A Reductio Against Universal Basic Income

Today, while reading Bryan Caplan's list of "common sense" policies that no presidential candidates are proposing, I had a thought about proposals for a "universal basic income" (UBI).

Libertarians often propose direct transfer payments as economically efficient alternatives to traditional welfare. Thus, the so-called "libertarian case" for a universal basic income involves scrapping all existing welfare programs and replacing them with the equivalent per capita universal basic income. In other words, if it costs the country, say $1000 per capita to deliver $500 in welfare benefits to qualifying citizens, we could improve economic outcomes by delivering $500 in cash to qualifying citizens at less than $1000 per capita. This is achieved by (1) eliminating the administrative burden of submitting welfare applications, (2) eliminating "means testing," i.e. investigating welfare claims, (3) cutting the workforce at the welfare departments, (3) passing the savings onto the taxpayers, and (4) empowering welfare recipients to spend their money however they think is best.

You might not agree with the above argument -- I sure don't -- but that's how it goes.

This underlying argument is often deployed on a much smaller scale. For example, when people debate steel tariffs, economists are quick to point out that steel tariffs cost American consumers hundreds of thousands of dollars per steel worker. If we really wanted to use policy to protect the steel workers, we'd be better off scrapping steel tariffs altogther and then paying each steel worker, say, $150,000 per year not to work at all. It sounds radical, and maybe it is, but the point of the argument is to show how costly steel tariffs are, and how we could lower domestic steel prices and still take care of American steel workers without creating large market distortions that hurt everyone.

In thinking about this, however, it struck me that the libertarian case for a universal basic income can be deployed against policies that libertarians traditionally do not object to. One example is the armed forces. It would be more economically efficient to eliminate the US armed forces and use the Department of Defense's annual budget to pay all Americans an annual self-defense stipend, which they could spend however they deemed appropriate. Some would squander the money on non-defense spending, but others would conscientiously invest in martial reinforcements for their homes and/or communities, thereby eliminating market distortions and saving taxpayers money.

The reason most people reject this proposal is because they broadly believe that the government does a better job providing military protection than home-grown militias do (for a wide variety of reasons). The idea that the government should just shutter one of its essential functions and "pass the savings on to you" suffers from the fatal flaw that it still leaves you without essential services, such as basic military protection against foreign military invasions.

What this reductio ad absurdum demonstrates is that there is at least one case to which the basic libertarian argument for transfer payments must be rejected: the case of "essential government services," whatever that terms happens to mean to you. Obviously, if you're a left-leaning person, you will tend to include more social welfare programs in the list of "essential government services," and if you're a right-leaning person, you will tend to include more national defense and infrastructure programs. In either case, the argument simply falls flat.

And the reason is simple. People don't care about saving money when it comes to essential care. They'd like essential care at the lowest possible cost, of course, but they would still rather ensure that the government spends money and provides essential care than eliminate essential care and disseminate the savings among the populace.

One group of people are clear exceptions to all of the above argumentation, however: Anarchists, especially anarcho-capitalists.