Comment on “Escaping the Middle‐Income Trap in Southeast Asia: Micro Evidence on Innovation, Productivity, and Globalization”

Abstract

Lee and Narjoko (2015) provide a comprehensive literature survey of empirical studies that use manufacturing plant-level or firm-level data to investigate the relationship between innovation and globalization in Southeast Asian countries. Their survey covers studies for Indonesia, Malaysia, Philippines, Thailand, and Vietnam because the quality and availability of the microdata in these countries are better than that in the other countries of the Association of Southeast Asian Nations. Based on their literature survey, Lee and Narjoko conclude that globalization via trade and foreign direct investment plays an important role in the firm's innovative activities and productivity improvements in these countries, and recommend that innovation policies should be directed toward the enhancement of productivity, particularly in export-oriented sectors. One of my major concerns with Lee and Narjoko's paper is in the causal interpretation of the empirical results. As the authors mentioned, all the results surveyed in this paper have not examined the causal relationship running from firms' overseas activity to their innovation input (e.g. research and development activity) or productivity. Further considering the fact that only a few studies have examined the linkage among firm innovation, productivity, and overseas activity by firm size, I think that the statements such as “Innovation policy targeted at enhancing productivity, especially in export-oriented sectors” (p. 143) and “SMEs might warrant special attention given that the productivity constraint to exporting is likely to be larger for these firms” (p. 143) are not sufficiently supported by the concrete empirical evidence. Another concern I have relates to the heterogeneity of the countries covered in their survey studies. The economic structures and the degrees of participation in international trade are widely different across the countries in the Southeast Asian region. This implies the possibility that the environments surrounding small and medium enterprises (SMEs) and the levels of foreign firm participation in each country can be also quite different. Given such heterogeneity among the economic conditions in each country, it seems rather difficult to provide a common policy recommendation. It is important for Lee and Narjoko to take the differences such as heterogeneity among countries into account in their discussion, and hopefully propose policy recommendations according to the different stages of development in each country. Related to the second point, one way to explicitly study the heterogeneity of firms in each country is to examine the dynamics of firm-level productivity measured by, for example, total factor productivity (TFP). More precisely, it would be informative to examine the catch-up process of low productivity firms (usually SMEs) in the country toward the national productivity frontier firms or global frontier firms (e.g. high productivity firms in the USA as in the literature). As one example, Fukao et al. (2011) employ the empirical model in Bartelsman et al. (2008), and examine such a catch-up process using Japanese, Korean, Taiwanese, and Chinese manufacturing firm data. Another concern about Lee and Narjoko's paper is that it does not provide clear explanation of the relationship between firm network participation and its innovative activity. Furthermore, previous studies in the region do not explicitly examine the mechanisms of how the internal and external linkages are related to a firm's innovation performances. As a reference, Ma et al. (2014), for example, analyze the mechanism of a firm's export participation on its productivity development using Chinese manufacturing data, and find that “firms become less capital-intensive but more productive after exporting, compared to nonexporters that share similar ex ante characteristics” (abstract). It would be useful if Lee and Narjoko implemented a similar empirical analysis using microdata in Southeast Asian countries. As a final remark, given that the region is facing the slowdown in TFP growth and find it important to overcome the middle-income trap, Lee and Narjoko should conduct the productivity growth decomposition and discuss the sources of the productivity slowdown in these countries. Such an analysis might allow the authors to confirm that improving the resource allocation in a country through trade liberalization is a better policy recommendation than increasing TFP growth rather than promoting more firms to participate in the export market. As one extant study in this context, Hsieh and Klenow (2009) find that both China and India can increase their aggregate TFP by reducing resource misallocation in each country.