What happens when a study shows that a minimum-wage increase is simply having its intended effect? When it’s found to raise the pay of low-wage workers without causing much in the way of the job displacements that critics rail about? Unfortunately, one thing that apparently happens is the findings get misinterpreted (though, as I’ll show, this is partly due to the omission of key statistical information).
The study to which I’m referring examines the impact of the first stage of the minimum-wage increase in Seattle. In April 2015, the city raised its minimum wage from around $9.50 to $11, on the way to $15 an hour by 2017 (for employers with 500 or more employees and certain other employers; the minimum wage for most Seattle businesses rose to $10 in April 2015, and $15 will not go into effect for all Seattle businesses until 2021). The pay of affected workers went up almost 12 percent, compared to a 5 percent increase for workers in nearby, similar places that weren’t bound by the increase. The study’s authors concluded that the increase raised the pay of affected workers by seven percentage points more than might otherwise have occurred.
The study also found that, relative to historical trends, the rate at which low-wage workers affected by the increase stayed employed rose by about three percentage points. For workers in the control group, it was up four points. Thus, absent the minimum-wage increase, there’d arguably be one percentage point more affected workers employed in Seattle.
Putting aside for a moment the critical question of whether these changes are actually meaningful in a statistical sense, these outcomes fit comfortably into a view well understood by minimum-wage advocates and increasingly accepted by economists: most increases have their intended effect of lifting the pay of low-wage workers with little in the way of job losses.
To be clear, the fact that the policy has its intended effect doesn’t mean every affected worker ends up ahead (there is no policy on Earth that is always and everywhere costless to its intended beneficiaries). It means that the vast majority of low-wage workers end up with higher earnings. Even if some workers lose some hours of work, their annual income often goes up (which, in fact, is another finding from the study).
Minimum-wage opponents who claim that increases will cripple local economies, either overall or even in their low-wage sectors, thus get no help from the Seattle results. The study’s authors point out that one challenge in teasing out minimum-wage effects was that the Seattle economy “boomed” over this period, posting growth rates that “tripled the national average” and “outpaced Seattle’s own robust performance in recent years.”
Yet, despite these expected, generally positive findings, the study’s press coverage has been pretty negative. Attacks from knee-jerk opponents of the minimum wage — who in some cases are paid by the low-wage employer and lobby to shoot at anything that moves, regardless of the evidence — were expected and are easily dismissed. But writers who are typically more careful have also erroneously declared that “employment went down” in Seattle as a result of the increase (employment actually went up in Seattle relative to past trends), or that it “did little to help workers.” In reality, what’s unfolding in Seattle thus far reflects the conclusion of a recent, exhaustive review of the minimum-wage literature by Dale Belman and Paul Wolfson: “While not a full solution to the issues of low-wage work, [the minimum wage] is a useful instrument of policy that has low social costs and clear benefits.”
The study also has several important limitations, noted by minimum-wage scholar Michael Reich. First, as Reich explains, “The authors did not report the standard errors of the estimates, even though their calculations indicated that the employment effect was not distinguishable from zero.” This is a serious omission; policymakers cannot make informed decisions without that information.
Second, because of data limitations, the study analyzes employment changes solely in single-establishment firms. That means, for example, that retail and restaurant chains — groups significantly affected by the minimum wage — are generally left out of the study (multi-establishment employers account for half of Seattle’s jobs). In fact, when all establishments are included in the analysis, employment outcomes were relatively more positive in Seattle than in the control group, both for all firms and for lower-wage firms.
These and other problems, especially the omission of standard errors, need to be resolved. But for now, here’s what we can conclude: After Seattle raised its minimum wage, low-wage workers’ employment, hours and wages all rose substantially. Neighboring areas that had similar trends in these variables before the increase — and that, by the way, were also bound by the highest state minimum wage in the country when the increase took effect — saw even larger employment and hours gains. In other words, relatively high minimum wages in Seattle and in Washington more broadly have had their intended impact and have been perfectly compatible with a strong economy, one that’s handily beating national averages.
As the Seattle minimum wage phase-in progresses, and the differential between the wage there and elsewhere grows larger, perhaps we’ll get results that really do fall outside the bounds of the standard outcomes. But that hasn’t been the case thus far.