UC Riverside

This dissertation is composed of three essays on the impact of foreign direct investment (FDI) on productivity growth, convergence, and the environment. Chapter 2 decomposes labor productivity

growth into components attributable to technological change, technological catch-up, foreign capital deepening, domestic capital deepening, and human capital accumulation, thus separating the effects of foreign and domestic capital deepening on productivity growth and convergence. We apply nonparametric production-frontier methods to a worldwide 1980--2005 panel and find that (1) foreign capital accumulation, together with human capital accumulation, is the driving force for productivity growth and increasing international dispersion of productivity, (2) technological change is decidedly non-neutral, with most technological

advancement taking place in foreign-capital-intensive countries, and (3) international polarization is brought about primarily by efficiency changes.

Chapter 3 develops a statistical procedure to select the appropriate

nonparametric efficiency model in terms of its dimensionality. The change of dimensionality is categorized into three cases: nested variable changes (expansion or contraction of a variable set), additive variable changes (aggregation or disaggregation of a variable set) and other non-nested model changes. A bootstrapping method is proposed to measure the size of the dimensionality

effect. Potential bias in raw efficiency scores owing to the dimensionality effect is corrected to reflect true efficient levels. An empirical illustration is presented with the Hughes and Yaisawarng (2004) (hereafter HY) data set.

Using U.S. state-level panel data from 1980--1994, chapter 4 estimates the impact of environmental stringency on the inflows of FDI in the U.S. The stringency of environmental policy is an uncontrollable variable in the operating environment. A three-stage model is proposed to evaluate state performance with environmental variables and reassess the pollution haven hypothesis. The three-stage model combines both data envelopment analysis (DEA) and stochastic frontier analysis (SFA), and can isolate the impact of luck (statistical noise) from those of managerial efficiency and environmental effect. This paper improves the second stage

SFA evaluation by using the Local Linear Least Squares (LLLS) estimator. The empirical result suggests a negative relationship between state-level environment standards and the distribution of foreign capital in the U.S.