Abstract
This chapter surveys the empirical growth literature on the selected Asian economies investigating the effects of a few important determinants of growth (e.g. the investment rate, trade openness, human capital, financial development, and macroeconomic policy environment). The summary provides some important insights but also presents conflicting findings, which unanimously do not always conform to the conclusions of the theoretical models. For instance, whilst investment is agreed to be an important determinant of growth, the literature puts emphasis on the type of investment in question and the important interactions between investment and other variables like trade openness and good policy environment. On trade openness, the absence of a strong theoretical basis, data limitations, and therefore, the use of right or wrong proxies shadow the impact of trade and growth. However, international data on trade patterns and other empirical evidences do support strongly that openness to trade has positive growth effects. There is no doubt of a strong theoretical support for the impact of human capital on growth. The reflective empirical findings also support this view, except that there are measurement problems for indicators like education attainment and quality of education. Similarly, the consensus is that credible policies and sound institutions are growth supportive. Cross-country findings have provided some uniform results, but country-specific findings show large variability on almost all the tested variables. Against this background, this chapter develops a pragmatic survey on five important hypothesized forces of economic growth.
TopBackground
Estimating growth effects of variables is of interest to academic and policy makers alike. However, a common point of reference for the validation of research findings is the available empirical growth literature. Theoretical developments have provided useful and consistent insights on growth effects, but country-specific evidences have varied. This chapter understudies the literature on five aforesaid determinants of growth and summarizes the recent findings of empirical evidences on these variables for the sample of Asian countries. This section discusses the literature on growth empirics for which any pragmatic summary is almost impossible. It must therefore be stated that any such survey is a tough undertaking because of the multitude of works that are available.
It must be noted that there are two distinct varieties of empirical works. The first uses cross-sectional observations to infer something about the determinants of productivity while the second proceeds by regressing country-specific growth rates on a potentially long list of country characteristics. In many of these studies, there are no underlying theories and therefore, variables included on the right hand side are dictated by previously obtained empirical evidences. Such studies at most provide interesting descriptive statistics. The empirical growth literature has gained prominence because while it is important to focus on long-term average rate of growth, the theory provides little guidance on theoretical specifications for growth in the short-medium term. Rogers (2003) thinks that this is less than ideal, but the complexities of the processes of growth and the lack of encompassing models make it a necessity. Further, for the second type, quality and consistent datasets are not widely available over long spans of time to confidently depend on long-term country-specific estimates. The empirical studies mostly use shorter spans of data although it is widely known that a country’s growth rates are not highly correlated in the short run. Easterly et. al (2004) state that determinants of growth themselves have low persistence. Therefore, it is hard to find long-run stable relationships with short samples of annual data, contrary to the usual findings in country specific studies. Panel estimates are also not immune to this limitation but have the advantages of higher observation count and larger variations in datasets.