|| Checking for direct PDF access through Ovid
This paper estimates the effect of diabetes on labor-force participation, hours worked, days-out-of-work due to illness, and earnings using data from the National Health Interview Survey. Findings indicate that diabetes, estimated wholly, is significantly detrimental to most labor market outcomes. However, separation of type I and II diabetes shows that much of the negative effect is due to type II diabetes. On average a female with type II diabetes can experience a wage penalty of almost 50% relative to a healthy individual. Additionally, estimates of specifically type II diabetes may be subject to endogeneity bias. To account for this, I utilize whether an individual's biological mother has been diagnosed with diabetes as an instrumental variable. This instrument provides both theoretical and statistical explanatory power to separate the causal effect of type II diabetes on labor-force decisions. Copyright © 2010 John Wiley & Sons, Ltd.