STONY BROOK, N.Y. — A while back I wrote in support of the $15 minimum-wage initiatives in San Francisco, Seattle, Los Angeles and a few other cities. Now it’s time for me to write the opposite article – the idea of a $15 minimum wage is being taken too far, too fast. Democratic Sen. Bernie Sanders just introduced a bill to mandate a federal $15 minimum wage, and he probably will make it a plank of his presidential campaign. This would be a bad idea.

At this point, some of my more left-leaning readers will probably get mad, and say that I am shilling for employers or the rich. They’ll list a litany of ways in which the poor and working class have suffered or been left behind in the last few decades, and demand to know why I’m opposing a measure intended to help the helpless and downtrodden.

But the argument against minimum wages isn’t that they hurt the rich; it’s that they can end up hurting the poor. Raise minimum wages too high, and you’ll eventually choke off employment, harming the very people who the policy is intended to help.

We’re not going to get an answer to the employment question until some cities try a $15 minimum wage. It might work, and it might not. Smaller minimum-wage hikes in the past weren’t too damaging, but $15 is uncharted territory. Experiments such as those in Seattle, San Francisco and Los Angeles will help us determine whether $15 is too high.

That’s exactly why a federal minimum wage is a bad idea. If some cities have $15 minimum wages and other nearby cities don’t, we can compare the former to the latter. We can look at their respective economic performances. We can see if people and businesses move from cities with high minimum wages to those with low minimum wages. If the minimum wage is $15 everywhere, we can’t make that comparison. A federal minimum wage ruins the experiment.

Now, with a federal minimum wage, there will be nothing to stop us from doing a before-and-after comparison of economic performance. But that probably won’t tell us much, because a lot of other stuff will be happening at the same time. There will be recessions or booms, asset prices will rise or fall, and a bunch of other policies will be enacted. We won’t get a clean test.

Suppose we enact a federal $15 minimum wage, and employment falls a lot during the next decade? Pundits will be arguing for years. Was it the minimum wage that caused the fall, they will ask, or too much regulation of business? Was it shifting patterns of trade? Was it declining business dynamism? The list of possible causes will be long, and the data won’t be able to distinguish among them.

The same thing will happen if employment doesn’t fall – minimum-wage opponents will argue that without the minimum-wage hike, it would have risen even more, while proponents will say that the hike was harmless.

In other words, local $15 minimum wages will resolve arguments. A federal $15 minimum wage will not.

Actually, there’s a big reason to think that a federal minimum wage isn’t a good policy in general. As Slate’s Jordan Weissmann points out, the huge differences in local costs of living mean that in some areas, $15 is only a small increase, while in others it’s a huge boost. This means that a $15 federal minimum wage would raise the wages of the urban working class only a modest amount, but would raise the wages of the small-town working class a huge amount.

Whether you’re a minimum-wage supporter or an opponent, you should be afraid of that kind of uniformity. If minimum wages don’t hurt employment, then a federal minimum wage is unfair to workers in big cities because their raises will be less than those of small-town workers. But if minimum wages do hurt employment, then small-town workers are going to be put out of a job – and that’s much worse. A much better idea would be local “living wage” campaigns, each one pushing for a number that is appropriate to specific job markets.

So although I’m glad to see cities experimenting with a $15 minimum wage, I think we should avoid implementing the policy at the federal level, at least until we see the results of the local experiments.

And even if the local experiments are successful, I think that targeted living-wage initiatives are a better idea than a federal minimum wage. If we’re going to help the working class, we need effective policies, not just well-intentioned ones.

— The Washington Post News Service with Bloomberg News