This paper extends the existing literature on military-labor supply in
three ways. First. using Current Population Survey (CPS) data, it dev
elops a more accurate estimate of the alternative wage faced by young
men. Second, it endogenizes the high-quality goal faced by recruiters.
As the United States Army Recruiting Command became increasingly bett
er at setting this goal during the 1980's, the assumption that the goa
l was exogenous to the recruiting process became increasingly suspect.
This paper presents evidence that the goal is indeed endogenous. Fina
lly, it presents estimates based on a switching simultaneous-equations
statistical specfication that allows behavior to vary across recruiti
ng environments to reflect the asymmetric incentives faced by recruite
rs. The findings show that the estimates of the parameters with the gr
eatest policy content are sensitive to each of these specification iss
ues.