For more than 80 years, many macroeconomic analyses have been premised on the assumption that workers' nominal wage rates cannot be cut. The U.S. evidence on this assumption has been inconclusive because of distortions from reporting error in household surveys. Following a British literature, we reconsider the issue with more accurate wage data from the payroll records of most employers in the State of Washington over the period 2005-2015. For every one of the 40 four-quarters-apart periods for which we observe year-to-year wage changes, we find that at least 20 percent of job stayers experience nominal wage reductions.