Every once in a while folks in the political corner of the blogosphere start talking about Hayek’s argument in The Road to Serfdom. As Matt Yglesias said Monday, lots of people, conservatives and liberals alike, say that Hayek believed that any welfare state inevitably leads to totalitarianism. Then some people who have actually read Hayek reply that he always supported social insurance, safety nets, public goods provision and many forms of regulation. Then confusion ensues. In this post I want to address some very recent wrangling on the subject and then try to explain how Hayek could consistently warn against the dangers of the welfare state and simultaneously support one.

I. The Present Discussion

To begin, consider Henry Farrell’s recent post at Crooked Timber following up on a similar discussion a few months prior. He seems prepared to insist that, contra Tyler Cowen, that Tony Judt had not been unfair to Hayek when Judt said:

Hayek is quite explicit on this count: if you begin with welfare policies of any sort — directing individuals, taxing for social ends, engineering the outcomes of market relationships — you will end up with Hitler.

Despite Hayek’s repeated claims that he does not think “welfare states” lead to totalitarianism, and Farrell’s repeated recognition of Hayek’s claims to this effect, Farrell insists that “I don’t think Judt was being unfair at all.” In a follow-up post, after reading a similar (but more academic) discussion between Andrew Farrent & Edward McPhail and Hayek Scholar Bruce Caldwell on the other, Farrell declares:

I’m calling this one unequivocally in favor of Judt – contra Tyler Cowen, he wasn’t being unfair at all.

Obviously Farrell and Judt’s claims are over the top due to their use of various “of any sort” “unequivocally” “at all” and “Hitler” modifiers. But instead of beating up on them, let’s use our collective annoyed-by-someone-on-the-internet energy in a constructive fashion: to see what we can learn about Hayek’s real arguments against socialism and the welfare state.

In his piece, Caldwell sharply distinguishes Hayek’s arguments against socialism and the welfare state (following Hayek’s lead). Hayek thought that the planned economy would lead to totalitarianism. Caldwell concludes, rightly, that Hayek was right about this. But he points out that Hayek’s criticism of the welfare state is subtler and involves two claims. The first problem with the welfare state is that it is a philosophically slippery target. As Hayek says in The Constitution of Liberty,

We cannot argue similarly against the welfare state, for this term does not designate a definite system (259).

But Hayek also argues that the aims of welfare state reformers can gradually lead to socialism because as Caldwell says interpreting Hayek,

The death of liberty is that of a thousand small cuts, each aiming at correcting some apparently flaw in the system. This is a very different argument from the one in The Road to Serfdom, and one should not mix them together (12).

So when Farrell reads this, he concludes that Hayek basically made the same claims about the welfare state and socialism, namely that both institutions will lead, eventually, to totalitarianism, even if the socialism gets us there sooner than the welfare state.

And if you cursorily read Hayek’s later writings, it looks that way.

But wait. In my last post, I pointed out that even the later Hayek defended a universal basic income, a policy considered solidly welfare statist. And he even supported the UBI as a condition of democratic legitimacy, not merely as a pragmatic measure. Thus, Hayek supported what we typically call a welfare state throughout his career.

Farrell thinks that Hayek argued that the welfare state will lead to Hitler and since it obviously doesn’t, that Hayek made a “rather idiotic claim.” But if you’re really itching to say that Hayek said something dumb, this is a poor example. Obviously something strange is afoot.

Of course, we could always follow Matt Yglesias, and throw up our hands:

The answer to this puzzle, I think, is that Hayek was inconsistent.

But that would be a major inconsistency, right? That Hayek would consistently warn against the dangers of the welfare state while simultaneously supporting one on moral grounds? It can’t be Hayek hagiography to wonder if he didn’t have some way of rendering these positions consistent. It seems to me that basic charity requires us to look for a resolution.

II. The Welfare State of Law and Welfare State of Administration

As I claimed in my last post, Hayek criticized social justice only as a claim that we can make specific evaluative judgments about whether certain domains of goods and services are justly distributed. Hayek had no problem evaluating a society’s basic structure and supporting a universal basic income to make that structure more legitimate. He explicitly says so.

In my view, then, Hayek’s target is not “the welfare state” as such, that is, not a social insurance or safety net state, but rather a state based on a robust conception of distributive justice applied to its economic components, not the system as a whole. Specifically, Hayek is concerned only about coercive redistributive institutions that permit constant state tinkering, not all coercive redistributive institutions.

This interpretation makes a lot of sense in light of the fact that Hayek insists, over and over, that law be based on general principles rather than expediency. Expediency is bad because it gives law over to administrators who, through many subtle alterations, will destroy a liberal legal order. Of course it will take time for this to occur, but it’s not hard at all to find insight in Hayek’s argument. A welfare state of bureaucratic tinkerers will lead gradually to socialism simply as part of the logic of their power. If bureaucrats can intervene to set coal miner’s wages, why not grocery store clerks? If they can subsidize corn, why not rice? If they can award patents to software engineers, why not musicians? If bureaucrats are pursuing a particular, patterned goal of, say, fair or equal wages, then we can see how the slide to serfdom might go.

Hayek’s critique of the welfare state simply falls out of his broader conception of the legal order of a free people. If you have a patterned principle of distributive justice, one that would license a welfare state of tinkering, then you’re going to have to constantly interfere with liberty in increasingly objectionable ways to get the distribution right. That is, the problem with certain kinds of welfare states is that they are one of many ways in which the strictures of general rules are relaxed.

So let’s distinguish between two kinds of welfare states: the welfare state of law and the welfare state of administration. Hayek’s preferred welfare state is limited by his insistence that the law be regulated by clear, public, general principles rather than administrative bodies. That’s why his safety nets are so general and uniform: because safety nets should follow these same general principles. In this way, Hayek endorses a welfare state of law.

Hayek opposes the welfare state of administration. First, it’s a vague target because it’s not exactly clear what principles it is based upon. That is, its general principles are insufficiently public. That’s one way (not the only way) that it can get away with tinkering, because the public lacks access to simple, public principles by which to evaluate and restrict the actions of welfare state bureaucrats. This opens the door to increasing social control (in the absence of ideological opposition).

But the second problem with the welfare state of administration is that it contains an internal dynamic that pushes in a socialist direction. Hayek argues in LLLIII, following Mancur Olson, that welfare states generate the accretion of regulations and interest groups. Of course, this is not totalitarianism by any means. For one thing, if citizens affirm even modest economic freedoms (as most members of liberal democracies do), then they will resist this accretion effect before things get too bad. And that’s the pattern we see: even in Scandinavian countries, people resist regulation due to their concerns about efficiency and, yes, concerns about property rights (sometimes more effectively than we supposedly libertarian Americans).

So in my view, Hayek identifies a unique social dynamic, one whose operation can be slowed but that if left unresisted would drag us to serfdom.

Hayek overplayed his hand by arguing that the tinkerer’s welfare state will inevitably lead to totalitarianism, but not by much. The most free and economically successful liberal democracies hybridize welfare states of law and welfare states of administration. They’re hybrids largely due to the fact that most citizens of liberal democracies endorse elements of both liberalism and socialism.

But if citizens of liberal democracies gave up liberalism entirely and stopped minding regulation so much, then I think the dynamic of the administrator’s welfare state would lead to significant authoritarianism that, while not totalitarian, would be uncomfortably close. Just think about India’s license Raj or the ridiculous, strangulating business regulations in many developing nations. Aren’t we serfs if we’re forced to wait fourteen years to start a business? Aren’t we serfs if our setting prices is subject to constant scrutiny? Aren’t we serfs if our medical choices are strictly controlled by the state? Aren’t we serfs if our government can swoop in at any time and nationalize our businesses? One doesn’t have to be a libertarian of any sort to answer yes.

But contrast this with being forced to pay a modest share of your income to provide a universal basic income, with clear, simple tax rates and clear, simple funds provided to those who fall below, again, a clear, simple, public threshold. That sort of welfare state does not seem to make us serfs or lead to worse forms of serfdom.