Suppose we want the government to help poor people. I think most of us do want to do this, to some degree.
How do we best go about doing it? Economists have been debating this for centuries, and the debate isn't about to end soon.
But one solution in the U.S. has stood out in recent decades as the most efficient method: the earned income tax credit, known by its even less elegant abbreviation, EITC.
First, some background. The main problem with just giving poor people money is that it tends to discourage them from working. There are at least two reasons we'd like people to work - first, work produces useful stuff, and second, work gives people a sense of dignity.
If you simply mail everyone below the poverty level a check, you're providing a strong incentive for people not to work, because working would put many people over the poverty line. At that point the checks would stop coming.
An alternative is the minimum wage. This has the benefit of rewarding workers, while removing the incentive not to work.
But minimum wages, like all price controls, bring distortions, locking some people out of the job market and pushing up unemployment. Maybe this distortion is small, in which case minimum wage hikes like the ones being implemented by Seattle, New York and other cities won't cause unemployment to rise very much.
But there is still going to be some limit to the amount that minimum wages can be raised before some low-skill people get thrown out of work.
There is, however, a third option, which is more subtle and complicated than welfare payments or a minimum wage.
This is the so-called negative income tax, a policy invented in the 1940s by a British politician and popularized by the famous American economist Milton Friedman.
Just as income taxes take away a certain percentage of your earnings, negative income taxes mean the government pays you a certain percent of your pay, on top of what you earn in the market. And just as income taxes discourage people from earning more, negative income taxes encourage people to earn more.
Of course, negative income taxes have to be phased out at some earnings level; otherwise there's no way to raise money for the tax! Not to mention that Warren Buffett doesn't exactly need poverty relief. But for poor people, the negative income tax provides money to live and an incentive to work. It really is a Goldilocks policy. Created in 1975 and expanded a number of times since then, the EITC is a big win for economic policy engineering.
There is still the question of how strong the work incentive is. Will people really toil more just to earn a few hundred extra dollars from the government? Some new evidence suggests that the incentive is large.
Using the quasi-experimental techniques that have recently become the gold standard in empirical economics, Hilary Hoynes and Ankur Patel show that increases in the EITC pay off big and that even modest increases in EITC payouts cut poverty while boosting employment. That confirms the basic result of previous studies.
So the verdict is in: The EITC accomplishes both of its goals - it gives poor people the money they need to live, and it encourages them to work.
The EITC does have one major problem: the way it's administered. EITC is embedded in the tax code, which means that you have to file a return in order to claim it. That involves paperwork and hassle.
It also means that poor people, who often don't file taxes at all (because they earn too little), will be unfamiliar with the tax system and therefore fail to claim the credit. It also means that many people won't even be aware that the EITC exists, since the Internal Revenue Service doesn't exactly do a good job of educating the public about how to use the tax system.
For this reason, the Tax Policy Center recently found that 26 million Americans are eligible for the EITC but fail to claim it.
One solution to that problem might be to restructure EITC as a wage subsidy - in other words, to have employers claim the credit and pay it out in the form of higher wages.
The problem here is that many poor people tend to work multiple part-time jobs, which will make the system overpay them because it will treat each part-time job as the recipient's only source of income. A better solution may be to send more workers and volunteers into poor communities to make sure everyone knows about the EITC.
In any case, the verdict of both theory and evidence is clear: The EITC is a good policy. The U.S. should use it more.
Noah Smith is an assistant professor of finance at Stony Brook University and a freelance writer for finance and business publications.