Essays on growth, stabilization, and the business cycle.

Abstract

This dissertation examines three important issues in macroeconomics in three different chapters. Chapter I attempts to resolve a long-lasting puzzle on the secular trend of capital utilization. While the convergence theory of economic growth predicts that low-income economies should have high rates of capital utilization, an extensive empirical literature documents that most developing countries have a shorter workweek of capital than capital-abundant countries. I show in Chapter I that these empirical findings are not anomalous but instead are fully consistent with a neoclassical growth model when we view the fast growth of low-income economies as a transition initiated by fast growth of technology and labor supply. In Chapter II, I investigate the causal relationship among government size, sector-specific income risks, and openness to trade. This study estimates a simultaneous equation system based on an empirical specification derived from two models with sector-specific income shocks and labor immobility. The estimation results indicate that the size of government reduces intersectoral income fluctuation, but at the same time, an economy facing higher intersectoral fluctuation will have a larger government. Furthermore, in stabilizing economic uncertainty, governments tend to resort to redistributive policies rather than spending although government spending is almost as effective as government subsidies and transfers for this purpose. It is also found that intersectoral fluctuation rises when a country becomes more open to international trade, and further increases as an economy is exposed to more intense external shocks. Chapter III investigates in a general equilibrium context the implications of on-the-job human capital accumulation for the estimation of intertemporal elasticity of substitution in labor supply (IES). Our business cycle model shows that on-the-job learning investment is countercyclical and thus the wage paid to production hours is less procyclical than the conventionally-measured wage rate. Furthermore, when the wage rate includes learning-by-doing effect of labor, the true returns to production hours are found to be less procyclical than the wage paid to production hours. This spurious procyclicality of the conventionally-measured wage rate creates substantial downward biases in the IES estimation.