Seattle’s painful lesson on the road to a $15 minimum wage
In the summer of 2014, the Seattle City Council unanimously passed a bill increasing the city’s minimum wage to $15 an hour. “No city or state has gone this far. We go into uncharted territory,” said council member Sally Clark. The City of Seattle, to its credit, actually made some effort to chart the waters as they went, funding an ideologically diverse research team at the University of Washington whose members range from Jacob Vigdor, who is a fellow at the conservative Manhattan Institute, to Hillary Wething, a graduate student who used to be a fellow at the left-wing Economic Policy Institute.
Since the law raised the wage in stages, they studied it in stages. The results of the first jump, from $9.47 to $11 an hour, were released last year, and seemed to show that the effects on earnings were pretty small — an increase of about $72 every three months — and that low-wage employment declined slightly. In the long-running battle over the effects of the minimum wage, this paper didn’t offer much ammunition for either side, and thus occasioned relatively little excitement in the wonkosphere.
Now the question has acquired a little sizzle. UW is not the only school studying Seattle’s experiment, and last week a report came out from UC Berkeley, focused specifically on food services. Last month,
that study reported: “Our results show that wages in food services did increase — indicating the policy achieved its goal. … Employment in food service, however, was not affected, even among the limited-service restaurants, many of them franchisees, for whom the policy was most binding.”
Seattle Mayor Ed Murray seemed ecstatic at the news; as Michael Saltsman noted at the time, he “conveniently had an infographic designed and ready to go for the study’s release. His office excitedly tweeted that the policy had ‘raised food workers’ pay, without negative impact on employment,’ linking to an uploaded study version on the Mayor’s personal .gov website rather than a University domain.”
Last week’s data gives ammunition to the mayor’s opposition. The University of Washington released its second study, this one covering the increase from $11 an hour to $13. And this study found huge effects: For every 1 percent increase in their hourly wage, low-wage workers saw a 3 percent reduction in the number of hours worked. As a result, they lost about $125 in earnings a month, clawing back the entire gain from the earlier hike and more.
Mayor Murray did not have an infographic ready to go for this study. Instead, he simply retweeted the infographic from the old one.
This is hardly the first time we’ve had dueling studies on the minimum wage; indeed, by this point, the dueling is so common that the minimum wage practically has its own rules of engagement. But it’s pretty rare for a city to fund one study and then try to rebut it with another. Worse still for the citizens of Seattle, a read of the new paper suggests that this rebuttal won’t work.
To understand why not, you first need to understand a little bit of the history of minimum wage research. Such research has often focused on restaurants, because they employ a lot of low-wage workers, which would seem to make them a good proxy for a variety of low-wage industries. The most famous study in the field is probably David Card and Alan Krueger’s study of fast food restaurants from 1994. When you hear someone on the left claim that “research shows” minimum wage increases don’t hurt employment, then you can almost certainly trace that statement back to Card and Krueger.
Reality was always more complicated. We are not working with a model in a textbook where undifferentiated widget manufacturers hire indistinguishable workers from a giant pile of bodies marked “labor.” Different industries, different firms within those industries, and different workers within those firms may all have very different experiences under a minimum wage: some unaffected, some better off, some driven into insolvency. So while Card and Krueger was important, it was by no means the final word that some took it for.
The new paper, unlike Card and Krueger, has broad data covering all of Washington’s workers, not just those employed by fast food franchises that happened to be operating at the beginning of the study. This is no slur on Card and Krueger, mind you; Washington State just happens to have unusually rich data available compared to most other states.
Leading labor economist David Autor told The Washington Post that “This strikes me as a study that is likely to influence people,” saying the study is “very credible” and “sufficiently compelling in its design and statistical power that it can change minds.” In other words: if you thought it was settled science that raising the minimum wage is good for workers, be prepared to think again.
And particularly be prepared to rethink very high minimum wages, like those supported by the “Fight for $15” folks. For as the authors note, the first round of hikes had relatively small impacts, while the second round had huge ones, suggesting that the effects may be nonlinear. And that makes sense. Relatively few people in this country make the minimum wage, so a small increase doesn’t make that much difference to most workers, or most employers. But a large jump affects more people, and the wage increases are much bigger for the lowest-paid staffers. If you make $9 an hour, but generate $10.50 in revenue for your boss, a law that raises the wage to $10.45 may cause her to shrug and decide it’s easier to keep you on as long as she’s making something. But a wage that forces her to pay you far more than you bring in? Continuing to employ you would just be bad business.
It’s worth noting that Card and Krueger’s famous study involved an increase in the minimum wage from $4.25 an hour to $5.05. That was a significant increase — about 18 percent. But Seattle’s minimum wage has already increased by 37 percent, and it still has roughly another 20 percent to go.
At some level, we all intuitively understood that this was true. If the minimum wage increases by a penny an hour, probably even most rock-ribbed conservatives would not predict mass firings. On the other hand, if the wage was arbitrarily set to $100 an hour, even ardent labor activists would presumably expect widespread unemployment to follow. You can’t flat-out say “minimum wages don’t increase unemployment,” because the size of the increase, and the level of the resulting wage, obviously matter at some margin.
Seattle may have discovered that margin. And unfortunately, it may yet discover even further, uglier margins when the data is in on the full increase to $15. That’s the danger of striking out for uncharted territory; sometimes, you end up where there be dragons.