Hu (2015) aims to provide an overview of the technological progress that East Asia has achieved over the past decades. The paper adopts South Korea as a case study to illustrate a nonlinear process of technological and economic catching up, and discusses the determination of technological innovation using a large sample of countries. Technological advancement in an economy depends on both innovation of new technology as well as imitation of the technology produced elsewhere. For many East Asian economies lagging behind in the technological ladder, the imitation and adaptation of advanced technologies provide good opportunities to catch up to more advanced countries. In contrast, for the advanced East Asian economies that have already narrowed the technology gap, technology creation becomes more important. Hu (2015) makes several contributions to the existing literature by assessing the dynamic process of technological imitation and innovation in East Asian economies, particularly in South Korea, and the role of policies and institutions for technological development. Despite its contributions, Hu's paper is hampered by several shortcomings. First, the paper does not explain technological diversity among East Asian economies. Hu uses the experience of South Korea, which transitioned successfully from technological imitation to innovation, as an interesting case study. In the case study, four major contributing factors to the South Korean success that Hu points out are the government's proactive role, its good-quality human capital in science and engineering, its well-developed link to a global production network, and as a more controversial point an international environment that was lax in enforcing intellectual property rights. It would be interesting to know to what extent the experiences of other East Asian economies, such as China, Indonesia, Malaysia, Singapore, and Taiwan, are different from that of South Korea in terms of the development of capabilities in technological adaptation and innovation. Some appropriate measures of national technological capacities, such as the number of patents, the payments and receipts of royalties and license fees, and research and development (R&D) expenditures, could be used to compare technological progress across East Asian economies over time. Discussions of national science and technology system and policies could also help evaluate to what extent the South Korean experience is applicable to other economies, considering that South Korea's technological innovation activities have been concentrated in large enterprises. Second, Hu's paper has very little discussion about the impact of technological innovation on economic growth. As is emphasized in recent endogenous growth theory, the role of technological progress is critical for economic development and growth (Romer, 1990; Hall & Jones, 1999). Lee and Hong (2012) present growth accounting results for East Asian countries that show the importance of total factor productivity (TFP) in output growth. There are also a number of recent papers available that measure TFP growth and assess the role of human capital, foreign direct investment, and R&D investment in TFP growth across countries (Lee & Shin, 2012; Park, 2012). Third, there are some improvements that could be made to the data and estimation techniques related to knowledge production function. Hu's paper constructs an unbalanced panel dataset consisting of input and output measures of innovation across countries over the period from 1996 to 2007, and estimates two types of knowledge production function. Based on the empirical findings, Hu argues that institutional and policy variables, such as the protection of intellectual property rights, high-tech products' share of exports, the openness to international trade, manufacturing's gross domestic product share, and the schooling attainment for adult population, while interacting with a country's stage of economic development, play important roles for the production of technology output. More explanations of the empirical specifications and data are needed to enable readers to understand the definitions and functional forms of the knowledge production function. The regressions must be subject to a number of methodological problems, such as endogeneity of the right-hand side variables, omitted variables, measurement errors, and sample selection. Appropriate empirical techniques should be adopted to address these issues. In addition, it is preferable to use period averages of the dependent variables, and to allow sufficient time for the influence from changes in input variables to affect the output variables.
Read full abstract