Essays On Labor Economics: Human Capital Risk And Labor Market Outcomes And Learning By Doing In Medicine
Tristao, Ignez Miranda
Rust, John P.
MetadataShow full item record
This dissertation consists of two essays. In the first essay I show that there are substantial differences in unemployment durations and reemployment outcomes for workers coming from different occupations. I argue that this variation can be explained by differences in occupational employment risk, arising from two sources: (1) the diversification of occupational employment across industries; and (2) the volatility of industry employment fluctuations, including sectoral co-movements. I define and estimate a measure of occupational employment risk (OER), which I relate to unemployment durations and wage losses. My results indicate that unemployed workers in high employment risk occupations, as defined by the OER measure, have 5 percent lower hazard ratios of leaving unemployment to a job in the same occupation and have 5 percent higher wage losses upon reemployment than workers in low OER occupations. Among occupational switchers, workers in higher OER occupations have 11.5 percent higher wage losses than workers in lower OER occupations. In my second essay, I and my co-authors estimate the effect of physician's experience on health outcomes. It is a common belief that experience can improve the level of skills, which suggests that there may be some learning by doing with practice. Economists have tried to empirically determine the existence of learning by doing in the medical area, because of its important policy implications. However, it is difficult to define and measure health outcomes since they are affected by patient selection and underlying conditions, making it hard to disentangle learning by doing from other effects. In this paper, we use a 'clean-cut' medical procedure that allows us to overcome those confounding issues. We use refractive eye surgery, an operation with a well-defined eligibility criterion and objective measures of previous condition and posterior outcome, which depend minimally on post-surgical care. The data used in the study is a two-year longitudinal census of refractive surgery patients collected by us from individual patients' chart. We find that the learning is coming more from the improvement in the surgical center's ability to translate the surgical plan into the desired eyesight correction rather than from the accumulation of the physician experience.