Essays in labor and health economics: Factors affecting the labor supply of registered nurses in the 1990s. Determinants of health expenditures in 12 OECD countries, 1960--1997

The first essay presents estimates of the determinants of labor supply for U.S. registered nurses (RNs) in 1992 and 1996. What factors cause RNs to work and when they do work, how many hours? Results will be useful to policy makers trying to formulate policies for alleviating RN shortages. The results are based on two large micro data sets including 30,000 RNs each from 1992 and 1996. Although corrections are made for potential selection bias, such bias was not of practical importance in the data. The reason is most likely because most RNs work and when they work they tend to work full-time. A key finding is that RN hours are insensitive to changes in the own-wage. Wage increases as a policy tool are unlikely to be a cost effective method for attacking RN shortages. The results also suggest that both labor force participation and hours worked given participation could be substantially increased by employer provided childcare.;The second essay concerns the determinants of the level of health care expenditures (HCE) in 12 OECD countries. Independent variables include not only national income, but also other demographic and exogenous factors that may influence HCE. Emphasis is given to institutional differences among those countries, by grouping them together according to similarities in the systems of financing health care, and some similarities in the methods of remuneration and provision. Determinants of health expenditures are estimated with the new 1999 OECD Health Database which allows estimation for the period 1960 through 1997. Panel techniques are used to estimate the HCE models. The estimations are based on two-way fixed effects models which account for both group (country) effects and time effects along with adjustments for autocorrelation. Estimated income elasticities associated with countries having social-insurance, predominantly private, and tax-based health systems are 0.87, 0.40, and 0.98 respectively. The first two exhibit characteristics of a normal, not a luxury good. Another key result is that the graying of the population is likely to have major effects on HCE.