R&D spillovers, information technology and telecommunications, and productivity in ASIA and the OECD

https://doi.org/10.1016/S0167-6245(00)00024-XGet rights and content

Abstract

This paper examines the role research and development (R&D) plays in technology progress for a sample of OECD and Asian economies from 1980 to 1995. An empirical model is estimated which relates total factor productivity to domestic and foreign R&D activity, trade, and information technology and telecommunications (ITT). Model estimates confirm a positive relationship between national productivity and R&D activity exists in the long run. Further, the benefits of R&D can spillover countries through trade, in particular, trade in ITT equipment.

Introduction

Endogenous growth models emphasise innovation and trade as vehicles for technological spillovers that permit developing countries to catch up to industrialised countries. The creation and accumulation of knowledge can improve sectoral and national productivity through the invention of intermediate goods (or by improving the quality of existing goods), which bring about the more effective use of existing resources (Grossman and Helpman, 1991). The role knowledge plays in technological progress has been the subject of much recent attention in the economic growth literature. Several studies find that the returns from investment in knowledge are positive, and they are greater than returns from investment in equipment, infrastructure and machinery (Griliches, 1988, Griliches, 1994; Nadiri, 1993, Coe and Helpman, 1995). Further, Coe and Helpman (1995) argue that own knowledge enhances a country’s ability to take advantage of innovation and technological advance elsewhere.

Coe and Helpman (1995) extend Grossman and Helpman’s (1991) ‘product variety’ model of innovation to show that national productivity increases with the accumulation of both domestic and foreign knowledge (see also Bayoumi et al., 1996). Employing annual cross-section, time-series data for 21 OECD economies and Israel from 1971 to 1990, Coe and Helpman (1995) demonstrate a significant positive relationship between total factor productivity (TFP) and knowledge, approximated by research and development (R&D) capital stock, exists. Further, the benefits of R&D can also spread to (spillover) other countries through trade when measured by the share of aggregate imports to gross domestic product (GDP). The predicted convergence occurs only when knowledge spills over perfectly between countries. This finding has important implications for national trade liberalisation and economic integration policy. For instance, Feenstra (1996) shows that when knowledge spills over imperfectly between countries, small open economies may experience slower rates of economic growth rates after trade liberalisation and integration.

Application of the above findings to particular national circumstances is somewhat problematic as received empirical evidence usually relate to OECD economies and aggregate imports are nominated as the sole channel for the transmission of R&D spillovers internationally. Coe et al. (1997) extend their sample to 77 developing economies, and estimate an equation that relates TFP to foreign R&D capital, imports of machinery and equipment relative to GDP, and educational attainment.1 This TFP equation does not include an argument for domestic R&D for developing countries since these data are scarce. Model estimates indicate that a 1% increase in the foreign R&D stocks of industrialised countries (in the ‘North’) raises output of the developing countries (in the ‘South’) by 0.06%.2 However, the assumption of negligible domestic R&D in developing countries is unacceptable for some ‘high income’ Asian countries contained in their sample. Madden and Savage (2000) address this potential source of error by estimating a model that relates TFP to both domestic and foreign R&D activity for a sample of OECD and Asian nations. They find that TFP growth in Chinese Taipei, India, Indonesia, Singapore, South Korea and Thailand is augmented by investment in own R&D.

Lichtenberg and Van Pottelsberghe de la Potterie, 1996, Engelbrecht, 1997a, Engelbrecht, 1997b and Madden and Savage (2000) suggest that Coe and Helpman (1995) and Coe et al. (1997) place too much emphasis on aggregate openness and education as channels for international knowledge transfer. For instance, Lichtenberg and Van Pottelsberghe de la Potterie, 1996, Engelbrecht, 1997a identify training received abroad, foreign direct investment, electronic information transfer, and imports of books and journals as potentially important transmission mechanisms for international R&D spillovers. Madden and Savage (2000) argue that information technology and telecommunications (ITT) is an important source of international knowledge transfer in an emerging global information economy. International trade in ITT equipment and services generates direct productivity benefits through lower transaction costs and improved marketing information, and indirect benefits due to accelerated information and knowledge diffusion across borders (Jussawalla and Lamberton, 1982, Antonelli, 1991). As such, ITT and trade policy are becoming a priority for many governments and international agencies endeavouring to improve national productivity and economic growth (European Bank for Reconstruction and Development, 1995, OECD, 1996a, Spiller and Cardilli, 1997).

This study examines the role that R&D activity plays in technological progress for a sample of OECD and Asian economies for the period from 1980 through 1995 using annual data. So as to assess the importance of alternative potential sources of TFP growth a series of models are developed from the Coe and Helpman (1995) specification. A composite model relates TFP to domestic and foreign R&D activity, trade, and ITT. Model estimates are used to calculate the TFP elasticity with respect to OECD and Asian domestic R&D, respectively, and individual country foreign R&D elasticity’s of TFP. The TFP equation and estimation method employed here differs from those of previous studies. First, the sample includes Asian measures of domestic R&D expenditures. Inclusion of Asian data allows further examination of the Coe et al. (1997) hypothesis that developing economies with low own R&D benefit from OECD R&D activity. Measures of ITT trade are also included in the TFP equation to allow alternative channels for the international transfer of R&D. Finally, the Im et al. (IPS 1997) group mean panel unit root test is used to establish the order of integration and cointegration of these panel data. The paper is organised as follows. Section 2 describes the econometric method used to estimate TFP elasticities. Data used in estimation are presented in Section 3. Section 4 reports estimation results, and elasticity calculations for TFP with respect to R&D. Section 5 presents conclusions.

Section snippets

Econometric method

Following Coe and Helpman (1995), the basic TFP equation used to assess the importance of R&D spillovers is: log TFPit0i+α1 log DRDit2 log FRDit+eitwhere i=1, …, n denotes countries and t=1, …, T the year, DRD is domestic R&D capital stock, FRD is foreign R&D capital (measured as country i’s bilateral import share weighted-average of the domestic R&D capital stocks of its trading partners), the α’s are unknown parameters to be estimated, and e is a white noise error term. The parameter

Data

TFP , , are estimated on a sample of: seven OECD G7 economies, Canada, France, Germany, Italy, Japan, United Kingdom (UK), United States (US); eight OECD non-G7 industrialised economies, Australia, Denmark, Finland, Ireland, Netherlands, Norway, Spain, and Sweden; and five Asian countries, India, Indonesia, Singapore, South Korea and Thailand. Annual data from 1980 to 1995 are collected for exchange rates, fixed capital, GDP, labour, aggregate imports, ITT equipment and service imports,

TFP equation estimation

The IPS group mean panel unit root tests indicate that log TFP, log DRD, log FRD, M/Y, COM and TEL are difference-stationary.8 To test for cointegration the IPS panel unit root test is applied to the residuals formed from the estimation of country-specific TFP , , , respectively.9

Conclusions

This paper examines the role R&D plays in technology progress for a sample of 15 OECD and five Asian economies from 1980 to 1995. An empirical model is estimated which relates TFP to domestic and foreign R&D activity, trade, and ITT. Panel unit root tests of TFP equation residuals indicate the existence of a long-run cointegrating relationship between national productivity and R&D activity. Further, TFP equation estimation shows that domestic R&D is an important source of TFP growth and that

Acknowledgements

Presented at the Thirteenth Biennial Conference of the International Telecommunications Society, Buenos Aires, Argentina, July 2000. Data support from the Indonesian Ministry of Industry and Trade, the Korean Ministry of Science and Technology, Singapore National Science and Technology Board and the Thai Office of Policy and Planning are gratefully acknowledged. The paper has also benefited from helpful comments from Terry Curtis, Hans-Jürgen Engelbrecht and Susan Simon. The usual disclaimer

References (44)

  • D Dickey et al.

    Likelihood ratio statistics for autoregressive time series with a unit root

    Econometrica

    (1981)
  • Edmond, C., 1998. Total technological productivity and panel data approaches to R&D spillovers. Department of Economics...
  • H Engelbrecht

    International R&D spillovers amongst OECD economies

    Applied Economics Letters

    (1997)
  • R Engle et al.

    Cointegration and error correction: Representation, estimation, testing

    Econometrica

    (1987)
  • Transition Report Update

    (1995)
  • S Greenstein et al.

    Modern telecommunications infrastructure and economic activity: An empirical investigation

    Industrial and Corporate Change

    (1995)
  • Z Griliches

    Productivity, R&D, and the data constraint

    American Economic Review

    (1994)
  • Z Griliches

    Patent statistics as economic indicators: A survey

    Journal of Economic Literature

    (1990)
  • Z Griliches

    Productivity puzzles and R&D: Another non-explanation

    Journal of Economic Perspectives

    (1988)
  • G Grossman et al.

    Innovation and Growth in the Global Economy

    (1991)
  • Im, K., Pesaran, M., and Shin, Y., 1997. Testing for unit roots in heterogeneous panels. Department of Applied...
  • ILO, various issues. Yearbook of Labour Statistics. ILO,...
  • Cited by (22)

    • The modern drivers of productivity

      2015, Research Policy
      Citation Excerpt :

      The international diffusion of ICT has also changed the value chain of most industries and expanded the available stock of entrepreneurial and scientific talents. These factors have contributed to increase both production efficiency and research engagement (Madden and Savage, 2000; Lee, 2009). More generally, advances in information technology are claimed to shift the innovation possibility frontier, rather than directly shifting the production frontier.4

    • E-commerce, r&d, and productivity: Firm-level evidence from taiwan

      2013, Information Economics and Policy
      Citation Excerpt :

      Both self-R&D and the R&D of technological neighbors naturally influence firm productivity through the spillover effect (Jaffe, 1986). This positive R&D spillover effect shows up in numerous empirical studies, e.g., Doring and Schnellenbach (2006) and Madden and Savage (2000). The adoption of information technology similarly exhibits network externality brought about by more users and thus benefits users, e.g., fax machine (Economides and Himmelberg, 1995) and cellular phone (Jha and Majumdar, 1999).

    • The changing pattern of industrial technology linkage structure of Korea: Did the ICT industry play a role in the 1980s and 1990s?

      2009, Technological Forecasting and Social Change
      Citation Excerpt :

      The ICT has carried out an important role to tow the overall economy; especially it has been a breakthrough for easing the currency crisis of Korea in the year 1997. The production and productivity effects of ICT on other industries and economy have been observed and empirically examined in many other countries such as New Zealand [13], Finland [14], UK [15], Italy [16], several OECD and Asian countries [17], etc. The OECD [18] has more emphasized that ICT has been, furthermore, an important driver of growth and productivity in its member countries.

    • Embodied and disembodied international spillovers of R&D in OECD manufacturing industries

      2004, Technovation
      Citation Excerpt :

      The results suggest that the R&D spillovers from the industrialized countries are substantial. For the information technology and telecommunications industry, Madden and Savage (2000) examined the role of R&D in technology progress using the data for Asian and OECD economies over the period 1980–1995. They found a positive relationship between productivity change and the spillovers through trade.

    View all citing articles on Scopus
    View full text