics, while also controlling for observable and unobservable characteristics of countries and industries that affect the potential for technology transfer. We find a positive and statistically significant effect of distance from the technological frontier on rates of productivity growth. Other things equal, the further an industry lies behind the technological frontier, the higher its rate of TFP growth. This finding is robust across a wide range of econometric specifications, to instrumenting distance from the technological frontier, to the use of alternative TFP measures, and to the inclusion of a series of control variables. A key advantage of our approach is that we are able to explicitly test whether a number of variables proposed as determinants of productivity growth (such as R&D, international trade, and human capital) affect productivity growth directly (through rates of innovation) or indirectly through distance from the technological frontier (technology transfer). Unlike much of the literature, which assumes that technology transfer occurs through a particular mechanism such as international trade, we test econometrically whether international trade plays a role against the alternatives that its pace is determined by domestic ‘absorptive capacity’ (in the form of human capital and R&D investments) and that it proceeds autonomously (independently of the economic variables considered). We find that R&D affects rates of UK productivity growth through innovation, while international trade facilitates the transfer of technology. These results are again robust across a wide range of econometric specifications and to instrumenting international trade in the UK with a measure based on OECD trade that captures exogenous variation in the degree of international integration in individual industries over time. Our preferred measure of TFP weights numbers of production and non-production workers in a country-industry by their shares of the wage bill. Once one controls for any increased productivity of non-production workers reflected in their wages (a private rate of return), we find no evidence of an additional effect of human capital on productivity growth (no evidence of externalities). This is consistent with the empirical growth literature's finding of a positive effect of human capital on growth, but suggests that this effect corresponds to a private rate of return. Taken together, our results emphasize the importance of technology transfer for countries behind the technological frontier. We find relatively rapid convergence of countries towards steady-state levels of relative TFP within individual industries, suggesting that observed cross-country productivity differences largely reflect systematic determinants, including fixed characteristics of countries and industries (such as location and institutions captured in the fixed effect), investments in R&D, and openness to international trade.