[Excerpt] In the summer of 1999, I first visited South Africa at the request of the South African government. The government was concerned about the nation's drastic unemployment situation, which in recent years has been estimated at 12-20% using the standard ILO definition (not working but actively looking for work) and which reached as high as 34% when account is also taken of persons who did not work, did not look for work, but who reported themselves willing to take a job if one were offered. Government believed that unemployment was caused by excessively high wages—excessive, that is, relative to market-clearing levels—so they asked us to estimate, inter alia, the wage elasticity of demand for labor. Note the role of both types of research here: government's core hypothesis came both from talking to business-people, who claimed that high wages discouraged them from employing more workers, and from prior econometric estimates.