- Arnold Kling, economist, and ex-employee of Freddie Mac and the Fed, makes an argument for Principle Based Regulation™, an argument he's been making for quite a few years.
- Sonic Charmer, mocker-in-chief of financial regulations, and industry insider argues that this constitutes Yglesiocracy, and is generally stupid.
- Foseti, master critic of the permanent bureaucracy and financial regulator, then piles on, further arguing that PBR is a bad idea.
- Now Aretae the epistemologist, with no experience whatsoever in either finance or regulation, attempts to suggest that neither Foseti nor Sonic Charmer is addressing Kling's essential point:
Kling's line is very simple:
- Ideal regulatory systems can be modeled as games.
- Regulators *try* to solve problem by creating regulations.
- The regulated participants *try* to circumvent regulations by following the letter of the law, avoiding the intent.
- In any system where there is enough money involved, there is no contest...the regulated participants necessarily win.
- Letter-of-the-law regulation is therefore *guaranteed* to fail in high $ industries like finance.
- This is a big problem, insofar as we can agree that an unregulated finance industry is undesireable.
Seems all true and obviously so.
The difficult part is that Kling continues:
- Therefore we should try a different form of regulation: PBR
Foseti and Sonic Charmer address point 7. Foseti also pulls out public choice, regulatory capture, and regulatory experience arguments, which call into question point #2. But even so...both Foseti and Sonic Charmer address only point #7...when the important part of Kling's argument is points 1-6.
Having made this minor nudge to the geranlly very well thought positions of Foseti and RWCG, this brings up the general principle.
Aretae's Rule 6: When listening to people....Listen to their identification/analysis of the problem. In roughly all cases, the identification and analysis of the problem is the high-value part of the discussion. The solution almost always sucks. This is because solutions are hard, and if you haven't tried it, you're almost always wrong.
Having made this minor nudge to the geranlly very well thought positions of Foseti and RWCG, this brings up the general principle.
Aretae's Rule 6: When listening to people....Listen to their identification/analysis of the problem. In roughly all cases, the identification and analysis of the problem is the high-value part of the discussion. The solution almost always sucks. This is because solutions are hard, and if you haven't tried it, you're almost always wrong.
Posts
11 comments:
You appear to be missing, or assuming away, the most important rule: Rule #0: The problem is worth 'solving', by government.
If you show me a 'problem' that the government attempts to 'solve' via a 'rule' and fails due to the game-theory you cite, some fraction of the time I'll be able to show you a problem the governmnent shouldn't have attempted to 'regulate' in the first place - rule-based or otherwise.
Essentially you are excluding the 'height-of-handrails' situation that arose in a prior discussion. But HoH situations clearly exist, the only question is how prevalent they are. Clearly I think they are widespread, and clearly I could be wrong and you could disagree - but that, and not Rule #6, would then be the locus of our disagreement.
Best
For points 1-6, as a libertarian, I see the conclusion there as being for the -futility- of such laws. Kling, although he doesn't seem to realize it, is arguing against the regulatory state, particularly given how bad his "solution" actually is.
Also, not surprised at all. (While the timing was impeccable, the post of mine which seems to address this was written in ignorance of this, inspired by an unrelated argument I was having with somebody on whether or not personal discretion should enter into the application of laws, which led to an "Aha!" moment about a lot of the arguments I've been having lately.)
1) Ideal regulatory systems can be modeled as games.
2) Regulators *try* to solve problem by creating regulations.
3) The regulated participants *try* to circumvent regulations by following the letter of the law, avoiding the intent.
4) In any system where there is enough money involved, there is no contest...the regulated participants necessarily win.
5) Letter-of-the-law regulation is therefore *guaranteed* to fail in high $ industries like finance.
6) This is a big problem, insofar as we can agree that an unregulated finance industry is undesireable.
7) Therefore we should try a different form of regulation: PBR
---
Let's start with some issues.
First, the libertarian response that #6 or #1-6 or whatever you want to define is unimportant because you don't believe in regulation and want free banking or something, then full stop. That's a pipe dream.
1) We, like every other first world county, live in a democracy.
2) In a democracy there is a market for banking regulation and central banks.
3) That market will be filled in some way. Either by bad regulation or good regulation, because "the voting public" will demand it, whether its good for them or they understand it or not.
4) As evidence of this I submit that all advance nations have central banks, legal tender laws, and banking regulators. It doesn't matter you opinion of these things usefulness, they are empirically proven to be the political economy equilibrium.
5) If you have to have regulation, better good regulation then bad.
6) The best way to get good regulation is to attack #4 & #5. Well funded, well paid, well respected financial authorities with broad mandates and independence from the political process. We know this works, look at Singapore. It's no contest because we don't even try. We don't try because guys like you don't want to try, and live in some fantasy world where central banks are going to go away.
7) If we are going to do #6, we best give our regulators the independence and authority to do their jobs, therefore principles based regulation is best.
Modeling it as a game is a good start. Specifically, it's akin to an advanced roleplaying game. If you've ever run such a game, you'd know that neither RBR or PBR is likely to work when the participants have a strong inclination to subvert the Rules as Intended of your system. Only a strong metarule (don't be a prick) works, and only by being able to self-select your own players. Since you can't do that in a nation at that fine level of granularity, it follows that the problem is insoluble. All you can do is limit the suckage created by it.
Also, I like your rule.
I have a philosophy regarding problems: If the solution isn't implicit in the problem statement, you haven't elaborated on the problem enough yet. If an elegant solution is implicit in the problem statement, you probably haven't elaborated on the problem enough yet.
You've arrived at the correct problem description when you're no longer sure whether you want to try to solve it or not.
RWCG,
I think your ordering is funky. I do think you ask a very important question, but I think that the order of questions is:
1. Is there a problem?
2. What is the problem?
3. Does the identified problem need solved?
Arnold was attacking #1.5, having already answered #1 in the affirmative. I am very happy with your desire to attack #3...but I still think Kling's point is very strong.
Orphan,
I guess I should mend my response to RWCG to include #4 is the Cost-Benefit sufficient.
Jehu,
In this case, I'm suspicious of RPGs as a good model. In RPGs, it's not DM against player...while in the regulation game, the DM is partly an oppositional participant...and substantially a corruptible ref.
Anon,
I assume you're the normal Anon here. I agree with your analysis that there is a market for regulation.
I am strongly of the opinion that you are over-certain of the efficacy of well-paid regulators. I think the Klingian position "Necessarily win" is either strictly true or nearly so.
Mmmm. Let me try Orphan's approach.
Problem description: "I am not a billionaire"
What comes next?
Aretae
Your new 1-4 are exactly right. But I see them as contained in my Rule #0. (I just didn't split it up; perhaps I didn't want to have to go to negative numbers :)
Best
Post a Comment