Author
Article Highlights
- Average wages went up in three of five low-wage sectors studied while total jobs declined in restaurants subsector in Minneapolis
- Analysis considers how anticipation of new minimum wage law in St. Paul may have affected wages, hours, and jobs
- Future analyses will investigate the welfare of specific employees and employers
When Minneapolis and St. Paul passed new minimum wage ordinances, the cities contracted with the Federal Reserve Bank of Minneapolis to evaluate the economic impacts of the new laws. While the target minimum wage of $15/hour won’t go into effect for all city businesses until 2024 in Minneapolis and 2028 in St. Paul, a team of economists from the Federal Reserve Bank of Minneapolis and the University of Minnesota have begun studying how the incremental increases that have taken effect to date have impacted wages, hours, jobs, and worker earnings in the Twin Cities. See Figure 1 for the phases of the cities’ minimum wage laws.
Source: Loukas Karabarbounis, Jeremy Lise, and Anusha Nath, “Economic Impact Evaluation of the City of Minneapolis’s Minimum Wage Ordinance,” Federal Reserve Bank of Minneapolis
Some wage gains but limited impact on earnings in Minneapolis
Principal investigators Loukas Karabarbounis, Jeremy Lise, and Anusha Nath focus their analysis on five sectors in which a significant share of Minneapolis and St. Paul workers—at least 30 percent—earn below the $15/hour target, as those are the sectors most likely to be impacted by the minimum wage law. See Figure 2 for more information about these sectors.
Source: Loukas Karabarbounis, Jeremy Lise, and Anusha Nath, “Economic Impact Evaluation of the City of Minneapolis’s Minimum Wage Ordinance” and “Economic Impact Evaluation of the City of St. Paul’s Minimum Wage Ordinance,” Federal Reserve Bank of Minneapolis
The economists then analyze how the average wages, total number of hours worked by all employees, total number of jobs, and total earnings of all workers in each of these sectors changed in Minneapolis compared with a projection of what would have happened without the minimum wage increase.
The results suggest that in three of the five sectors, average wages grew around 6 percent to 7 percent more in Minneapolis than they would have otherwise. The impacted sectors are retail trade, administration and support, and other services. (“Other services” includes repair and maintenance shops, personal and laundry services, and various civic, professional, and religious organizations.)
The only other statistically significant result at the sector level was an increase in total worker earnings in the “other services” sector. The analysis does not find that the total number of jobs or total hours worked in any of the five sectors changed more in Minneapolis than would have happened in the absence of the minimum wage ordinance. See Table 1 for a summary of the Minneapolis results.
| Sector | Wages | Jobs | Hours | Earnings |
|---|---|---|---|---|
| Health care and social assistance | – | – | – | – |
| Accommodation and food services | – | – | – | – |
| Administration and support | ▲ | – | – | – |
| Retail trade | ▲ | – | – | – |
| Other services | ▲ | – | – | ▲ |
| Full-service restaurants | ▲ | ▼ | – | – |
| Limited-service restaurants | ▲ | ▼ | – | – |
Because the data are reported quarterly, the investigators’ main results focus on what happened between the first quarter of 2018, when the Minneapolis ordinance took effect, and the first quarter of 2020. That was when the unprecedented health and economic crises caused by the coronavirus as well as the protests after the murder of George Floyd began affecting the city of Minneapolis in unique ways, and that impact reverberated for the remainder of the year. Isolating the impact of the minimum wage during those unprecedented times poses a significant challenge and raises questions about how to interpret data from 2020. In future reports, investigators will use additional data and sources of variation with the goal of disentangling the effects of the minimum wage from the effects of the pandemic and racial justice protests.
Restaurant job losses in Minneapolis
At the request of the city, the investigators also looked specifically at the economic impacts on full-service and limited-service restaurants, which are part of the accommodation and food services sector. At limited-service restaurants (establishments where food is ordered at the counter, such as McDonald’s), a full 80 percent of employees were earning below $15 an hour in 2017.
The analysis suggests that wages increased about 4 percent more for full-service and 9 percent more for limited-service restaurant workers than they would have in the absence of the policy. At the same time, the total number of jobs declined by 12 percent more at full-service restaurants and 18 percent more at limited-service restaurants. The investigators calculate that this translates to roughly 1,700 fewer jobs at full-service and 1,200 fewer jobs at limited-service restaurants over the 27 months analyzed. The impact on total earnings for restaurant workers was not statistically significant, meaning the investigators cannot be confident whether total earnings changed or not.
Anticipation effects in St. Paul
St. Paul’s minimum wage law took effect only a couple of months before the pandemic began, making it difficult to isolate the minimum wage’s impact in this highly unusual time. Instead, the investigators analyze economic outcomes between the first quarter of 2018 and the last quarter of 2019, prior to the implementation of the new law. Minneapolis had already begun implementing its wage ordinance at this time, and the mayor of St. Paul had publicly stated his intention to adopt a similar one. Many businesses were thus aware of this possibility and could potentially have acted in anticipation of rising labor costs, perhaps by adopting technology that replaced certain jobs or by relocating outside the city.
The two sectors that saw statistically significant impacts in St. Paul were retail trade and accommodation and food services. In both sectors, wages increased while hours declined. The number of jobs also declined in accommodation and food services, while total worker earnings declined in retail. Restaurants saw the largest impact, with significant declines in jobs, hours, and earnings. Interestingly, the “other services” sector saw a large increase in jobs, a result that will require additional study to fully understand. See Table 2 for a summary of the St. Paul results.
| Sector | Wages | Jobs | Hours | Earnings |
|---|---|---|---|---|
| Health care and social assistance | – | – | – | – |
| Accommodation and food services | ▲ | ▼ | ▼ | – |
| Administration and support | – | – | – | – |
| Retail trade | ▲ | – | ▼ | ▼ |
| Other services | – | ▲ | – | – |
| Full-service restaurants | – | ▼ | ▼ | ▼ |
| Limited-service restaurants | – | ▼ | ▼ | ▼ |
The estimates of job loss in the restaurant industries of both St. Paul and Minneapolis are particularly large compared with other studies of minimum wages, implying that at least some businesses in the Twin Cities were quite sensitive to the actual or imminent increases in labor costs. Future research will examine economic outcomes of specific workers and establishments of different types in the Twin Cities, which could help explain why certain businesses appear so sensitive to labor costs at this time and place.
Constructing a counterfactual
The main challenge in all empirical evaluations of a new policy—a minimum wage, rent control, a new sales tax—is that it depends on a counterfactual that is impossible to observe: What would have happened in this place, at this time in the absence of this policy? In order to identify the effects of a policy, then, researchers attempt to identify a control group that behaves similarly to the group that will be “treated” by the policy. Karabarbounis, Lise, and Nath use Minnesota cities outside Minneapolis and St. Paul to construct their control group.
However, because economic trends in that control group diverge from trends in the Twin Cities in the period 2001 to 2017, before the new minimum wages went into effect, the economists use a weighted average of the control group to try to construct a “synthetic control group” that behaves similarly to Minneapolis/St. Paul prior to the introduction of the minimum wage. If the trends in the synthetic control match those in Minneapolis/St. Paul closely prior to 2018, then economists feel comfortable in assuming that the differences in outcomes after 2018 are due to the minimum wage increase. For this reason, the accuracy of the synthetic control group—the counterfactual—is the most important part of the analysis.
How closely does the synthetic control group mimic trends in the Twin Cities prior to 2018? In their report, the investigators show how well the synthetic control “fits” the trajectory of wages and jobs in Minneapolis and St. Paul in each sector between 2000 and 2017. The synthetic control accounts for the majority of the variation observed in these economic outcomes over time, although the fit does vary by sector. What constitutes a “good enough” fit remains an active area of research by economists who use this methodology.
Detailed data enhance analysis
This minimum wage evaluation was facilitated by access to detailed administrative data from Minnesota’s Department of Employment and Economic Development (DEED). These data provide quarterly earnings and hours worked for each employee of businesses that file unemployment insurance wage detail reports with the state. Previous analyses of minimum wage policies often studied only quarterly earnings because many states do not collect hours worked. But without that information, it is difficult to understand why earnings go up or down—Have wages changed? Hours? Jobs?
In addition, by merging the DEED data with the Quarterly Census of Employment and Wages, the researchers are able to identify the specific establishment, industry, ZIP code, and city where an individual works. This provides several advantages. Notably, the researchers can include firms with multiple establishments across city borders in their analysis because they can tell in which city an individual works. In contrast, other minimum wage studies have sometimes had to throw these data out because the data do not indicate at which establishment an employee works. Having more data allows researchers to make calculations with greater confidence.
Next steps
Importantly, the analysis in this initial report looks at averages and aggregates to paint a picture of the economic impact of the minimum wage ordinance on sectors and on the economy as a whole. The analysis does not yet tell us about the welfare of specific employees or employers, and it’s likely that in both groups there are some who benefited and some who did not. What happens to individuals who lose jobs, and what does their economic welfare look like after they return to work? On the employer side, what happens to profits after the minimum wage increase? Do firms alter health or other benefits when their labor costs go up? What impact might the minimum wage have on prices that consumers face? Future reports will use data from the Department of Revenue and Department of Human Services to look more closely at these and other questions.
Minimum wage laws can be applied at a city, county, state, or national level. A city minimum wage might allow employers to move their business several blocks—from Minneapolis to Bloomington, say—and bring their workers with them. This means that perhaps jobs weren’t really “lost,” they simply moved. It’s also possible that some people may have transitioned from being an employee to being an independent contractor, a group that is not included in this analysis. Switching to gig work could mean a job didn’t disappear—but it also could have negative ramifications for workers if it means less schedule certainty, lower wages, and no employer-paid benefits.
“Economists generally can’t do controlled experiments—all we can do is observe. And the trouble with trying to draw conclusions from economic observations is that at any given time and place lots of things are happening,” Paul Krugman wrote in a recent column describing the research that won the 2021 Nobel Prize in economics. One of the winners, David Card, conducted a pathbreaking study with Alan Krueger in the 1990s of the effects of a minimum wage law, pioneering a methodological approach that has informed subsequent economic analysis when a true controlled experiment is impossible. Another 2021 winner, Guido W. Imbens, has made further contributions to this type of analysis, including advancing the synthetic control methodology used in the minimum wage analysis discussed here. Now as then, lots of things are happening. Now as then, these results deepen, but do not conclude, our understanding of the ways minimum wages affect our economy and the people in it.
Additional Resources
Economic Impact Evaluation of the City of Minneapolis’s Minimum Wage Ordinance [PDF]
November 1, 2021
Economic Impact Evaluation of the City of Saint Paul’s Minimum Wage Ordinance [PDF]
November 1, 2021
Two-Pager—Labor Market Effects of Minimum Wage Increase in Minneapolis and St. Paul [PDF]
Lisa Camner McKay is a senior writer with the Opportunity & Inclusive Growth Institute at the Minneapolis Fed. In this role, she creates content for diverse audiences in support of the Institute’s policy and research work.
