Nicole Kelly, left, interviews Angelo Falcone for a bartender job at a job fair in Miami.

“Wage rigidity” is an important feature of many models of the macroeconomy because the term refers to how firms try to deal with their employees when recessions hit. During economic downturns, output declines as firms sell fewer goods and services, which in turn reduces a firm’s revenue—leading to the need to cut costs. Given the large role of labor costs in overall firm costs, that means cutting labor costs. Now there are costs to firing and then rehiring workers, so firms ideally would try to keep as many workers as possible on staff and reduce wages to cut down on costs. But empirically that’s not what happens. Wages don’t seem to change that much and instead workers get laid off to reduce costs. That’s why such wage rigidity causes unemployment to spike during recessions.

Some research on wage rigidity challenges this assumption. Pointing to data on individual wage growth, some economists argue that the wages of new hires is more important. If wages are really rigid, then the inflation-adjusted wages of new hires won’t vary as recessions come and go. Yet these researchers can point to data showing the wages of new hires moving up during economic expansions and down during recessions. So perhaps wages are more flexible than some think.

Now comes a new paper that shows how the cyclical nature of the wages of new hires isn’t really evidence against wage rigidity. The working paper, by economists Mark Gertler of New York University, Christopher Huckfeldt of Cornell University, and Antonella Trigari of Bocconi University, was released earlier this month. The three economists’ major point is to show that looking at the wages of all new hires in the United States is lumping together two groups of workers with different experiences. There are new hires who were previously unemployed and then there are new hires who were previously employed.

Using data from the Survey of Income and Program Participation, the authors can see how individual workers’ wages change over time. What they find is that the trends in wages for these two different groups of new hires are clearly different. The wages for new hires from the unemployment line don’t vary much more over time than the wages of already employed workers. But the wages of new hires from the ranks of the already employed do vary. This phenomenon, however, is less about flexible wages and more about workers moving up the job ladder, which mostly only happens during economic expansions, and is the reason why wages for these new hires move with economic cycles.

Outside of the implications for macroeconomic models of the labor market during recessions, the results from Gertler, Huckfeldt, and Trigari are also a reminder of the effects an economic downturn can have on workers’ career earnings. Recessions hinder the hiring of already employed workers, which hurts their chances of climbing the job ladder and future wage gains. Downturns don’t just harm the workers who lose jobs, but also the ones who keep their jobs.