The debate over the Seattle experiment has generated more heat than light to this point. A new report from Jacob Vigdor and his colleagues at the University of Washington attempts to shed some light on the effects of the first incremental stage of the increase. They use data from the state’s Employment Security Department from when the law was passed through the fourth quarter of 2015, at which point the minimum wage stood at $11 per hour. This does not include the second stage of increases that took place January 1, 2016, or the further increases that will eventually bring it to $15 per hour and much higher thereafter. The early results show the mixed effects of the first incremental increase, there does not appear to be much evidence of firms being driven out of business, and some low-wage workers have seen their hourly wage increase, it also reduced the employment rate and hours, with the end result for these low-wage workers being “ambiguous and likely fairly small.”
The report only analyzes the first initial stage of the scheduled minimum wage increases, as the authors note and as illustrated in Figure 1. In addition, due to the timing of the study, it can only capture the short-run effects of this first incremental increase. As such, this analysis cannot provide insight into the impact of future additional increases to the minimum wage or what the longer-run effects might be.
Figure 1
Seattle Minimum Wage Schedule and Period Analyzed
Source: Seattle Minimum Wage Study Team, 2016.
They employ multiple strategies to try to discern the impact of the minimum wage increase. In the most straightforward, the “observable change,” they simply compare trends in Seattle before and after the increase took place. This might be affected by differences in underlying trends that would lead to spurious findings, so they include a number of other comparisons, like King County excluding Seattle, but their preferred specification is a “Synthetic Seattle” consisting of an aggregate of zip codes in the state with similar levels and trends to Seattle. They then compare what happened in Seattle, which was subject to the minimum wage increase, to “Synthetic Seattle” which had no minimum wage increase but in other ways was very similar.
As might be expected, the share of workers with wages below $11 an hour decreased significantly, but this was also true to some extent outside of Seattle, which suggest much of this decline might be due to improvements in the broader economy rather than directly attributable to the minimum wage increase. They estimate the minimum wage increase responsible for a $0.73 rise in median hourly earnings.
While the employment rate of low-wage workers in Seattle increased by 2.6 percentage points over the period, it increased less than it did in their preferred comparison Synthetic Seattle, where it increased by 3.8 percentage points, leading them to conclude “the Minimum Wage Ordinance modestly held back Seattle’s employment of low-wage workers relative to the level we could have expected.” Evaluations simply looking at the total number of low-wage jobs in Seattle before and after the increase took effect would have observed a substantial increase in the total number of jobs, and may have erroneously concluded that there were no adverse employment effects of the minimum wage increase. While it is true that the discrete level of jobs increased, there is some evidence that it reduced the employment rate of lower-wage workers compared to what it otherwise would have been.
A similar dynamic can be seen with hours worked for this lower-wage group: due to an healthy broader economy, the hours worked increased relative to Seattle’s history, but improved less than it did in the control groups, and “on balance, it appears that the Minimum Wage Ordinance modestly lowered hours worked” by 4.1 hours per quarter compared to their preferred comparison, Synthetic Seattle.
Table 1
Impact on Low-Wage Workers
Source: Seattle Minimum Wage Study Team, 2016.
In aggregate the authors find some evidence of a reduction in hours per employee, and minimal evidence of an impact on the number of persistent jobs in industries with a high proportion of low-wage workers. They also find an increase in the rate of business closures and business openings, and suggest this could be in line with other research that found that minimum wage increases prompt in firm composition from more labor intensive companies to those that rely more on capital.
We don’t yet know what the long-term effects of the minimum wage increase will be. This study only encompasses the first step of the phase-in, and later scheduled increases will raise the wage floor to levels that are outside the scope of most past U.S. experiences, making it difficult to estimate the magnitude of potential effects. Some of the minimum wage literature has found that the long-run effects of an increase are greater in magnitude, as the authors of this report note saying “in the long-run, certain industries affected by the minimum wage, such as the fast food industry, have more opportunity to relocate, change the composition of their workforce, or invest in technologies that reduce their need for labor.” In addition a high wage city like Seattle implementing this incremental step during a time of relatively strong broader economy limit how applicable these findings might be for other places.
The initial results suggest that, at least in this initial stage, the sky is not falling, but there are signs that the increase did have unintended consequences, reducing opportunities for low wage workers that leave the net effect of the increase on these targeted workers ambiguous. Because the minimum wage is poorly targeted to poor households, think teenagers in relatively affluent families who are affected by the minimum wage increase, this ordinance likely has even less of an impact in terms of poverty reduction. These are just preliminary results from the first step of a minimum wage increase in a relatively high-wage city. The adverse effects of further increases and the long-run responses could be much larger and more costly.
