Abstract
Using the panel data from 1981 to 2000 of 17 Organization for Economic Co-operation and Development (OECD) countries, we investigate the role of the trade flow of information technology (IT) commodities in international knowledge spillovers. We use the group mean fully modified ordinary least squares (OLS) estimator, proposed by Pedroni (Citation2000), that allows for greater flexibility in the presence of heterogeneity of cointegrating vectors among countries under panel cointegration. The results of estimation confirm statistically significant positive effects of IT imports on international knowledge spillovers. In contrast, non-IT imports turn out to have negative effects on the productivity of importing countries.
Notes
1 Only two-digit classifications are available.
2 The total number of technology fields is 36.
3 Although some previous studies, including those by CH and LP, include service sectors in their TFP calculations, we focus on the manufacturing sector because productivity in service sectors is not well defined and sufficient data on R&D investment are not available.
4 Seven out of 340 observations are missing.
5Lee (Citation2005b) also uses the GM-FMOLS estimator to show the robust positive effect of international R&D spillovers through the channel of intermediate goods trade.
6Engelbrecht (Citation1997) points out this as a caveat of CH and allows for country-specific spillover effects using interactive dummy variables. Madden et al. (Citation2001) follows the same approach.
7 This selection is implemented automatically by the RATS code that is used for GM-FMOLS estimation. For details, refer to the RATS homepage: http://www.estima.com.
8 Negative spillovers through goods imports were also found for UK manufacturing sectors by Mcvicar (Citation2002).