On television last week arguing against an increase in the minimum wage, I found myself making the case that it was a blunt instrument.

So it is — and so, I should have mentioned, is the alternative policy that I suggested and that has been attracting a lot of attention lately.

An increase in the minimum wage would force higher pay for affluent teenagers working summer jobs and for senior citizens with savings and Social Security checks who are working part-time to keep busy and earn a little extra money. Better, I suggested, to raise the earned-income tax credit, which would boost the income of the working poor in a targeted way, without the job-killing effect of a minimum wage or the benefits to non-poor workers.

A few days later, one of the richest businessmen in the world, Warren Buffett, published an op-ed in the Wall Street Journal making a similar argument. Under the headline, “Better than Raising the Minimum Wage,” Mr. Buffett wrote, “The better answer is a major and carefully crafted expansion of the Earned Income Tax Credit.”

Mr. Buffett, in turn, was echoing, though he may not have known it, one of the points made in a classic New York Times editorial from 1987, which ran under the headline, “The Right Minimum Wage: $0.00” The Times wrote then, “The working poor obviously deserve a better shake. But it should not surpass our ingenuity or generosity to help some of them without hurting others. … Washington could enlarge the existing earned income tax credit, a ‘negative’ income tax paying up to $800 a year to working poor families. This would permit better targeting, since minimum-wage workers in affluent families would not be eligible. “

The earned-income tax credit as a more efficient alternative to a minimum wage increase has a long history as a conservative economic idea. I first heard of it in an undergraduate introductory economics class taught by a veteran of the Reagan administration, Martin Feldstein. It has roots in Milton Friedman’s idea of a negative income tax. It became law in the administration of a Republican president, Gerald Ford. Expanding it has support from high-profile Republican politicians such as Senator Marco Rubio and Rep. Paul Ryan, as well as from high-profile economists who served in recent Republican administrations, such as the chairman of Harvard’s economics department, Gregory Mankiw, and the dean of Columbia’s Business School, Glenn Hubbard.

Democrats such as Presidents Clinton and Obama are also big supporters of expanding the earned-income tax credit.

The earned-income tax credit may indeed be better than a nationwide minimum wage increase to some level as high as $15, which is what Los Angeles City Council just passed. But — and here is the key point often omitted — the EITC has its own formidable problems, perverse incentives, and unintended consequences. Anyone — politicians, Warren Buffett, economists — recommending an expanded EITC as an alternative to a minimum wage increase would do well to keep those problems in mind, because tinkering with the details of the EITC has the potential to create plenty of trouble.

The policy affects family dynamics. There’s a marriage penalty in the EITC, and there’s a bonus (up to a point) for having children. The Clinton welfare reform was a success in reducing the incentive that the Aid to Families With Dependent Children program created to have single-parent families with non-working mothers. But the earned-income tax credit expansion under Clinton created a different incentive — to have single-parent families with working mothers. Making the EITC more generous for childless single workers, as some propose, would likely accelerate existing trends in terms of delaying household formation. This may seem ridiculous — no one decides whether to get married or have children based on a tax incentive of a few thousand dollars a year, right? — but the people who qualify for these programs turn out to be surprisingly familiar with how they work and with what behavior maximizes their benefits.

Changing the EITC from its current annual nature to a monthly program of the sort recommended by Mr. Buffett or Mr. Rubio has its own risks. The Brookings Institution points out that the current system has a “forced savings” dynamic — families “can’t touch their EITC money until tax time. If the families received payments in small portions throughout the year, they would be tempted to spend it, rather than put it away for a rainy day.”

The phase-out of the EITC, along with the phase-out of other benefits such as the MediCaid, the ObamaCare health insurance subsidy, and food stamps, means that low-income workers are subject to effective marginal tax rates that are in some cases higher than the 50%-plus rates that hit high-income earners in high-tax locales such as California or New York City. That, in turn, adversely affects incentives. If the government is going to take away 60 cents of every additional dollar earned, why should a low-income worker take an on-the-books second job, work extra hours, or invest money in education or training to earn a higher wage?

Then there is the matter of cost. The money for a refundable EITC, as for all federal programs has to come from somewhere.

So while Mr. Buffett may be correct when he says an expanded EITC would be better than a mandated higher minimum wage, he’s overstating it quite a bit when he claims that the EITC “does not distort market forces” or that “there is no disincentive effect.”

It’s something for policymakers across the ideological spectrum to remember as they grapple with the question of how best to help the working poor.

Mr. Stoll is editor of FutureOfCapitalism.com.