Last August I noted their study of the effects of the first nine months of the minimum wage in 2015 ("Higher Local Minimum Wages: Early Results from Seattle," August 8, 2016). Now the group has published on its website a follow-up study that includes data for the first nine months of 2016--that is, it includes the period just after Seattle's minimum wage rose from $11/hour to $13/hour. The most recent study is "Minimum Wage Increases, Wages, and Low-Income Employment: Evidence from Seattle," by Ekaterina Jardim, Mark C. Long, Robert Plotnick, Emma van Inwegen, Jacob Vigdor, and Hilary Wething (June 2017; it's also available as NBER Working Paper #23532).Here's their central finding:
"Our preferred estimates suggest that the Seattle Minimum Wage Ordinance caused hours worked by low-skilled workers (i.e., those earning under $19 per hour) to fall by 9.4% during the three quarters when the minimum wage was $13 per hour, resulting in a loss of 3.5 million hours worked per calendar quarter. Alternative estimates show the number of low-wage jobs declined by 6.8%, which represents a loss of more than 5,000 jobs. These estimates are robust to cutoffs other than $19. A 3.1% increase in wages in jobs that paid less than $19 coupled with a 9.4% loss in hours yields a labor demand elasticity of roughly -3.0, and this large elasticity estimate is robust to other cutoffs. ...
Consequently, total payroll fell for such jobs, implying that the minimum wage ordinance lowered low-wage employees’ earnings by an average of $125 per month in 2016."
The finding is getting a fair amount of attention, because the researchers have some especially interesting data. A lot of minimum wage studies have data on total earnings--which is collected for Social Security and tax purposes--but not on hours worked. However, the state of Washington collects data both on hours and earnings. The authors note:
"The analyses use data from the State of Washington Employment Security Department (ESD), collected quarterly from all employers in the state. Washington is one of only 4 states that collects hours worked, in addition to earnings, in these data. These unique data allow us to identify jobs that pay low wages and workers who earn low wages. Prior studies without hours data have had to define `low wage jobs' imperfectly (e.g. by focusing on the food service industry or teenage workers)."
The key difficulty in any minimum wage study lies in choosing a comparison group. For example, one approach might be just to compare Seattle before the minimum wage increase and after. But this "time series" approach has largely fallen out of favor, because so many different factors might be affecting Seattle's labor market at any given time, and it's hard to sort out the minimum wage effects from other factors.
Thus, the study takes an alternative approach and selects several comparison groups. For example, one comparison is between Seattle and the rest King County that surrounds Seattle. Another comparison is to three other counties that surround King County, but do not actually border Seattle. Yet another comparison is carried out by doing a statistical comparison of how Seattle varied in the past compared to other counties across the state and building up to a "synthetic" comparison group by weights those other counties according to how much they have been coordinated with Seattle. All of these comparisons try to look at areas that have similar economic influences to Seattle, and thus would have similar ups and downs, except for the change in Seattle's minimum wage laws. None of these approaches is perfect by itself, of course. But looking at a group of potential comparisons, and thinking about the strengths and weaknesses of each comparison, gives a reasonable range of outcomes.
Every economic study comes hedged about with warnings and issues, which the authors take care to enumerate. Here are a few of them:
1) The estimate of how much a higher minimum wage reduced hours and jobs in this study is fairly high, compared to other minimum wage studies. Why might this be so? The authors point out that Seattle had already started from the nation’s highest state minimum wage at $11/hour, and then raised it substantially higher. Their previous study, looking only at the minimum wage increase to $11/hour in 2015, found much smaller effects. Thus, one way to interpret these findings is that moderate raises in the minimum wage up to about $11/hour have smaller effects on hours worked, but pushing substantially higher will have a noticeably negative effect.
2) Obviously, this is a study of Seattle, where the unemployment rate is a rock-bottom 3.1%. The effects of a minimum wage are likely to be different in a place where the unemployment rate is substantially higher.
3) An issue for a city-level minimum-wage law is that it's fairly easy for a number of employers to focus their hiring outside the city, thus avoiding the law. The study can look for signs of this effect: for example, if this kind of shifting is a big deal, it will probably be happening to a greater extent from Seattle to the areas of King County around the city, rather than shifting to the counties that are further away or to the "synthetic" comparison group. But it's hard to analyze such shifts. In addition, a state-level or federal-level rise in the minimum wage law would not be as susceptible to such shifts.
4) There are a number of other tricky issues like the fact that the state-level data doesn't cover earnings in the "informal" off-the-books sector of the economy.
As other cities experiment with higher minimum wages, other studies will arrive. As I've said before, I'm happy to let the empirical patterns tell me what to believe. Based on this evidence, jacking up what is already a very high city-level minimum wage to even higher levels is bad for the total earnings of low-wage workers in that city.