Drawing on data from the Current Population Survey (CPS), David Neumark (2002) finds that living wage laws have brought substantial wage increases for a high proportion of workers in cities that have passed these laws. He also finds that living wage laws significantly reduce employment opportunities for low-wage workers. We argue, first, that by truncating his sample to concentrate his analysis on low-wage workers, Neumark’s analysis is vulnerable to sample selection bias, and that his results are not robust to alternative specifications that utilize quantile regression to avoid such selection bias. In addition, we argue that Neumark has erroneously utilized the CPS data set to derive these results. We show that, with respect to both wage and employment effects, Neumark’s results are not robust to more accurate alternative classifications as to which workers are covered by living wage laws. We also show that the wage effects that Neumark observes for all U.S. cities with living wage laws can be more accurately explained as resulting from effects on sub-minimum wage workers in Los Angeles alone of a falling unemployment rate and rising minimum wage in that city.
Available at: http://works.bepress.com/robert_pollin/27/