DISQUS

CraigMcGillivary · 4 months ago

A lot of people spend time criticizing the moral behavior of politicians, but I think that this is vastly overstated. They do lots of bad stuff, but its usually because they are deluded and think they are indispensable. Often times the bad things they do are just an expression of the public's will. I actually think that it is political commentators who are the real scum of our system. Some of these people on both sides seem to have no regard for the truth. Why would you become a political commentator other than to try to help people understand politics better?

New Wind Blowing · 4 months ago

I find this an excellent perspective on politicians... how then do we change "who" we elect to high public office? http://bit.ly/48VxMg

heather_33 · 4 months ago

I don't think it is that hard to get into the heads of politicians. No matter what party they do as the people that fund their campaigns tell them to do

Alan Gunn · 4 months ago

I don't see public choice as saying things about what goes on in people's heads. It's more of an evolutionary thing: those who don't behave in certain ways don't get and keep office. If you aren't the sort of person who will do the kinds of things politicians do, you won't be a successful politician (especially in an age when success in politics takes big bucks). This isn't really different from ordinary economics. Nobody thinks businessmen know what the marginal costs of their products are and try to set the price there; it's more a matter of if you don't come up with that price, one way or another, you'll lose the race and be gone. Even birds (looking for food) are "rational" in the economists' sense; nobody thinks they're smart, though (hence "birdbrain" as a term of opprobrium).

Hunter · 4 months ago

Actually, some species (particularly corvids such as ravens and crows) are really smart. They communicate abstract information among their peers (such as that particular human was harassing me the other day, where the person is recognized by his/her face; this is from studies at the U of Washington, if memory serves), construct tools, solve multi-step tool-use problems, etc. (the BBC has been covering a lot of this research and has some pretty incredible videos). So, in fact they appear to be 'rational' in more than just the economist's sense.

LarryM · 4 months ago

I do think, on the other hand, that public choice theory is too harsh when applied to the average bureaucrat. Mind you, this is less than a full throated defense of bureaucrats (insert here every libertarian critique of the state intervention), but explains why sometimes we do get regulation which at least achieves its stated purpose (i.e., regulatory capture is a real problem but not by any means universal).

Rob · 4 months ago

Public choice theory isn't too harsh for bureaucrats - all it implies is that like any other individuals they, on balance, respond to incentives.

Dan Miller · 4 months ago

Think of it this way--I don't think this critique need be limited to politicians. I'd argue that, say, a tobacco CEO who goes on working despite being worth millions is just as alien to the average person. So even in the absence of government regulation, we're going to be led by people with outsized ambition and less conscience than is optimal. I think this goes a long way to explaining why liberals are more fond of technocracy--it's an attempt to avoid the rule of those who really, really want to rule.

Hunter · 4 months ago

Yes, yes, a thousand times yes! Procedural and institutional liberalism (like, you know, the way the US gov't was set up in the first place... checks and balances and such) are exactly attempts to solve that problem. Glibertarians' (not including our host in that number) refusal to see/admit that it even is a problem is one of the major things that so infuriates those of us slightly to their left.

navigator · 4 months ago

To me, the broader implication of the idea, “we’re fated to be ruled by the sort of people who are really desperate to cling to power” is twofold:

(1) The powers of any government at any level should be severely restricted, and never go beyond what is explicit in its constitution. If an additional role or power is needed because of changing times, that constitution should be amended by an orderly public process.

(2) In a representative government, every elected office, along with its staff and top level bureaucratic appointments, should have a two-term limit. We should always be governed by amateurs and never by experts. Political expertise is part of the problem.

As long as any public office is able to easily expand its power and long enough in duration for moneyed interests to get a long-term return on investment from office-holders, we exacerbate the problem inherent in the political mindset.

J R · 4 months ago

will,

at first read, i couldn't help but think that you were being somewhat coy in this post. you live and work in DC. is it really that difficult to look at the crop of fresh-faced twenty-somethings who come here every year ready to change the world and then envisage the process that takes them from naive and idealistic to older and pragmatic, and eventually takes some of them to a place where they are so caught up in this world that they really begin to believe that their own well-being and that of the general public are one in the same?

obviously this is something that would be quite difficult to empirically model, but if i am reading your post correctly your problem is not with the modelling but with the underlying psychology.

TGGP · 4 months ago

It's probably w waste of time trying to get inside people's heads and wondering what "really" motivates them. Like Robin Hanson, I think rational choice economics does a good enough job explaining behavior even if people find it unrealistic with respect to their subjective experiences.

Neel Krishnaswami · 4 months ago

Robin Hanson does not believe rational choice explains behavior, since he's a strong proponent the idea that signaling is pervasive -- and signaling models are incompatible with rational choice.

If you go back to your Schelling, you'll see that signals are credible only when they are costly -- that is, the cost of sending the signal must exceed its benefits for it to be credible. This means that a reputation is something that is valuable to have but imprudent to buy, and so to the extent that you accept costly signaling as an explanation for a behavior you are rejecting the idea that it is rational optimization at work.

Robin is probably right about irrationality being pervasive, though I am more skeptical about signaling being a comprehensive explanation of it.

TGGP · 4 months ago

Costly signals are not incompatible with rational choice. Being costly does not means that costs exceed benefits (on an individual basis). If they did, nobody would engage in them! Robin actually believes that when costs for some forms of "irrationality" like overconfidence exceed benefits, we act more rationally.

Neel Krishnaswami · 3 months ago

Costly signals are not incompatible with rational choice. Being costly does not means that costs exceed benefits (on an individual basis). If they did, nobody would engage in them!

This is the fallacy of assuming the consequent. The rationality of this behavior is precisely what is in question.

Signals are not credible unless their costs exceed their benefits. If the benefits of a behavior exceed its costs, you have an incentive to engage in this behavior regardless, and so your opponent cannot infer anything about your true intentions from it. That's why talk is cheap.

Signaling is precisely a way to escape the Nash equilibrium: if you give credible evidence that you are not rational, then the common knowledge assumption fails, and so the theorem that says you can't do better than a Nash equilibrium now no longer holds. Of course, to send such a signal requires...not being rational about costs!

This is why Schelling was not a particularly mathematical game theorist. His idea of strategy made heavy use of manipulations of perceptions of rationality -- things which mathematical game theorists had to assume a fixed form for (common knowledge of rationality) in order to get viable solution concepts at all. This is also why many of Schelling's suggested strategies are not reflectively stable. For example, he suggests that a winning strategy in the Chicken game (drive two cars straight at each other, with the loser being the one who swerves first) is to remove your steering wheel, so that you have credibly committed to being unable to swerve away. If you do this, and your opponent is rational, then you win the game. However, if both sides do this, you get a crash, which is worse than the Nash equilibrium.

Seriously, read Schelling's The Strategy of Conflict. His prose is elegant and clear.

As an aside, this also gives a handy heuristic for judging policy debates: if someone starts talking about "credibility", that's basically an admission that their policy's costs exceed benefits.

TGGP · 3 months ago

Have you read much evolutionary theory? Animals evolve to give signals. If the costs of those signals exceeded the benefits, the mutations responsible would be weeded out. Then there's the question of why anyone believes the signals. Signals are not noise, they give actual information, generally because the costs DIFFER depending on how accurate the signal is. If I DON'T signal overconfidence, that indicates that I probably have unusually low confidence. The experienced are more overconfident, so it is rational for observers to infer greater experience on the part of the more confident.

Neel Krishnaswami · 3 months ago

Have you read much evolutionary theory? Animals evolve to give signals. If the costs of those signals exceeded the benefits, the mutations responsible would be weeded out.

Yes, I have. Evolutionary models describe a form of bounded rationality, not full rational choice -- for example, evolutionary processes can get stuck in local optima, whereas perfectly rational agents will not. That is, an evolutionary model is an explanation based on a deviation from rational choice.

If I DON'T signal overconfidence, that indicates that I probably have unusually low confidence. The experienced are more overconfident, so it is rational for observers to infer greater experience on the part of the more confident.

This is plausible, but actually establishing that this is a true explanation is very, very difficult. The problem is that this is not a stable solution concept -- you're assuming that the observer is rational, but the counterparty is not acting strategically.

Now, what happens if the counterparty decides to project confidence strategically? And what happens if the observer decides to take into account the possibility that the counterparty is acting strategically, and so on? If you treat the degree of mutual information as a parameter to the model, then the model can predict either that the signal is meaningful or that it's not meaningful, depending on the degree of common knowledge -- if you don't iterate, you get the solution you propose, and if you do, then you get the Nash equilibrium. In between, you get a transition depending on exactly how you've formulated the game.

This means that you need a method for measuring the degree of common knowledge before you can decide whether this is a true explanation or not. Without out, you have to remember that there's a word for models that can be tuned to predict anything, based on a hard-to-observe parameter with many plausible values: non-falsifiable.

TGGP · 3 months ago

That it is possible to get stuck on a local maxima doesn't mean you are, so evolutionary explanations are not inherently dependent on bounded rationality.

Furthermore, bounded rationality can explain why one animal might BELIEVE a signal, but not why the mutation to send out the signal spread. The benefits for signalling must have exceeded the costs.

Hanson's theory is that it is more costly for the inexperienced to signal confidence than the experienced. The experienced learn how to signal credibly. Because they actually are more experienced, it is rational for observers to consider it credible.

Vlad Tarko · 4 months ago

This was an interesting post. However, perhaps the focus on politicians' psychology is misguided, and we should focus instead on what determines their success of staying in power, i.e. on the voters' psychology (the way Bryan Caplan does). I'm thinking that the psychology of successful politicians is probably a consequence of voters' psychology (by means of a natural selection process that favors certain types of politician).

uknowbetter · 4 months ago

But, but those power-mongers often say stuff like 'they care' and that gets Yglesias off.

uknowbetter · 4 months ago

The House should be chosen by lottery every 5 years and each person gets a salary of $400,000 adjusted for inflation. Let the power-mongers aspire to the Presidency, Senate, and local school boards. The purse strings of the country should be in the hands of the people.

Brian Moore · 4 months ago

"So I’m curious what Matt takes to be the broader implications of the idea that “we’re fated to be ruled by the sort of people who are really desperate to cling to power.”"

This may seem libertarian-trite, but it seems to me that the implication would be "let's give them less power where possible." I'm sure the response to his original post was condescending (I certainly was) but when you say "gosh I just noticed that politicians are horrible, immoral power-mongers!" and then attempt to draw no conclusions about how this might impact the many policies that you endorse that increase that power, I suppose you should expect it.

There wasn't even a "yeah, I can see how this power might be abused, but here's why I think policy X was worth it." Many extremely smart people on Yglesias' side seem to pretend that policies will be executed precisely as they imagine them, with no abuse or scope creep. I'm trying not to portray those who disagree with me as stereotypes, but this is making it difficult.

DMonteith · 3 months ago

I'm sure the response to his original post was condescending (I certainly was) but when you say "gosh I just noticed that politicians are horrible, immoral power-mongers!" and then attempt to draw no conclusions about how this might impact the many policies that you endorse that increase that power, I suppose you should expect it.

Yes, but when you say "gosh, politicians are horrible!" and then fail to note that robber-barons and generals and popes, etc., are also horrible then you're left sounding libertarian-trite. I think it's safe to say that institutions that concentrate power are a permanent fixture in human societies. Reducing the power of "horrible politicians" just gives other horrible people in positions of power that much more freedom to operate and provides that much more incentive to power-mad assholes to move into alternative institutions and do their thing there.

Given the high barriers to entry into politics, the various advantages that incumbents enjoy, and the preponderance of assholes seeking entry, elections are far from ideal, but they do present a superior way to separate horrible people from power than exists in many other institutions that concentrate power such as religious organizations, corporations, the military, etc.

Many extremely smart people on Yglesias' side seem to pretend that policies will be executed precisely as they imagine them, with no abuse or scope creep.

Well, many extremely smart people on Wilkinson's side seem to pretend that "smaller government" won't lead to a game of whack-an-immoral-power-monger with a mallet even smaller than the one we currently have. "Eternal vigilance is the price of liberty" is trite, too, but a utopian elimination of concentrations of power isn't one of its prerequisites. I also think you'll find that most of the smart people on Yglesias' side are on board with the eternal vigilance bit.

Margaret · 3 months ago

I recently came across your blog and have been reading along. I thought I would leave my first comment. I don't know what to say except that I have enjoyed reading. Nice blog. I will keep visiting this blog very often.

Margaret

http://lotterymegamillions.net