« A list of children's books which contain the word "scrotum" | Main | Executive pay fact of the day »
Behavioral public choice: the next subfield in economics
Jane Galt writes:
The post below also applies to behavioural economics, which the left seems to believe is a magical proof of the benevolence of government intervention, because after all, people are stupid, so they need the government to protect them from themselves. My take is a little subtler than that:
1) People are often stupid
2) Bureaucrats are the same stupid people, with bad incentives.
There are few subfields in economics that have not been fleshed out with every possible combination of mechanisms, but this is one of them. Yes, it is hard to come up with generalizable results about the psychological and behavioral biases of bureaucrats, but that hasn't stopped many other areas in economics (like, um...behavioral economics) from taking off. In fifteen years someone will write a JEL survey on Behavioral Public Choice, and you will regret not having written at least one of the early papers in the field.
I might add that Behavioral Public Choice gives us a better sense of when government programs actually work. When morale is high, many people in government will "feel they matter," even if they do not, and do a very good job. So Behavioral Public Choice is not just government-bashing, although there is a place for that too. Behavioral factors also help account for why corruption becomes the norm in some settings but not others; psychological propensities are one way to narrow the set of possible equilibria.
Posted by Tyler Cowen on February 20, 2007 at 06:42 AM in Political Science | Permalink
Comments
"1) People are often stupid
2) Bureaucrats are the same stupid people, with bad incentives."
This is one of the main reasons I'm a libertarian.
Posted by: Jacqueline at Feb 20, 2007 7:46:12 AM
Step 1: Watch the Wire
Posted by: Keith at Feb 20, 2007 8:11:45 AM
"I might add that Behavioral Public Choice gives us a better sense of when government programs actually work well."
When exactly is it that government programs do work well? Do we have any examples of when a government programs has been efficient and effective?
Posted by: anne at Feb 20, 2007 8:14:32 AM
Libertarians are also stupid people with bad incentives.
Posted by: Barbar at Feb 20, 2007 8:24:27 AM
Barbar: Good thing we libertarians are generally of the mind to keep ourselves and our government out of peoples' business. I'd hate to think of what might happen if our stupidity spilled over into governance. Personally, I try to confine my stupidity to home improvement and and the stock market. :)
Posted by: billb at Feb 20, 2007 8:32:04 AM
1) People are often stupid
Um, maybe it should be "People don't conform to theoretical predictions. Theorists are stupid". And I say this as someone who likes theorists.
Part of the reason behavioral economics arose was to explain non-Nash behavior in experiments like the Prisoner's Dilemma, public goods games (really just n-person Prisoner's Dilemmas as they are currently formulated) etc. There is a problem with calling these people "stupid" because they don't follow Nash predictions - the problem is that they make more money than they would if they had played using strictly Nash behavior. Who's the more foolish - the fool who makes more money by using non-Nash behavior, or the fool who insists the individual who makes more money using non-Nash behavior is a fool because that person did not use Nash behavior?
Yes, there are other types of experiments (Ultimatum game, Dictator game) where the Nash predictions are not Pareto dominated by some non-Nash outcome, but we have some idea as to how to remove the confounding effects to make predictions more Nash-like (there are papers by Hoffman, McCabe, and Smith and probably a bunch of others by now that show this).
Unfortunately people don't interpret the results this way - they put theorists on a pedestal and assume that because people do not conform to theory "people are often stupid".
Apologies to John Nash - it's not his fault people have taken his theories too literally. There are probably some games in which Nash himself would use "non-Nash behavior" (there's a quote in the Handbook of Experimental Economics, pg. 10, that suggests this).
Posted by: AZ at Feb 20, 2007 8:46:48 AM
I am interested in a behavioural macroeconomics and a new psycho-economics which is different from neuroeconomics and based on social psychology.
Posted by: GVV at Feb 20, 2007 9:18:49 AM
"Behavioral public choice" would be an example of economists reinventing the wheel. There is a huge literature on the biases that creep into bureaucracies (and, for that matter, businesses) but it is in fields called sociology and political science. It is, of course, entirely possible that economists will do a better job at this than sociologists and political scientists but it would be nice not to ignore that literature (as with the sudden discovery in the 1990s that "institutions matter" for economic growth).
Btw, Philip Tetlock's Expert Political Judgment is a very good place to start and look at what you can do with systematic biases in the judgments of elites (although it is not directly on bureaucracies). To see an early attempt to integrate findings from cognitive psychology into the way foreign policy bureaucracies make decisions, try Robert Jervis' Perception and Misperception in International Politics (from the 1970s).
Posted by: Z at Feb 20, 2007 9:58:44 AM
The Besley-Ghatak work on intrinsic motivation in bureaucracies also seems related
Posted by: hbi at Feb 20, 2007 10:16:52 AM
"The post below also applies to behavioural economics..."
So I guess if you write for the Economist you start spelling things English-style?
Posted by: Zachary Skaggs at Feb 20, 2007 11:25:02 AM
just a spare thought on
"Bureaucrats are the same stupid people, with bad incentives."
as it was discussed earlier ( Clark related topics ) it seems that 'quality' of people affects the economic outcomes. And also this 'quality' could be measured ex by IQ.
Now there is a meritocratic approach for governance of country, we could find the one in Singapore. So there might be a way to select 'proper' bureaucrats.
it really quite uncommon way to treat the situation. Still a possibility to look for better bureaucrats changes the weight of argument and also requires to show that the selection of bureaucrats is unrealistic so that the argument was valid.
Posted by: Sergey Kurdakov at Feb 20, 2007 12:04:59 PM
"When morale is high, many people in government will 'feel they matter,' even if they do not, and do a very good job."
Good point. This probably applies to any job with no direct market feedback mechanism as to performance; e.g. teachers, priests, college professors, etc.
Posted by: cllam at Feb 20, 2007 12:24:09 PM
Tyler,
Prescient post. For those not sure what is up here, I think the issue is
doing experiments with people playing roles of bureaucrats or other
government agents (not all of them are bureaucrats). There has not
been much work of this sort, but I am beginning to see some papers
moving in that direction. Some of the public goods game work is
already partly there. Another area that is already active involves
experimental work on corruption, which often takes the form of that
old fave of pub choice, rent seeking.
As some of the papers I have seen are under editorial review, I am
not at liberty to cite them.
anne,
You probably do not approve of what it is doing, but the Social Security
Administration has extremely low administrative costs, less then 1/2% of
the monies it hands out. You may not think it should be handing out the
money it does, but it does so very efficiently.
Likewise Medicare, which is arguably more efficient in some ways (even if
it delayed on covering prescription drugs) than private insurance because
it does not spend huge amounts of money and labor time trying to show that
patients do not deserve to get covered for that treatment they had, and
so forth, not to mention all the pooling of risk issues.
Posted by: Barkley Rosser at Feb 20, 2007 1:42:29 PM
Right, Medicare may have lower admin costs, and it may have other efficiencies. But I wonder if Medicare subsidizes less effective care precisely in order to save on the admin costs that would be required to see which care if effective. As the say in the military, "you get what you inspect, not what you expect." I think it's a major trait of government to be really efficient in the measured areas and be really inefficient in unmeasured areas.
It really all comes down to this: If we can get the admin costs savings and risk pooling advantages of a government-run system, that's great. But not if it costs too much in innovation.
Posted by: Keith at Feb 20, 2007 2:05:30 PM
Behavioral theories are not particularly good at explaing the behavior of trained technocrats.
Posted by: Michael Greinecker at Feb 20, 2007 3:17:28 PM
Keith,
Well, there is that rather substantial and much debated bottom line
regarding medical care where the US pays far more per person than any
other country in the world, a good 60% more than such competitors as
Norway, Switzerland, and Luxembourg, but we are something like 17th
in life expectancy and 32nd in infant mortality. Of course there are
lots of other factors in all that, but offhand the argument about
the supposedly unmeasured quality of medical care paid for by private
medical insurance is something less than overwhelming.
Michael G.
Do you have any evidence that behavioral theories "are not particularly
good at explaining the behavior of trained technocrats," or even a
vague theoretical argument beyond mere assertion?
Posted by: Barkley Rosser at Feb 20, 2007 5:19:32 PM
I'm not convinced the field of behavioral public choice is as fertile as it might initially appear. Most of the examples suggested in Jane Galt's comments thread are just standard invocations of public choice theory wherein rational bureaucrats pursue goals incompatible with the public good.
A true theory of behavioral public choice would involve situations where administrators systematically make decisions inconsistent with rational maximisation of their own utility.
However, most of the classes of decision making bias identified by behavioral economics are quite specific to choices affecting oneself, rather than choices affecting others. A bureaucratic decision making process will do a pretty god job of removing loss aversion and time horizon effects from the policies it suggests, because the policy makers are asking a fundamentally different question about, say, risk, saving, or health than individual consumers of investments or cholesterol.
I can see arguments about escalating commitment holding in relation to policy design, but beyond that I'd be surprised to see the field advance much beyond its vapid "bureaucrats are stupid too" starting point any time soon.
Posted by: Paul at Feb 20, 2007 6:46:07 PM
Economists have done some work on behavioral public-choice theory; see my "Interpreting Federal Reserve Behavior" in Journal of Behavioral Economics, vol 17, Winter, 1988, and the literature cited there.
Thomas Mayer
Posted by: Thomas Mayer at Feb 21, 2007 3:51:11 AM
The problem with stupid, glib theories like Jane's is that they are incapable of explaining why we're not worse off than we are.
Oh, and of course they also would apply to corporate bureaucracies as well.
Posted by: Mike Huben at Feb 22, 2007 8:40:41 PM
大家好，我是臺灣人，從臺灣一個人搬家來到美國，環境很陌生，感覺很孤單。以前在臺灣幾家知名的徵信社工作過，我是一個優秀的徵信工作者，希望早點找到適合自己的工作。希望通過貴站，認識更多的朋友。
Posted by: 謝文豪 at Apr 1, 2008 9:55:26 PM
Posted by: at Oct 22, 2008 10:11:26 PM
http://www.batteryfast.co.uk/toshiba/satellite-a105.htm toshiba satellite a105 battery,
Posted by: herefast123 at Oct 27, 2008 6:12:37 AM
New BATTERY For TOSHIBA 5000 5100 PA3123U PA3178U laptop battery
Posted by: herefast123 at Oct 28, 2008 5:23:00 AM
Posted by: at Oct 29, 2008 3:03:46 AM
Battery fits COMPAQ Presario 1200 1600 1800 116314-001 laptop battery
Posted by: herefast123 at Oct 30, 2008 11:47:00 PM
