David R. Henderson  

Feedback Loops in Government

PRINT
There Is No Great Stagnation: ... Jason Furman at CEA...

Public Choice in a Nutshell

Government officials are happy making and executing plans that affect the lives of millions, but when things go wrong, well ... they're willing to accept the responsibility, but they're not willing to take the blame. What's the difference? People who are to blame lose their jobs. People who are "responsible," do not. The blame, such as it is, winds up deflected on to The System, or something else suitably abstract.

But when you cut the linkage between outcomes and experience, you make learning much more difficult. When you were a toddler learning to walk, you fell down a lot. This was unpleasant: shocking, at least, and often painful. Thus, you learned to fall down a lot less often.

But imagine if falling down didn't hurt. You wouldn't have learned not to fall, or at least, you would have accumulated a lot more bruises along the way.


This is from an excellent (and short) op/ed by Instapundit Glenn Reynolds in USA Today. The title: "Penalties for Politicians."

How did this incentive system come about? Reynolds writes:

The problem is that they don't have, in President Obama's words, "skin in the game." When it comes to actual wrongdoing, they're shielded by doctrines of "absolute immunity" (for the president) and "qualified immunity" (for lesser officials). This means that the president can't be sued for anything he does as president, while lower-ranking officials can't be sued so long as they can show that they were acting in a "good faith" belief that they were following the law.

Such defenses aren't available to the rest of us. And they're not even the product of legislation passed by Congress after considered judgment -- they're judicially created. (Judges gave themselves absolute immunity, too, for good measure.)


Reynolds goes on to advocate some changes in the incentive systems facing politicians. His changes might make sense but I would want to think carefully about the unintended consequences that might arise and I would also want changes that could actually be enforced. For example, his proposal that the U.S. president not be allowed to travel when there's not a budget could cause the president to sign on to almost anything if he likes travel enough. Also, who would enforce this provision? Eric Holder?

I've written previously about one huge bad incentive facing U.S. presidents, an incentive that President Ford helped create and that President Obama added to.


Comments and Sharing


CATEGORIES: Public Choice Theory



COMMENTS (19 to date)
Ken B writes:

I have to say I worry about prosecuting prior administrations turning into criminalizing policy disputes. And the distinction can be tricky; David in the linked second article wants to prosecute "policymakers" (my bold) who condoned alleged torture, and implies they would argue the legality of the policy decision in court. (There is no way they could assert the legality of their actions without discussing it as a policy decision.) I for one see a lot of bad incentives there. Qualified immunity for such things makes some sense, as a restraint on the abuse of power.

This is quite different from prosecuting Eric Holder for perjury. There is no reason for immunities for that kind of thing. And I expect it has gotten worse in recent decades.I doubt you could convict a former attorney general and send him to prison now, but it's been done in the past.

David R. Henderson writes:

@Ken B,
Point taken, but it seems to me that the equilibrium we have reached is on one end of the spectrum. There are tradeoffs. You point them out. But then we should trade off.

Ken B writes:

David, I mostly agree, which was the point of my John Mitchell allusion. Poindexter is the last guy I can think of actually held to account. Even locally in Detroit it is like pulling teeth to prosecute a wildly corrupt former mayor.

David R. Henderson writes:

@Ken B,
Got it.

Aaron Zierman writes:

Good article. I had a discussion about this just this past weekend.

One point I'd like to take further: Not only do the government bureaucracies not take blame when things go wrong, they actually use these problems to grow larger! After all, if they only had more power, more authority and more resources, then the problem would have been avoided. How many times have we heard this?

Government failures should not be a reason for larger government. But how can we change the incentives? That becomes complicated...

Hazel Meade writes:

In the market, a kind of natural selection also takes place. The people who get punished by losing their jobs are effectively weeded out. So even if the DON'T learn anything, the weeding process in itself will select for people who perform well. Politicians don't even need to learn anything from their mistakes, just removing them from their positions will tend to thin the herd of the incompetent.

Mike Rulle writes:

I agree with David versus Glenn.

Further, on the political front, striving for perfection is a loser in the short run. I simply want government at all levels to gradually shrink. This is very difficult to accomplish when a majority of the populace believe the government is primarily a force for good. I think in many ways it is.

But it is also a corruption machine with no "responsibility". Any institution can figure out how to do the same or even more with less. The S&P 500 companies grew earnings largely through "productivity". Sequestration on a constant basis is one good method to gradually shrink. Frankly, I do not think the government accomplishes much. It interferes and wastes much. I will trade permitting them to continue to do so, if they will do it with 25% less money.

David R. Henderson writes:

@Hazel Meade,
In the market, a kind of natural selection also takes place. The people who get punished by losing their jobs are effectively weeded out. So even if the DON'T learn anything, the weeding process in itself will select for people who perform well.
Well put, Hazel. You've just stated the essence of one of the classic economics articles written by the late Armen Alchian: "Uncertainty, Evolution, and Economic Theory."

Tracy W writes:

I think the "lose your job" incentive though is too blunt. Everyone makes mistakes, an entrepreneur or a capitalist typically just loses some money from mistakes. That allows them to keep learning even so. An employee loses their bonus. For really big mistakes you do lose your job, or go bankrupt, but government lacks more I think those more subtle feedbacks.

Tom West writes:

I'm not all that sure about there being a whole lot of difference between the private and public sector.

Short of going bankrupt, being let go after making a big error in either sphere has as much to do with your relationship with your employer as with the nature of the mistake. I don't see that as much different in either domain.

As for politicians, presumably they're not fired on the spot, but they are up for assessment every four years for their entire career in a way that almost no private employee of any worth would countenance. At that point, we the public judge whether their errors are egregious enough to cause them to be removed.

Thus I am not certain whether the incentives to avoid errors are really all that different between public/private sphere.

I'll say that in my personal experience (although only at the bottom), there didn't seem to be all that much difference, and what difference I did see was that the public sector workers were *more* cautious about making errors (at the expense of innovation in my book, but it would seem a tad unfair to tar them for both being too innovative (i.e. making errors) and not innovative enough (i.e. not making errors) in the same post :-)).

Bostonian writes:

"For example, his proposal that the U.S. president not be allowed to travel when there's not a budget could cause the president to sign on to almost anything if he likes travel enough."

Congress has passed a law stating that the President must propose a budget by a certain date. The President's invitation to give the State of the Union Adress before a joint session of Congress should be revoked when he has failed to present a budget. Like many other things, all this proposal requires from the GOP is guts.

David R. Henderson writes:

@Bostonian,
Congress has passed a law stating that the President must propose a budget by a certain date. The President's invitation to give the State of the Union Adress before a joint session of Congress should be revoked when he has failed to present a budget.
Here's the problem: I think that's unconstitutional. See Article 2, Section 3. Now, what they could do is say that they won't attend. That would be quite interesting.

Les Cargill writes:

All organizations face the same problems. Think GM in the '70s ( assuming we can use the term "private enterprise" to describe that ). You can't take seriously the premise that "the incompetent get fired".

The problem with politics is the endless series of narrative fallacies that are coin of the realm in that "market". They may be more than "coin"; perhaps more like "air".

Every Memorial Day, I sit and resist the urge to answer every yellow-ribbon thing said with "maybe we shouldn't send as many people overseas." Do you think the people who say "support the troops" want to hear that? Not in my experience.

I could be wrong, but it sure looks like cognitive dissonance to me. If that is cognitive dissonance, then it's probably due to narrative fallacies, most probably mistaken for "values".

My inner Mencken says this is not only intractable, it's *profoundly* intractable.


Andujar Cedeno writes:

[Comment removed pending confirmation of email address. Email the webmaster@econlib.org to request restoring this comment. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Ken B writes:

I want to slightly change my vote. David is right on the issue of torture. This should be a matter of law not policy. My own view matches John McCain's: torture should be illegal, and if a ticking tiome bomb situation arises you go ahead and plead necessity. That does not match the kind of thing David is discussing, where he is envisioning a more cold blooded thorough-going policy of using torture in less exigent situations. I believe that is illegal under current law(?), which means it's not a legitimate "policy" decision. When I think about the necessity defense I see it is just what David is asking for.

As a further caveat I don't think water-boarding is torture.

Tom West writes:

> As a further caveat I don't think water-boarding is torture.

I think that's pretty much the summary of the problem right there. I'm not certain there are a lot of people who "support torture", but the line between pressure and torture isn't all that obvious.

After all, I absolutely consider water-boarding torture in the classical sense, but given there's obviously a substantial portion of the populace who disagrees with me, it's guess it's not open-and-shut.

Ken B writes:

@Tom West:
I agree this a gray area. But from what I have read, from people who allowed themselves to be water boarded is
1. they felt no pain
2. they suffered no damage
3. they recovered almost instantly

I am curious if you can cite any of the "classical" tortures of which any of that can be said. Wrack, thumb-screws, bastinado, I have tried quite a few examples against these criteria ...

So while it's something, and it's unpleasant, and it suffers from the same "say anything" problems torture has it's not quite the same thing. We need another word.

Tom West writes:

I am curious if you can cite any of the "classical" tortures of which any of that can be said.

You have a point there. On the other hand, I'm fairly certain I'd be willing to go through a low-level beating (which is considered torture, albeit at the lower end of the scale) than get water-boarded.

As I said, there's obviously room for disagreement, which makes criminal prosecution rather iffy.

I believe the word your looking for is "enhanced interrogation". Obviously though, people fighting against water-boarding want it called 'torture'.

Ken B writes:

@Tom West:
A bit off track perhaps but ...
One of the real problems with "enhancements" of any sort is their tendency to elicit whatever the interrogator wants to hear. In the 30s the NKVD found a simple way to get old bolsheviks to confess to any crime at all: sleep deprivation. Bukharin and Zinoviev were not beaten, wracked, or even water-boarded. They were kept awake.
Now the system had a flaw compared to torture: when rested the victims would recant. However the world is full of people who would apologize for anything the Soviets did or said, so the confessions served Stalin well.

In short, I agree there are issues even absent what we are calling "classical" torture.

Comments for this entry have been closed
Return to top