One reason the debate over the minimum wage is so contentious is that data about how it affects employers and workers are murky. So a new study showing that increases in the wage can force restaurants out of business is welcome — but it doesn't show what minimum wage critics claim it does.
The working paper by the husband-and-wife team of Michael Luca of Harvard Business School and Dara Lee Luca of Mathematica Policy Research examines the closure rates of restaurants in the San Francisco Bay area from 2008 through 2016, a period in which various communities in the region raised their minimum wages at different points.
This enabled the Lucas to isolate the effect of wage increases from other factors affecting the industry, such as rents and general economic conditions. What’s most innovative about their approach was that they mapped restaurant closures against restaurant quality by using customer ratings on
What they found is that minimum wage hikes have an especially pronounced effect on restaurants rated mediocre or worse — that is at the median rating of 3.5 stars or below. Restaurants with ratings of 4 or 5 stars were almost entirely insulated from the impact of higher wages.
The data, they say, "suggest that a one dollar increase in the minimum wage leads to a 14% increase in the likelihood of exit for a 3.5-star restaurant...but has no discernible impact for a 5-star restaurant." Lower-quality restaurants, which are "closer to the margin of exit, are disproportionately impacted" by the higher minimum wage. To put it another way, the higher minimum wage does drive restaurants out of business— but chiefly restaurants that weren't very good to begin with.
What they did not conclude was that a higher minimum wage hurts employment. In fact, they didn't study that factor.
That seems to have gone over the heads of the anti-minimum wage lobby, which has seized on the Lucas' paper to declare that it shows that a higher minimum wage kills jobs. That's not what the paper found, Luca told me.
"I wouldn't feel comfortable making sweeping conclusions about unemployment based on our findings," he says. He conjectures, in fact, that employment losses resulting from the closing of bad restaurants might be at least partially absorbed by the survival of better restaurants, which might have to hire more workers to meet demand. But nailing down that effect would require further study.
Minimum wage critics have strained to find good evidence that increases have a major employment effect; most studies find either no effect, or a very slight impact on the least-educated and -experienced workers. The restaurant lobby tends to rely on the "stands to reason" argument, which is where you go when the empirical evidence isn't on your side.
The Lucas examined restaurant failures in 13 Bay Area communities, including San Francisco, where the minimum wage, currently $13 an hour, will rise to $14 on July 1 and $15 a year later. They followed the life arcs of more than 35,000 restaurants over the study period with average ratings of 3.5, including about 6,000 in San Francisco itself. They noted at the outset that restaurants are the leading employers of minimum-wage labor, and about one in five of all restaurants close within a year of opening.
The authors are careful not to take the statistics further than they warrant. "We can say that at any wage level, 3- and 2-star restaurants are more likely to go out of business than 4's and 5's," Luca says. Their profit margins and the loyalty of customers are probably lower. "They have less wiggle room to start with." Lower-quality restaurants, "are already closer to the margin of exit," the paper states.
The data couldn't tell the authors if the higher-rated restaurants are insulated from higher minimums because they pay higher wages, or whether they're willing to earn lower profits and pass the difference on to employees.
The Lucas believe they've made an important contribution to economic research by identifying a new kind of data set to complement the Bureau of Labor Statistics and census data commonly mined for minimum-wage studies. "It's hard to get something granular enough to look at the kinds of businesses affected by the minimum wage" from those traditional data, Luca says.
The care the Lucas took with their data hasn't been reflected in much of the public commentary about their findings. Over at Barry Ritholtz's Big Picture website, blogger Invictus compiled a short roster of tendentious spin from minimum-wage critics.
At Forbes, for instance, Tim Worstall erroneously concludes that the paper "tells us that the old story is correct — the higher minimum wage kills jobs." It tells us nothing of the kind. Worstall also asserts that it "isn't a huge surprise" that mediocre restaurants are disproportionately affected, since they "are generally going to be the cheaper ones after all."
Actually, no. "It's not the case that all burrito places are 2-star and steak places are 5-star," Luca says. The authors controlled for price, so they could compare 5-star burrito places with 2-star burrito places, and so on. And Yelp's ratings don't correlate with price anyway — it's not uncommon for a good burrito truck to rate higher than a lousy steak place.
Over at the Wall Street Journal's op-ed page, Michael Saltsman of the Employment Policies Institute, which has been accused of acting as a front for the restaurant lobby and other big industries opposed to minimum wage increases, made the same error. He asserts that the Luca paper indicates that "San Francisco's minimum wage experiment may be dangerous for your favorite white-tablecloth restaurant— the kind of place where the food is exquisite and can command a premium — but it's downright deadly for your local white-apron diner."
National Review went even further by declaring that as the San Francisco minimum wage rises over the next two years, its restaurant industry "will likely shrink, as nearly 6,000 restaurant employers contemplate whether paying $15 per hour salaries is feasible." The paper makes no such finding.
Although the paper finds that a higher minimum wage generally raises the bar to entry, data from the city indicates that the restaurant industry has experienced net growth in establishments during the first few years of increases. In fact, one factor that restaurateurs identify as a cause of failure is the explosion in restaurant competition over the last four years. In 2012, there were 3,600 restaurants in San Francisco; last year there were 7,600.
It's only natural that competition on this scale would make mediocre or poor restaurants much shakier. A one-star increase in rating, the Lucas found, is associated with, if not the cause of, a decline in the likelihood of exit of about 70%, as well as a 5% to 9% increase in revenue. That explains the title the Lucas gave their paper: "Survival of the Fittest."