Right Reason

A Rant on Deceptive Market-Failure Arguments

 

 

 

 

Over at the Niskanen Center, Samuel Hammond has written one of the worst articles I’ve ever read on mutual-aid societies and the welfare state. So let’s begin my rant:

The conventional libertarian argument has always been a bit hand-wavey, pointing to eras in history where, yes, government was smaller and mutual aid societies were bigger, but where poverty was nonetheless widespread and abject.

Yes, people were a lot poorer in the late 18th and 19th centuries. By our standards, poverty was indeed "widespread and abject" during that period because society was *a lot* poorer. We're simply not comparing apples to apples when we compare the effects of a 1950's style welfare state with 1890's style mutual-aid societies. A lot has changed in the meantime, including a substantial increase in the standard of living having very little to do with either the welfare state or mutual aid societies.

It would have been honest to point out that poverty would have been "nonetheless widespread and abject" even if there were a welfare state. Alas, as we shall see below, there is a lot of deception going on in this article.

Furthermore, we should count this post as yet another example where someone uses 'market failure' as pseudo-scientific code for 'markets not producing what I want:

Thus the reason all modern economies have some sort of pension system is not because private charity comes up short, or to steal from the rich to give to the poor. That’s a category-mistake. Rather, it’s because there is endemic market failure in the private annuities market caused by adverse selection and information asymmetry (among other things, you have a much better idea of how long you’re likely to live than an insurer).

Before I even discuss the theoretical problems with this sentence. Let me point out that it is obvious Mr. Hammond hasn’t even read the source he cites to support his market-failure hypothesis. To lend academic credence to his argument, he cites Jonathan Gruber’s Public Finance and Public Policy, who argues that the market-failure rationale is not the reason why most modern economies have social-security programs: “While annuities market failure is the classic economic rational for Social Security, the true reason that most policy makers favor the program is pateralism; that is, they are concerned that individuals won’t save enough for their own retirement.” So clearly Mr. Hammond couldn’t be bothered to read his source when he claimed why these kinds of pension-systems exist!

Now onto matters of first principles: Markets are non-teleological. That markets evolve towards ends no one intends is a basic lesson of economics that should be drilled into every high-schooler’s mind by teachers so that citizens don’t fall into this fundamental error.

Being non-teleological systems, they don't fail when they don't accomplish what we want them to accomplish. Instead, hey fail when they systematically don't realize potential mutual benefits to specialization and trade. What Mr. Hammond describes isn't a market failure because it simply isn't profitably for insurers to insure everybody. That isn't a market failure because there isn't mutually beneficial trade being left on the floor. It's just that the mutually beneficial trade is exhausted before Mr. Hammond is willing to say the market is doing whatever job he believes he has the authority to tell the market to do.

In a perfectly free market for insurance, some people might not be able to find insurance. But the reason why they cannot find insurance is probably because there is no mutually beneficial opportunity for them to buy insurance. Free and self-interested agents want to manage their risk-pooling so that they maximize their own benefits. Some people may therefore be too unhealthy to ever find a health-insurer willing to insure them or may be too dangerous of a driver for a car-insurer to do the same. That those high-risk people are not insured is not an evidence of a market failure. It’s evidence that there’s no opportunity for mutually beneficial trade.

 

 

Mr. Hammond calls for the government to fix something he perceives as wrong in the markets. This is an example of what Nassim Taleb has called lecturing birds on flying. More specifically, Mr. Hammond is lecturing birds on how they should not be able to fly. HE comfortable lecturing us about how insurance is broken, and clearly has ideas about what insurers should be doing. But, as of my knowledge, he has no actual experience in the market for insurance. If Mr. Hammond, and others really think that there are $20 bills strewn across the floor in the market for insurance, they should become entrepreneurs and use their knowledge to get rich by picking up all of that money.

However, we clearly see that Mr. Hammond is not actually all that good about thinking about insurance when he makes an attempt at demonstrating how asymmetric information might destroy the realization for opportunities for mutually beneficial trade. He writes: "among other things, you have a much better idea of how long you’re likely to live than an insurer." However, there’s no reason to actually believe this. Insurers probably actually know better than most people how long they will be likely to live because they know how to analyze actuarial tables. Whereas most people are likely to self-delude themselves as special when computing how long they are likely to live, insurers will just use an actuarial table and will probably get a better answer!

Of course, Mr. Hammond may proclaim that the problem is insoluble, but I’m not particularly convinced. Just because someone with no experience in a market can write some article about ‘moral hazard’ and ‘adverse selection’ does not mean that entrepreneurs don’t solve those problems on a daily basis, even if those articles had called the problem insoluble.

 

Mr. Hammond may want there to be insurance provided to people deemed extremely risky by the markets. He may want different outcomes than the ones generated by the market process. However, that isn’t evidence of a market-failure. A market-failure would imply that there are Pareto improvements to be had, here there is not.

As used, the entire language of market-failures is deceptive: Under realistic assumptions, there isn’t a Pareto improvement to be had here. Some people *are* going to be made worse off and, in a free market with exit options (I repeat myself), those people would have an incentive to leave the ‘social insurance’ scheme and create their own association that pools their low-risk people together. When those people are pooled in with higher-risk people, they are going to have to pay higher premia and are going to be worse off. The language of ‘social insurance’ and ‘market-failure’ obfuscates that some people will be worse off, and that maybe those people want a better deal than the one Mr. Hammond wants them to accept.

 

I could then rant about how the author mistakes regression estimates as averages and not as differences at the margin, but that would be clearly getting too pointy-headed for a criticism of think-tank's blog post. Looking at a regression coefficient as an average and then using it to compute a counter-factual at the scale of the entire whole is simply a gross abuse of statistical reasoning. In a complex system such as human society, regression coefficients only work at the margins because of a ceteris-paribus assumption. However, when we assume that an entire magnitude is going to disappear, there simply isn’t any certeris to hold paribus. Everything is going to change and the statistical model is no longer going to be a good estimate.

 

A Hayekian Welfare State?

 

What goes under [the name of the welfare state] is a conglomorate of so many diverse and even contradictory elements that, while some of them may make a free society more attractive, others are incompatible with it or may at least constitute potential threats to its existence.

-Friedrich Hayek, The Constitution of Liberty

 

In its January issue, Econ Journal Watch held a symposium on economists’ opinions on the welfare state, titled “Economists on the Welfare State and the Regulatory State: Why Don’t Any Argue in Favor of One and Against the Other?” In response,  Sam Bowman of the Adam Smith Institute voices his sympathy with Andreas Bergh’s contribution (no, not that Andreas Bergh): “Yes, There Are Hayekian Welfare States (At Least in Theory).” In “Let’s have a Hayekian welfare state,” Mr. Bowman writes: 

[W]e should have a predisposition against regulation, even regulation that appears to solve problems, if it holds people back from experiments. There are also serious examples of regulations leading to bad things that are even worse because everyone has been forced to make the same mistake, which strengthens this predisposition even more. The more complex a system is is, the more we should value pluralism.

All of this has to do with having limited knowledge in a complex world, not incentives, though of course there may be good incentives-based arguments to be made on a case-by-case basis against certain regulations.

But this doesn’t tell us very much about the distribution of wealth in a society. To use Bergh’s terminology, redistribution may be something that can be done with relatively low amounts of knowledge. That doesn’t mean that it can’t fail – clearly it can, very easily, if the level of redistribution is set too high (or too low) – or that the system itself be badly designed.

The particular distribution of wealth in an economy may be an efficient reflection of who is most productive, and interventions that try to correct for that are likely to fail for the same reasons that other interventions designed at improving market efficiency will fail.

But we may have non-economic concerns about the distribution of wealth as well. An economy in which everyone is paid according to their productivity may be very brutal for people who are not very productive and cannot change that. We may wish to redistribute income for their welfare.

 

Myself, I’m sympathetic to Messrs. Bowman and Bergh’s notion of a Hayekian welfare state. I certainly believe that the most humane way of helping the poor politically today is by ending policies that have put up walls to them improving their own stations in life, whether those walls be vocational licensing or the war on drugs. Nevertheless, in a nation as wealthy as the United States, a welfare state can advance the general welfare by helping those who would otherwise down and out. The welfare state describes such an institutional means. 

When advancing a Hayekian welfare state, we have to be careful to keep in mind that the welfare state is not beneficence. Politicians who claim to be beneficent in their support of welfare policies are lying. Beneficence must be freely given. In the welfare state, it is not and politicians would have nothing to give if it weren’t for the tax revenue, at the end of the day, coerced from the general population. 

Nevertheless, keeping in mind that the welfare state is not an instrument of beneficence, I am sympathetic to the Hayekian welfare state, that is to a welfare state existing side-by-side to a minimal regulatory state. Importantly, Messrs. Bowman and Bergh argue, in my mind successfully, that there is a plausible economic reason for opposing the welfare state as there is for the regulatory state. Instead, the issue is one of morality and expediency. On both counts, there is a case for the welfare state.

 

On this issue, I think that contemporary libertarians, otherwise opposed to any welfare programs, should take a Whiggish view on the matter. They need to accept the political changes around them, even if their hearts may not be fully on the side of those changes. The existence of the welfare state is something that liberal political theory needs to adapt to, if it is to be relevant. Concerns about knowledge do matter in providing a, shall I say, efficient welfare state. There is much the Hayekian position can add to welfare policies. Taking a pig-headed opposition to it as the sole acceptable position only serves to cut off the value of that position from political discourse.