When should the government choose?

Megan at From the Archives has written two pieces on libertarians and choice that are, predictibly, stirring up a lot of angst in the libertarian blogosphere. From the first piece:

People live in denial, do not do good risk analysis (as evidenced by my erratic use of bike helmets.) They do not conscientiously save against medical emergencies, even though they could. They do not have the capacity to compare fancy-dancy medical treatments (I should figure out what chemo regimen is best for me? I DO NOT WANT TO, because that is outside my expertise and BORING. I want to trust an expert, if it comes to that.), especially if the pain has already started. They do not have any interest in comparing not-fancy treatments. (When I broke my arm, I realized I had no information whatsoever on which of the four local emergency rooms had good reputations. None. I had never cared until it was too late.) I derive zero utility from comparison shopping for health care; I want someone else to handle it.

I figure people are roughly like me, non-savers, bad risk assessment, more than willing to delegate their health care. (I am not willing to delegate my fitness or nutrition, but that is different from disease or injury.) You know what makes good sense for that model of the individual? Government based health care that does a decent job by me. You know what doesn't make sense? For profit insurance agencies who do not have my best interests at heart.

And from the second:

This is the other thing I don't get about small government types. You protest so vociferously that government takes choices away from you. But a whole lot of choices are BORING. If I never once think about car bumper safety standards for 25mph crashes, I will never miss it. I do not want to carefully match my car safety standards to my most likely driving patterns and save two grand in the process. I would not enjoy that process. (Perhaps you would, and you would rather have the money.) I've never been a comparison shopper or a meticulous consumer. Maybe my model of the individual is too biased by my experience. But I don't want to figure out how much coliform bacteria I can tolerate on my spinach, given my health. I don't want to do that even if it saves me money. I don't want to figure out what goes into paint in nephews' toys. I don't even want to handle my health care.

People talk about being rational health care consumers, but they are maximizing some combination of health outcomes and money. I want to maximize my utility. My utility is optimized by going outside to play while someone who is interested in health care gets paid to balance my health care and money. I'll pay a little extra to cover that person. I come out well ahead in that deal*.

This is true of almost everybody. My model of the individual is that they do not always make the best choices. (Obviously). Most people are not driving the car, watching the television, or taking the vacation that would make them happiest. Most people pay too much to avoid risk (witness the preferences for low insurance deductibles and extended warranties), assume that anything their neighbours do is probably a good idea, and spend money on expensive features they then do not use. And I think that virtually everyone, from anarcho-capitalist to anarcho-syndicalist, agrees that this is so.

But this is not an argument for turning those choices over to someone else, necessarily. Most people probably do a better job at picking what is right for them, most times, than even a polished expert. Have you ever gone shopping with a friend who had beautiful taste in clothes, only to end up with a closet full of things that you can't quite pull off?

On the other hand, in many cases there may be an argument for turning over the choice to someone else. I neither know, nor care, about the inner workings of my Bose sounddock; I bought it because a friend I respect said it was the best thing short of an expensive stereo kit I couldn't afford.

Megan is making an argument for turning many of those choices over to the government. Libertarians would argue that most of those choices could be safely left in the hands of consumer reports or an equivalent. My question is, when does the government work, and when do private agencies work just as well or better?

Megan works in one of the areas where I, at least, would say the government is probably the best solution: water management. There's a reason that irrigation systems seem to have given rise to the first large states: water is no respecter of property lines. Given the coordination problems and hazily defined boundaries inherent in water rights, I'm pretty comfortable having the government take care of this, though as Megan might put it, I find the topic too BORING to decide whether I think the government is doing it well.

(Sorry, Dad. Try to think of it as my outsourcing my opinions to you. Gains from trade!)

On the other hand, Consumer Reports does a pretty good job with my blender. And what they don't take care of, tort law seems to be pretty good at. Jack in the Box doesn't try to track down the source of e. coli outbreaks in its food because it's afraid that the government will shut it down; it's afraid the government won't need to, because people will already have stopped eating there.

Private ratings firms do have drawbacks. For example, I was told by a marketing professor one story about Yahoo in the early days, which used to sell its search rank order. There was one chap who wanted "Garden City Financial Planner", which wasn't, as you can imagine, a frequently requested search. They checked their listings and saw that that search term was only requested about 25 times a month, and decided to charge him a dollar a search--back in those days, the prices were also set kind of haphazardly. (Note: as I'm telling this from memory, the numbers are very approximate). The next month, he cheerfully came back and reordered. And the next month. So they doubled the price. He went on paying. They doubled it again. He ponied up without a protest. Eventually, they signed a two year contract at something like $2,000 a month. And at the signing, they asked him to explain why he was willing to pay so much for so few hits.

"Easy," he explained. "People who use that search term really want a Garden City financial planner. Out of those 25, maybe 15 click on the link. Ten of them ask me for information. And five of them sign up. Those five have a median net worth of $350,000, and I make about $5-10,000 a year providing them advice and services. They'd be worth it to me at twice the price."

How many of those five would have paid $400 for an unbiased search? As long as it is worth more to the provider to keep you misinformed, than you are willing to pay to be informed, private ratings are subject to corruption. Particularly since if you knew how to tell whether you were getting a good rating, you wouldn't need the rating. That's why equity research ratings from investment banks were useless: no consumers were willing to pay as much as the fees the banks got from security underwriting and merger advisory, and as long as the market kept rising, they couldn't tell the difference between good and bad advice anyway.

Also, private firms may just not care. Does it really matter to them if your blender breaks in two weeks?

Megan clearly places a higher faith in the incentives of government agencies than in private firms or ratings agencies. I'd say government regulations have different problems, not better ones. Usually, government bureaucrats care; no one goes to work in the engineering department of the NHTSA unless they're pretty interested in cars, or the water bureaucracy of the state government of California unless they think water's pretty fascinating.

But government bureaucracies have other problems that private firms suffer less from, or not at all. Many of them get a new boss every two to four years. They are generally punished only for the things they have allowed to happen, but not for the things they have prevented from happening, so they tend to be far too risk averse, even when that reduces welfare. This is compounded by the fact that the people who staff bureaucracies tend to already be risk averse; many chose government over the private sector at least in part because they value the job security. There's nothing wrong with being risk averse--but remember, risk averse does not equal "better outcomes". It just means "less variance". It can, and often does, mean "worse, but more predictible outcomes".

At some level, for example, most of us seem think that the FDA is probably too cautious. I know this because contra Megan, most cancer patients seem to be fairly eager to get into clinical trials of unapproved drugs, especially once they have exhausted the approved treatments. These trials take place after the drug has been tested for toxicity; what they represent is the FDA delaying approval on treatments that could save peoples' lives, not for safety reasons--though nasty side effects sometimes do show up in later clinical stages, this is of fairly small concern to terminal cancer patients--but in order to put its "this works" stamp on it. If this ordering of values were really the majority preference, we'd all go home to die instead of trying remedies, whether they are drug trials or alternative therapies, that haven't gone through double-blind testing. This could happen to private firms, as well, of course, but competition and price signals seem to curb that tendency. Private insurance firms are quicker to approve "off label" uses of drugs than the FDA, which requires an entirely fresh round of clinical trials to do so.

And while private firms are probably in more danger of outright corruption, government ratings agencies often do serve corporate interests rather than the consumer. They do so in a different way: usually, they protect insiders against competition, which raises costs and lowers quality for everyone. It's a phenomenon called regulatory capture, and it's nearly universal, though to varying degrees. It's not malicious--it's just that the bureaucrats spend a huge amount of time with the companies they regulate, and so naturally, they sort of come to see things their way. Also, all entities, government or otherwise, hate change in their industry. Change is uncomfortable, and makes years of carefully acquired expertise useless. Regulators, however, have slightly more latitude than most organizations to prevent change from occurring.

Governments also have the problem of divided attention: none of the people voting on the laws are experts on what they're voting on. Low profile issues often see the laws written by lobbyists; high profile issues often see them written by panicked staffers who don't really know what they're doing. That's why the financial regulations written after crises are often so, well, weird; whatever the appropriate level of financial regulation, it's hard to see what benefit America derived from separating underwriting from banking for 60 years, which was a prominent feature of Glass-Steagall, the 1930's Congress' response to the Great Depression.

Knowing all that, it's hard to come up with a pithy theory of when government agencies will do a better job of providing information or regulation. Tenatively, I'd offer this: government works better when the coordination problems are large, and failures are hard to observe, such as dealing with e. coli outbreaks; private firms work best when the failures are easy to observe and directly affect the purchaser, such as auto safety and blenders. This being about what the public choice people have been saying. Of course, that still doesn't explain why financial regulation is such a mess.