A recent working paper from the National Bureau of Economic Research by the economists Jeffrey Clemens, Lisa B. Kahn, and Jonathan Meer should make us pause and question the wisdom of higher minimum wages. The economists explore how minimum wages affect the probability of employer-provided health coverage and find that a chunk of the increased earnings for workers who get higher wages will be offset by a reduction in employer-provided health coverage.

There are a lot of ways minimum wages can hurt the people they’re supposed to help. By raising the price of something in a competitive market, people will demand less of it. And while there is some mixed evidence, it does appear to be the case that minimum wages reduce employment.

In the simplest versions of the labor market, we assume that workers work for one thing and one thing only: wages. We know this isn’t strictly true, but it’s a good enough approximation that brings us some important insights into how labor markets work.

Of course, there’s a lot more to a job than wages. People want work that is meaningful or enjoyable. They might especially value safety, comfort, or flexibility. People can also get a lot of non-wage benefits like health coverage, scholarship opportunities, and paid vacation. Workers can (and do) “buy” these perks by accepting lower wages than they would require if the job weren’t as pleasant, meaningful, or safe or if the fringe benefits weren’t as good.

In short, workers don’t live on wages alone, and minimum wages might not change what workers get paid but rather how they get paid. Minimum wages mandate that cash wages take up a bigger part of employee compensation.

The mandated higher wages won’t be a free lunch. Consider someone who earns $12 per hour in total compensation but only $8 per hour in wages (the rest is things like health benefits and things like pleasantness of the job, safety of the workplace, flexibility, and so on). A minimum wage of $10 per hour doesn’t cost her her job, but it changes how she is compensated: instead of $8 in cash wages and $4 in benefits, she gets $10 in cash wages and $2 in benefits.

The effect doesn’t show up in fewer hours worked or a lost job, but she is worse off. It goes in other directions, too: if she is required to take part of her compensation in health insurance, she then gets lower wages or adjustment on some other margin.

Clemens, Kahn, and Meer are limited by available data, so they can’t measure everythingcomprehensively. They narrow their focus to the relationship between minimum wages and employer-provided health coverage and find that for Very Low wage workers (e.g. food service), about 9% of increase in earnings due to a $1 per hour increase in the minimum wage is offset by the lower probability of employer-provided health coverage. For Low Wage workers (e.g. retail sales), its 16%. For Modest Wage workers ( clerks and food service supervisors), it’s 57%—which is unsurprising since workers with higher wages get a smaller bump from minimum wage increases.

Henry Hazlitt, author of Economics in one Lesson, wrote that

“The art of economics consists in not merely looking at the immediate but at the longer effects of any act or policy; it consists in tracing the consequences of that policy not merely for one group but for all groups.”

In the case of the minimum wage, the art of economics also consists in identifying all of the important effects and not just those that are most politically salient. In this case, we would do well to try to measure all of the important effects on things like insurance coverage, job quality, fringe benefits, schedule flexibility, and so on before we decide that it’s wise to “#FightFor15.”