Much of President Obama’s fiscal-2015 budget request will stoke familiar partisan battles over taxes and spending. One major exception is Obama’s proposal to enlarge the Earned Income Tax Credit (EITC), a “negative tax” wage subsidy celebrated by Democrats and Republicans alike.

Launched in 1975 and expanded several times since then, the EITC is a “refundable” credit, which means it can be delivered to beneficiaries as a direct payment if their net federal-income-tax liability is zero. Designed to incentivize and reward work among disadvantaged Americans, it has been widely hailed as one of our most effective anti-poverty tools. In a recent interview with The New Yorker, Obama declared that the EITC “has probably kept more people out of poverty than a whole lot of other government programs that are currently in place.” According to University of Chicago economist Jeffrey Grogger, not only did EITC expansions help reduce America’s welfare rolls during the 1990s, they also “appear to be the most important single factor in explaining why female family heads increased their employment over 1993–1999.”

Obama’s 2015 budget would make the EITC significantly more generous both to childless workers and to workers without custody of their children, thereby addressing a popular complaint about the program, while also expanding eligibility by raising the upper age limit to 67 from 65. The administration estimates that its proposal “would lift about half a million people above the poverty line and reduce the depth of poverty for 10 million more,” at a ten-year cost of $60 billion.

Many conservatives would view such an EITC expansion as an alternative to a minimum-wage hike. (For example, the Wall Street Journal editorial page argues that the EITC “was designed to be a substitute for the minimum wage that wouldn’t price workers out of the job market.”) By contrast, Obama wants to expand the EITC in addition to raising the federal minimum wage from $7.25 to $10.10 per hour.

Here we have two policy reforms that share a common goal — boosting incomes and slashing poverty rates — but use very different methods to achieve it. Writing in the New York Times in early January, Harvard economist Greg Mankiw explained why Plan A (expanding the EITC) is superior to Plan B (raising the minimum wage):

If we decide as a nation that we want to augment the income of low-wage workers, it seems only right that we all share that responsibility. Plan A does that. By contrast, Plan B concentrates the cost of the wage subsidy on a small subset of businesses and their customers. There is no good reason this group has a special obligation to help those in need.

Indeed, one might argue that this group is already doing more than its share. After all, it is providing jobs to the unskilled. Asking it to do even more, while letting everyone else off the hook, seems particularly churlish.

But even putting fairness aside, there is reason to doubt the efficacy of Plan B. Taxing businesses that hire unskilled workers would alter their behavior in ways that would hurt those we are trying to help.

Several weeks after Mankiw published his column, the Congressional Budget Office (CBO) projected that, according to its “central estimate,” raising the minimum wage to $10.10 per hour “would reduce total employment by about 500,000 workers.” White House officials immediately challenged that figure, with economists Jason Furman and Betsey Stevenson citing “the overall consensus view of economists which is that raising the minimum wage has little or no negative effect on employment.” But Washington Post columnist Robert Samuelson dismissed their argument as “fairy-tale economics”:

Many studies find negative job effects. The CBO didn’t make them up. As important, the CBO shows — and this is its real contribution — why many recent studies may not be relevant to today’s proposal. The reason: The proposed increase is much “larger than most of the increases that have been studied.” Even after inflation, it would likely be about a third. Moreover, the minimum would be indexed to inflation, rising automatically with prices. This, too, is new.

All these differences suggest larger job effects, says the CBO. Cutting jobs involves one-time costs and disruptions that companies may avoid for small increases in the minimum — but not for big increases. Similarly, more workers would be affected than in the past (about 10 percent of workers compared with 5 percent for increases since 1980). Finally, indexing the minimum wage to inflation implies a permanence that may inspire firms to make deep cuts in labor costs. Companies won’t hire unless new workers are profitable.

Samuelson went on to tout the merits of an EITC expansion, observing that it “would eliminate hiring disincentives and focus benefits on the poorest workers.” Research by economists David Neumark of UC-Irvine and William Wascher of the Federal Reserve has found that the EITC reduces poverty (in Neumark’s words) “not simply through the payment of the credit, but also through the incentives that the credit creates to work more.” Put another way: “Even before we account for the credit payment,” the EITC establishes “pro-work incentives” that “lead to more families earning their way out of poverty.”

To be sure, the EITC is hardly perfect. For one thing, it is exceedingly complicated; for another, it is plagued by high levels of improper payments. According to a 2013 report by the Treasury Department’s Inspector General for Tax Administration, the IRS made close to $62 billion worth of improper EITC payments in fiscal 2012 alone. Indeed, improper payments amounted to anywhere from 21 to 25 percent of all EITC payments that year. Meanwhile, the participation rate among those eligible for the credit was estimated at 78 to 80 percent.

Like other means-tested anti-poverty programs, the EITC also imposes marriage penalties and steep effective marginal tax rates (EMTRs) on certain beneficiaries. “Single parent households that receive the EITC face some of the highest (positive) marginal income tax rates in the United States,” concluded a 2010 Richmond Fed study. The ideal EITC reform would therefore seek to reduce or abolish all marriage penalties, mitigate the EMTR problem, and simplify the entire program.

Despite its flaws, the Earned Income Tax Credit remains one of America’s best vehicles for fighting poverty. Rather than jeopardize potentially hundreds of thousands of jobs by hiking the minimum wage to $10.10 an hour, Congress should reform and expand the EITC instead.