Many studies find a correlation between how much a worker is paid (per hour) and how many hours she works. The usual causal interpretation of this correlation is that those with lower wages choose to work fewer hours. However, it is also possible that those who work fewer hours are offered lower wages. We find evidence that the latter interpretation may be relevant as well, at least for older workers. This means that researchers are potentially overestimating the response to work incentives of high wages. This is important because tax cuts are viewed as similar to increases in the wage. Therefore, unless labor supply analyses account for the fact that part-time workers earn less than full-time workers, the work disincentives of taxes may be overestimated. This Chicago Fed Letter summarizes the nature of this problem, our empirical solution, and some public policy issues related to it.