Less wrong economic policy
Yesterday I heard an interesting story on the radio about US President Obama’s pick to head the Office of Information and Regulatory Affairs, Cass Sunstein. I recommend checking out the story, but here are a few key excerpts.
Cass Sunstein, President Obama’s pick to head the Office of Information and Regulatory Affairs, is a vocal supporter of [...] economic policy that shapes itself around human psychology. Sunstein is just one of a number of high-level appointees now working in the Obama administration who favors this kind of approach.
[...]
Through their research, Kahneman and Tversky identified dozens of these biases and errors in judgment, which together painted a certain picture of the human animal. Human beings, it turns out, don’t always make good decisions, and frequently the choices they do make aren’t in their best interest.
[...]
“Merely accepting the fact that people do not necessarily make the best decisions for themselves is politically very explosive. The moment that you admit that, you have to start protecting people,” Kahneman says.[...]
The Obama administration believes it needs to shape policy in a way that will keep us all from getting hit by trucks — health care trucks, financial trucks, trucks that come from every direction and affect every aspect of our lives.
At the risk of starting a discussion that will be wrecked by political wrestling, I’m always hopeful when I hear about governments applying what we learn from science to policy. Not to say that this always generates good policies, but it does generate the best policies we have reason to believe will be good (so long as you ignore the issue of actual politices that might get in the way).
What surprises me a lot is how strongly most Americans are opposed to soft paternalism—as if every bad thing should either by banned outright with huge penalties, or allowed and completely unregulated. Soft paternalism of government adjusting incentives to reduce harmful activities is widely accepted in Europe, and seems to work much better than either of the extreme.
One obvious example is opt-out system for organ donations. Who can honestly say it would have worse results than either forced organ donations, or opt-in?
Well, the progressives strongly support soft paternalism. And hard paternalism.
Meanwhile, those americans who are not progressives, of course, chafe at progressive paternalism, which is often geared toward remaking society along progressive aesthetic and cultural lines. Thus, we have progressives arguing for taxes on soda and fast food, but not for 500 calorie biscotti, or 750 calorie Soy Chai mocha lattes from Starbucks.
American opposition is not generally a policy level opposition, but a meta-level opposition to granting these powers generally to the government. There is a common recognition among US conservatives, libertarians, and many liberals that governments do not produce optimal policies, so granting them vast new powers is incredibly risky.
Anyway, I agree America would be better off in some ways with more soft paternalism. But it would predictably be worse in some ways, as well. It isn’t obvious that the positives would outweigh the negatives.
I strongly agree. Many of my friends, nearly all of them “liberals,” exhibit knee-jerk opposition to things like road congestion pricing, and pigou taxes on alcohol, cigarettes, and fossil fuels.
Based on London experience I’m quite strongly opposed to congestion pricing.
First, cost of operating the system are enormous. Of £929.8m in 2001-2007 revenue, TfL got ridiculously low £275m in net income, and after subtracting capital costs it’s barely positive return of £10m. If you add transaction costs on charge payer side, the ratio of revenue to waste is even higher.
This makes congestion charge one of the worst ways of raising money. This billion pounds is pure loss. Taxing alcohol and cigarettes on the other hand is one of the cheapest ways of getting extra revenue.
And second, there’s very little evidence that this billion of pounds had significant impact at traffic in London. For example “A report by TfL in early 2007 indicated that there were 2.27 traffic delays per kilometre in the original charging zone. This compared with a figure of 2.3 before the introduction of the congestion charge.”, so it failed quite spectacularly at long term way to reduce congestion. There are some shifts in traffic patters, but they don’t seem to have terribly much impact.
These results surprise me, as I expected significantly better revenue to waste ratio (at least 4:1, not 1:4), and significant impact on congestion levels. But I feel proven wrong and now I’m strongly against congestion charges.
Thanks for the info, it does dampen my enthusiasm a bit, but I’m still an advocate for carefully considered congestion pricing.
The cost of a well designed system should fall over time because of improvements in technology.
Congestion pricing can already be implemented on highways at very low costs. Even if it were currently impractical for a large area like the one in London, there is still much to be gained by implementing it where it is already inexpensive to do so.
One big improvement would be to charge more (or only charge) during rush hour. In some cases this makes administration and enforcement more expensive but I’m convinced we can drive the costs down.
London already charges only during work days between 7 am and 6 pm.
Technology might improve revenue to waste ratio, but right now it would need to get over order of magnitude cheaper to be practical, and that sounds unlikely. Also—if there are hardly any changes in congestion, it’s still a stupid tax even if it became much cheaper. If you just want to charge traffic, gasoline taxes and car taxes are much easier way to go.
It’s not obvious that toll roads really reduce highway congestion. I know in Poland they introduced them as a backdoor tax to fund building highways, in no way related to congestion issues. The costs include not only operating costs, but all the slowdowns to pay the toll, and extra travel due to invariably reduced number of highway exits. I’m not so sure if revenue to waste ratio is really that good.
“on a weekday, the average car driven into Manhattan south of 60th Street causes a total of 3.26 hours of delays to everybody else. (At weekends, the equivalent number is just over 2 hours.) No one car is likely to suffer excess delays of more than a few seconds, of course, but if you add up all those seconds for the thousands of affected cars and trucks, it comes to a significant amount of time.”
I had to add to this thread because I found this quote in what appears to be a good analysis of Manhattan’s transportation externalities and a proposal that includes congestion pricing.
If this is accurate, does it change your opinion on the need for congestion pricing? How much?
Would congestion pricing significantly reduce amount of traffic? I quite doubt it. People are willing to stay extremely long in the traffic, a few extra dollars of congestion charge on top of that won’t make much of a difference, and cost of operating the scheme would be enormous.
Unless the congestion charge is really extremely high, like $100 a day, that would probably work well enough to reduce traffic, but it’s unlikely they’ll ever do that.
America does have pigouvian taxes on all of these things. At the federal level, and additionally at the state level.
To be fair, many of these sorts of taxes can be regressive, so even if you’re fine with paternalism, you might legitimately have qualms about them. In theory of course, you can alter the overall system of taxes and benefits to redress this, but in practice it seems unlikely to work that way.
Obviously we should discuss the impact of each pigou tax I mentioned: alcohol, cigarettes, congestion pricing, and fossil fuels, needs to be discussed separately and in detail. The distribution of the tax burden would be affected far more by taxing fossil fuels than any of the other categories, but cap-and-trade has similar effects on people’s pocketbooks. As far as I can tell these are the only two reasoned approaches to global warming.
In many cases the poor experience the negative externality most heavily, so its unclear that taxes which appear regressive, actually are.
Those who are against easy organ donation often argue that it would provide incentives for doctors to strive less to save people in accidents or suffering from issues like brain tumors, since on strict utilitarian grounds, that person’s death might save several others. This isn’t something you know is happening except statistically, which means no redress for those (or their heirs) so affected. Lots of people don’t want to feel as though they cannot trust their doctor, even if more lives might be saved overall.
Unless there are technical subtleties in the organ transplantation process I’m not aware of, this sounds completely insane to me.
Whatever accidental cognitive goldbricking doctors are guilty of, they’re most likely to be guilty of it now, when organs are very scarce, making it highly likely that each organ recovered from a goldbricked patient will be given to some other needy person. If organ donation were the norm, the supply would outstrip demand, and recovering organs wouldn’t be a big enough deal to (accidentally) risk your career and your humanity over.
It sounds to me like opponents of organ donation [1] are just voicing squeamish emotions without bothering to make sense.
[1] I think this phrase is actually a complete, isomorphic formulation of the problem. “Who could possibly oppose organ donation?” and so on.
[2] I’ve restricted my commenting to HN for too long. How do I make pretty superscript footnotes?
And people who think so can opt-out obviously. Unless you think organ donations as a whole are a bad idea, opt-out is the right thing to do.
I actually do not necessarily agree with your last sentence. I think that organ donations as a whole are a good idea, but I don’t think takings are a good default for all cases. I’m not sure about this case, partly because of the sticky issue of ownership and transfer, which I won’t bother going into unless someone asks.
It’s good that they’re aware that humans often make bad decisions. But the major take-away point from that article is the quote, “The moment that you admit that, you have to start protecting people.” And that’s a different issue.
I wouldn’t call this an encouraging awareness of bias. I call it an abuse of science to push an ideological agenda.
You seem to be making assumptions by what is meant by “start protecting people”. In particular, I’m guessing that you think this implies taking action that reduces ones freedom or ability to choose, but that’s not necessarily the case.
In the book Nudge, the author gives the example of laying out food in a school cafeteria in such a way that healthier food is more prominently displayed than unhealthy food. It has a huge effect on what people actually eat, and yet all the same food is still available.
The bigger problem is that regulators and others in gov’t are just as biased and make just as bad decisions as anyone else. And the effects are made worse by being imposed on so many others (at least some of whom would have better decisions if left to themselves).
In this case they are applying it to meta-policy, rather than policy: how to get people to go along with their policy.
Here is another excerpt:
I took Cass Sunstein’s Behavioral Law and Economics class a year ago. I’m at the University of Chicago, but the syllabus was almost identical to when he taught it at Yale.
I was quite impressed with him in person, and he publishes more than one man should be capable of publishing.
monetary incentives trump the incentive to tell the truth for most people. therefore it is a bad idea to mix science and politics. you just wind up with bad science being promoted.
see also
“How Obama Is Using the Science of Change” http://www.time.com/time/printout/0,8816,1889153,00.html
This is, of course, only true for some ethical frameworks, and it does little good to handwave that away as he does here. Many people do not hold ethical systems that involve maximizing the utility functions of others.
Taken to its extreme, imagine that someone made all your decisions for you. You would seem to have a higher utility, but you would have no free will. You would be more like a character in a book than a living person.
I think that maybe there’s some way in which the amount of aliveness you have is a function of the amount of free will you have, and that your “super-utility” is utility * aliveness. So a life with less freedom could have higher utility, yet be less valuable.
Is there a reason you can’t just redefine utility to capture the value of freedom?
Many (even most?) people do have freedom as part of their utility function, but at different weights, so redefinition is unnecessary if you grant a moral imperative to increase the utility of others.
The problem is that most people interpret such an imperative as “increase the utility others would have if they shared my own utility function”, which is not at all correct. Simply redefining utility in the general case to include freedom is in this class of mistakes.
Amartya Sen has written extensively about how to do just this, though he wouldn’t call it utility either (it’s one of the cornerstones of the capability approach). He formalizes it in terms of the real option sets available to an individual rather than “free will” though. The main difficulty is how to quantify and value different option sets. (You can’t just look at the size of the sets, because different options are likely to be differentially valuable qua options, and you need to incorporate that somehow.)
You seem to be handwaving the definition of “free will” a bit here. On some level, the laws of physics “make all my decisions”, but this clearly doesn’t bother me. Is it really free will that matters, or the perception of it?
If Omega felt sorry for someone and (based on their own utility function) started making subtle interventions in their life, blocking off the possibility of bad choices and opening doors for good choices, all without them noticing—they’d still be reacting to their environment and would have a higher utility. Is that bad?
My own view is that there’s no particular reason it couldn’t be bad, if the individual concerned happened to value Omega not doing that sort of thing.
Technically, you are correct, but on a personal note, the media quotes one of my heroes, speaking casually, expressing a sentiment that many people have, and you want to bust his balls?
I don’t want to bust his balls. Hell, it’s even a sentiment I mostly agree with, modulo a few details.
I’m quibbling with a statement, not criticizing the man.
“Merely accepting the fact that people do not necessarily make the best decisions for themselves is politically very explosive. The moment that you admit that, you have to start protecting people,”
You mean they haven’t admitted that until now? How were they justifying protecting people before?