How prevalent is downward rigidity in nominal wages? International evidence from payroll records and pay slips

Abstract

For more than 80 years, many (though far from all) influential macroeconomic analyses of the labor market have been premised on the assumption that nominal wages cannot be cut. Some classic studies that used longitudinal household surveys to track job stayers from year to year measured a high incidence of wage cuts, but this evidence reasonably was discounted on the grounds that the measurement of frequent wage cuts could be an artifact of survey response error. The main point of the present paper has been to synthesize a more recent international collection of studies that have sought out more accurate wage data from employers' payroll records and pay slips. Outside of circumstances where nominal wage cuts have been legally prohibited or rendered irrelevant by very high price inflation, most of this evidence has continued to show that nominal wage cuts occur more frequently than has commonly been supposed. Most of us are surprised by this finding, not only because of the persistent influence of Keynes's (1936) contrary assumption in The General Theory but also because introspection, casual empiricism, and Bewley's (1999) interviews tell us that workers really do dislike nominal wage cuts and employers are therefore reluctant to impose them. But is this obvious aversion to wage cuts so extreme as to bind even when inefficient layoffs into unemployment are the alternative? The accumulated international evidence showing that nominal wage cuts occur frequently should inspire reconsideration of the commonly invoked assumption that nominal wages cannot be cut even when efficiency of allocation decisions is at stake. Of course, because the evidence reviewed here is based on longitudinal tracking of job stayers, it pertains directly only to wage rigidity for incumbent workers. As discussed above, a related question is how flexible wages are for the hiring of new workers. Some recent models have assumed that wage rigidity for incumbents spills over into wage rigidity for new hires. In that light, the evidence reported here is indirectly pertinent for hiring wages. If nominal wage cuts are feasible for incumbent workers, why would they not be for new workers? The development of theoretically coherent and empirically relevant accounts of what happens in the labor market over the business cycle remains a crucial mission for economic research. We hope to support that effort by providing a more accurate picture of the frequency and nature of nominal wage cuts.