Encouragingly, last week's dive into the evidence on work effort and basic income schemes has sparked a bit of follow-on discussion. While I was attempting to rebut the conservative argument that giving everyone a certain amount of money would reduce peoples' willingness to work, a lot of the criticism that followed has come from lefties. Here are three of the most thoughtful replies, and why I don't think they kill the idea.
What is the plan, even?
Kevin Drum is skeptical due to the lack of details being proposed. I didn't lay out a formal proposal in my initial post because (a) it was meant as a response to criticism of the idea as a whole (b) there are so many that have been proposed and (c) I'm not an economist equipped to do revenue or budgeting estimates. (b) is particularly important. Drum asks, "How would it phase out with income?" and the answer is honestly, it depends on what plan you're looking at. Negative income tax proposals would. Pure basic income proposals wouldn't.
But it's a fair request. So here are three model proposals, and rough cost estimates.
Negative income tax
If you want a negative income tax proposal, you could do worse than looking at the Family Assistance Plan of 1971, a bill supported by the President and which passed the House. It's not just a decent model proposal; it almost was the law of the land in the United States. That provided about $10,000 for a family of four and taxed it back at a 50 percent phase out rate; it included a work requirement, which is unorthodox for negative income tax proposals, but probably a realistic feature of any one that'll come close to passage.
Pat Moynihan, one of the designers of the plan on President Nixon's staff, estimated that the negative income tax itself would have cost $5 billion then ($30 billion in today's dollars), not including supplemental welfare spending. That's implausibly low, and Gary Burtless' estimate of the additional cost of a negative income tax set at the poverty line of 1.5 percent of the economy (or $200-300 billion a year) is probably closer to the mark. All the same, that's hardly an unimaginable figure.
Basic income advocate Allan Sheahan proposed a $10,000 per adult, $2,000 per child basic income back in 2006 and laid out specifically what he thought it would cost, and how to pay for it. The cost comes to around $1.9 trillion a year. That's over 10 percent of GDP, and would constitute a rather massive expansion of government. He proposes paying for it by eliminating basically every tax deduction in the book — including the mortgage interest deduction, the charitable deduction, the standard deduction and personal exemption the Earned Income Tax Credit (EITC), and the child tax credit — going back to 1994's tax rates, subjecting all wages to payroll taxes, and a bunch of other tax increases.
This is an extremely ambitious proposal; that much is beyond dispute. But he's hardly the only person arguing for a 10 percent of GDP expansion of the welfare state. Lane Kenworthy, an sociologist at the University of Arizona, released an outstanding book last year, Social Democratic America, laying out an expansion of the welfare state exactly that big, which he argued would be necessary for the US to become a true, European-style social democracy; I interviewed him about the details here.
If you (like me) accept that basic argument, that a government capable of meeting human needs must be significantly larger than the one the US has now, then it's worth considering how best to structure that expansion. I think a basic income makes as much sense as any other idea for how to do it.
Negative income tax (for conservatives)
Conservative welfare critic/scientific racist Charles Murray wrote a book in 2006, In Our Hands, laying out what a conservative/libertarian basic income would look like. His plan has no child allowance and gives $10,000 a year to everyone over 21, but unlike Sheahan, he wants to keep the government the same size or smaller. It has a 20 percent phaseout rate, making it a negative income tax, formally speaking.
He accomplishes this by gutting much of the rest of the welfare state. He'd get rid of Social Security, disability insurance, workers' compensation, unemployment insurance, Medicaid, S-CHIP, veterans' health care, school lunches, housing assistance, food stamps, the Earned Income Tax Credit and child tax credit (this he and Sheahan have in common), Pell Grants, and Head Start, among many many other programs. That doesn't quite pay for it, so he throws in some other budget cuts too. The result is still more expensive than the current set-up, but he accepts that, given that basic income's costs will grow more slowly than a lot of the welfare state, especially medical benefits.
Benefit cliffs and inflation
Stanford med school's Keith Humphreys is one of my favorite policy analysts, and his critiques on any topic are worth taking seriously. He has two on basic income plans. The first is that a plan only targeting the poor risks creating a benefit cliff: if one makes too much to qualify, they'd lose benefits, which creates a work disincentive. Here's Humphreys:
Imagine someone who is working hard 2000 hours a year at $11/hour, generating an income of $22,000. Then the government creates a minimum income of $20,000, which the person could receive if s/he stopped working. The person is now working 2000 hours a year for an extra $2000, i.e., for $1/hour.
Some people would keep working at that low effective wage because they like their jobs or they think they will move up someday or they just believe in the moral value of work. But other people — potentially most people — would be glad to free up 2000 hours a year at such little cost.
A minimum income that only kicked in below a certain income threshold or which only applied to those not working would be, indeed, a horrible, horrible idea. But I know of no basic income advocate who supports that. The idea is for the cash grant to be either unconditional (basic income plans) or phased out gradually to avoid horrific work disincentives (negative income taxes). The negative income tax experiments worked a little differently, since researchers didn't have enough resources to give everyone in a community money (and then tax it back), regardless of their income. So researchers wound up only working with low-income people in most cases. But just because that's a next-best way to do a demonstrative experiment doesn't mean it's the way the policy has to be structured — and indeed, only giving the grant to poor people would be a mistake.
Humphreys' second criticism troubles me more. He notes that a basic income would likely require low-wage employers to bid up wages, to convince workers that staying employed is worth it even when they don't have to to avoid starvation. The first-order effect of that is very positive indeed. It's good for low-income workers to get a raise! But Humphreys notes that it could promote inflation for the kind of consumer goods that low-income workers rely upon disproportionately; after all, low-wage employers need to fund those raises somehow, and price increases are one likely mechanism.
I have two broad responses here. One is that I think the government agency that should be tasked with worrying about wage inflation is the Fed, not Congress; as Danielle Kurtzleben explains here, the Fed is plenty comfortable using its rate-setting powers to tighten when it suspects wages are rising too quickly and jeopardizing price stability. Indeed, the Fed has arguably been overly aggressive about this over the past thirty years, to the detriment of workers' wage growth. I think a tighter Fed policy in combination with a basic income should be able to capture most if not all of the benefits of the policy while avoiding this particular problem.
The second response is that there's an upper limit to how much a basic income can bid up wages. If McDonald's finds that the only way to hire people to make hamburgers is to pay them $100 an hour, its response will not, in fact, be to pay people $100 an hour and quintuple the price of a Big Mac. Its response would be to start automating its food production more — perhaps using one of Momentum Machines' hamburger-making robots, perhaps another company's, perhaps developing technology of its own. Not only does this put an upper limit on the wage inflation problem, it serves a valuable policy purpose in its own right, by serving as an incentive to develop labor-saving technological advances.
What is the point?
Max Sawicky argued against a basic income from the left when Ezra and I put the topic up for debate at the Wonkblog debates in November, and put his notes up here, which he's reupped in light of recent discussion on the topic. Sawicky's case is extremely thoughtful, and helps clarify the deeper philosophical reasons for liking a basic income.
Sawicky notes that there's little chance of a basic income in its pure form ever being implemented. True enough, though I think the Nixon administration experience suggests it's possible we'll again enjoy a political climate in which a negative income tax could plausibly pass. I also think a tiny pure basic income could potentially pass in today's political climate. Imagine if we converted the standard deduction and personal exemption to refundable credits. That could be done in a revenue-neutral way, and would amount to a small basic income (albeit one financed on the backs of middle class taxpayers, with the rich paying nothing). Indeed, the only pure basic income proposal to ever be proposed in Congress (by then-Congressman, now-disgraced-former-San-Diego-mayor Bob Filner) was a refundable $2,000-per-adult, $1,000-per-child credit. It's not hard to imagine the Democratic caucus getting behind something like that.
But Sawicky is right that arguments over a basic income are more of a conceptual exercise than a concrete political one. "The way to read the UBI is not as a real proposal, but as a message about something else: our existing system," he writes. "But the implicit critique of the existing system underlying UBI advocacy is not well-founded." That critique, Sawicky says, is that the existing system involves too much overhead, or is too inefficient.
I suspect I'm more sympathetic to those claims than Sawicky is, especially with regards to housing programs which for whatever reason seem more random in benefit distribution than most, but he's right that most of our welfare state, including Medicaid and Medicare and Social Security and food stamps, is actually very efficient, and fairly well targeted. And I think most basic income fans would, like me, grant that point. Food stamps are awesome. Social Security is a wonderful program that should be expanded. But Sawicky is right that there's an implicit critique at work in basic income proposals. The critique, however, isn't of overhead. It's a critique of paternalism.
Part of what makes Social Security so great is that it gives you cash. It was set up to address a concrete problem — senior poverty — and chose the simplest way to do that, affording beneficiaries the most possible choice over how to use their benefits. Now suppose that, as part of a debt deal or some such, it was decided that Social Security would not be distributed this way. Instead, everyone on Social Security would get a food stamp equivalent, a Section 8-style housing voucher, and a slew of other in-kind provisions that add up to the cash value of their old Social Security income.
No would stand for it. Who is the government to decide how we should spend our Social Security checks? What if I bought a house outright over the course of 30 years and don't need the housing vouchers? Why can't I use that for food or to pay the utility bill instead? Why is this better than cash?
The answer is that it isn't better than cash. It's worse, in every way, for the people getting it. It's less capable of being tailored to their particular wants and needs. And so it is for other welfare programs. For programs involving adults who are capable of caring for themselves, it makes sense on epistemic grounds to default to giving cash rather than specific goods. The idea that politicians who have, for the most part, never endured financial hardship in their lives know what kind of assistance poor people need better than poor people themselves do strikes me as very foolish indeed. You could make a deeper moral argument here too, if you're not so utilitarian in outlook. In some ways, in-kind provision is an assault on the dignity of poor recipients. It's the government telling them that it doesn't think they're capable of taking control of their own lives. It's a real restriction on their autonomy.
But the funny thing is that I think, by the end of his essay, Sawicky appears to come around to some form of negative income tax. If you want to give people cash, he writes, "federalize TANF [Temporary Assistance to Needy Families] and establish it as an individual, adequate cash payment to which every resident has a legal right. To constrain its cost, limit eligibility to families with dependent children and phase it out as other income grows." I don't much care for leaving childless people out of the benefit, but otherwise, that strikes me as a very fine proposal indeed. It seems we basic income supporters are winning converts after all!