What Would a Basic Income Actually Cost?

Recently Elizabeth Stoker and I wrote about the poverty reducing potential of a universal basic income at The Atlantic. We show that sending every person a $2920 check each year would cut official poverty in half. Yesterday, Danny Vinik at Business Insider and Annie Lowrey at the New York Times also wrote pieces about the potential of a basic income. If this continues, a big trend piece will be in order soon.

When folks write about the basic income, they inevitably talk about how much it will cost and provide a dollar figure or a fraction of GDP. But when they do so, they are not always clear on what exactly they mean by "cost." Obviously, a big dollar figure will show up on the federal budget for a basic income, but in what sense is that a cost? In particular, I am curious as to what Annie Lowrey has in mind when she writes this in her piece yesterday:

There are strong arguments against minimum or basic incomes, too. Cost is one. Creating a massive disincentive to work is another.

This bit of text tells us at least two things about Lowrey's thinking. First, she thinks cost is a strong argument against a basic income. Second, and more importantly for our purposes here, she thinks the cost argument is something different from the disincentive to work argument. These two points taken together seem to reveal some confusion.

To explain why, let's first distinguish between two meanings of "cost" in this context. The first meaning understands something to be a "cost" if it trades off with some other government spending. So a dollar spent on a basic income is a cost because the government could have spent it on something else (let's say a road). But crucially, under this definition of "cost," a dollar that goes untaxed is also a cost. The biggest cost to the government each year is thus the amount of tax revenue it foregoes by not jacking taxes to their revenue-maximizing levels. That policy costs the government massive amounts of money each year because the dollars foregone trade off with using the dollars on some other government spending (let's say a road).

This is not some sleight of hand either. Suppose the government passes a law tomorrow that taxes me $1 each year and passes another law that gives me $1 each year. Does this policy cost the government $1? Sure. It could take that $1 it gives to me and spend it on something else. If the government changes the policy to repeal the $1 tax and $1 benefit, we find ourselves in exactly the same situation. The government is undergoing a cost of $1 by not taxing me. But would anyone ever endeavor to argue without any real explanation, as Lowrey has done here, that one strong argument against the policy of not jacking taxes to their revenue-maximizing levels is its enormous cost? I doubt it. It's certainly not a strong argument, as Lowrey claims.

So this brings us to the second meaning of "cost" in this context, which is a reduction in economic output. So, for instance, we can say that a basic income would "cost" the economy because it would, relative to having no basic income program, reduce overall GDP. How might it do this? By reducing the labor supply. Both the tax increase and the benefit increase would probably cause some people to work less and therefore to produce less. The possibility of a huge fall in labor supply and the cost to GDP it entails could be reasonably considered a "strong argument" against a basic income (or really an argument against making the basic income too high).

But Lowrey cannot mean "cost" in this sense of the word. In this sense of the word, "cost" is literally the same thing as "disincentive to work." But recall, Lowrey has said cost is one argument against it and disincentive to work is a totally separate argument against it.

So what does she actually mean by "cost" then? If she means "cost" in the sense that it trades off with the government spending that dollar on something else, then that argument would also entail that cost is a strong argument against the policy of not upping taxes to their revenue-maximizing level. Surely she doesn't think that. If she means "cost" in the sense that it may reduce labor supply and thus reduce GDP, then it does not make sense for her to have "discentive to work" as a separate point. Like I said above, something is confused here.

How to Think About Basic Income Costs

As always, when you are thinking about a big macro program, the best way to think about it is to ignore money and dollars altogether. The main thing a basic income does is change the distribution of real resources. Implementing a basic income moves real resources (on net) down in the income distribution. That is all that is going on. The distribution of stuff is being altered so as to be more even.

When a government runs a basic income program, it is not actually commanding real resources. It is not like when a government makes a road. The making of that road pulls resources out of private use and directs them towards making the road. Here, the government just moves real resources around. It just changes the distribution of them. It does not direct their use.

Thus there are only two meaningful costs to a basic income program. The first is the cost of administering the program. The government would have to pay people to collect taxes and write checks. As Social Security shows, this can be done really efficiently. This is a real cost because the people doing these tasks could be doing other tasks instead. The second meaningful cost is the cost to the economy from a reduction in labor supply. This cost is the factor to look towards when determining how big to make the basic income.

Other than administration and labor supply effects though, the basic income doesn't cost anything, not really. With a basic income, we are talking about changing the distribution of income, not undertaking a government project like a road that actually commands and directs real resources. If you want to talk about its "cost" when referring to its budget line item, that's fine I suppose. But saying a big line item will show up on the budget is not a "strong argument" against it, and we really should not let ourselves get confused into thinking that. The word "cost" has many meanings, and by failing to distinguish between them, we can easily make serious mistakes when analyzing the basic income and cash transfer programs in general.