Dani Rodrik is a Liar
He's a First-Best Economist, That's what he is !!!! I know this because I read his book
Crooked Timber is planning on doing a seminar on Dani Rodrik's book "One Economics, Many Recipes". Go Buy The Goddamn Book!
In the meantime I'm gonna jump the gun. Because I'm sort of upset. We do economics because we like precision, objectivity and the ability to disagree with each other while maintaining respect for our fellow economists whatever crazy notions they may hold at one time another (um... we all have exceptions). They at least have thought it through.
Which is why we should make certain terms precise. What exactly is a "second best economist" as opposed to a "first best economist"?
I thought there was a difference there, and maybe there still is, but after going though the book all I can say is that I agree.
But. This is Dani's basic framework;
You should identify the areas where the DIRECT distortion are the biggest and focus on those.
( Added: As Dani notes in comments, I should be clearer (leave it to me to be imprecise in a post where I ask for precision). By "biggest" I mean those with the highest shadow price in terms of welfare loss. Those are not necessarily the same as the ones where there is the largest gap between marginal social benefit and marginal private benefit. The example in the book is one of opening up the capital account )
The not so implicit warning is that if you try to reform some other distortions the theory says that either a) the likely gains are going to be small or b) while you make things better in one area you might screw things up in another. Or a or b. If you at least focus on the biggest distortions, you might, just might, get it right (though actually choosing among the probabilities in this uncertain world (and the uncertainty of the world is what Danny Rodrik stresses)) is exactly what makes this hard.
As far as practical advice goes it's about as good as you can get unless you're the sort who goes for Utopian visions.
It's a humble message which is what economics as a field of study should aim for. But it's not "second best economics". It's still "first best". It just says that you should be careful in your thinking about what constitutes first best.
And that's why I get upset.
A First-Best Economist uses basic economic theory to guide her decision. Yes, even "Econ 101" theory sometimes. But this is same approach behind "Growth Diagnostics".
A Second-Best Economics is a different animal. She's is not the fauna that Dani hopes for. It's not about identifying a menu of options, choosing from that menu, considering how one choice affects another and how the feedback mechanisms work ... arghhhh I quite dislike putting this into prose. George Goddamn Akerloff and Kenneth Freaking Arrow are both First Best Economists! They're the First among the First Best Economists!
A Second Best Economist insists, and keeps on insisting, that no matter what, INTRODUCING MORE DISTORTIONS CAN MAKE THINGS BETTER!!!
AND THE MORE DISTORTIONS THE BETTER!!!! They want to up the lambda's just like the Washington Consensus folks don't even pay attention to the magnitude of lambda!
The politics:
There's a big difference between saying that;
a) reform everything everywhere in accordance with neoclassical economic theory
b) DOWN WITH NEOCLASSICAL ECONOMIC THEORY!!! Do the opposite of what it says. Distortion everywhere. Therefore distort MORE!!!!
c) Heavy Metal!!!!!!
d) Think about it
Obviously any thinking person would choose c) ... wait...
Um. It's a loaded question. Obviously c), wait no, d) is the right answer
Crooked Timber is planning on doing a seminar on Dani Rodrik's book "One Economics, Many Recipes". Go Buy The Goddamn Book!
In the meantime I'm gonna jump the gun. Because I'm sort of upset. We do economics because we like precision, objectivity and the ability to disagree with each other while maintaining respect for our fellow economists whatever crazy notions they may hold at one time another (um... we all have exceptions). They at least have thought it through.
Which is why we should make certain terms precise. What exactly is a "second best economist" as opposed to a "first best economist"?
I thought there was a difference there, and maybe there still is, but after going though the book all I can say is that I agree.
But. This is Dani's basic framework;
You should identify the areas where the DIRECT distortion are the biggest and focus on those.
( Added: As Dani notes in comments, I should be clearer (leave it to me to be imprecise in a post where I ask for precision). By "biggest" I mean those with the highest shadow price in terms of welfare loss. Those are not necessarily the same as the ones where there is the largest gap between marginal social benefit and marginal private benefit. The example in the book is one of opening up the capital account )
The not so implicit warning is that if you try to reform some other distortions the theory says that either a) the likely gains are going to be small or b) while you make things better in one area you might screw things up in another. Or a or b. If you at least focus on the biggest distortions, you might, just might, get it right (though actually choosing among the probabilities in this uncertain world (and the uncertainty of the world is what Danny Rodrik stresses)) is exactly what makes this hard.
As far as practical advice goes it's about as good as you can get unless you're the sort who goes for Utopian visions.
It's a humble message which is what economics as a field of study should aim for. But it's not "second best economics". It's still "first best". It just says that you should be careful in your thinking about what constitutes first best.
And that's why I get upset.
A First-Best Economist uses basic economic theory to guide her decision. Yes, even "Econ 101" theory sometimes. But this is same approach behind "Growth Diagnostics".
A Second-Best Economics is a different animal. She's is not the fauna that Dani hopes for. It's not about identifying a menu of options, choosing from that menu, considering how one choice affects another and how the feedback mechanisms work ... arghhhh I quite dislike putting this into prose. George Goddamn Akerloff and Kenneth Freaking Arrow are both First Best Economists! They're the First among the First Best Economists!
A Second Best Economist insists, and keeps on insisting, that no matter what, INTRODUCING MORE DISTORTIONS CAN MAKE THINGS BETTER!!!
AND THE MORE DISTORTIONS THE BETTER!!!! They want to up the lambda's just like the Washington Consensus folks don't even pay attention to the magnitude of lambda!
The politics:
There's a big difference between saying that;
a) reform everything everywhere in accordance with neoclassical economic theory
b) DOWN WITH NEOCLASSICAL ECONOMIC THEORY!!! Do the opposite of what it says. Distortion everywhere. Therefore distort MORE!!!!
c) Heavy Metal!!!!!!
d) Think about it
Obviously any thinking person would choose c) ... wait...
Um. It's a loaded question. Obviously c), wait no, d) is the right answer
posted by YouNotSneaky! at 12:50 AM
12 Comments:
Dani Rodrik is a first-rate economist because he understood the theory of the second best better than most economists, and the mark of his success and victory is that so many more economists today can no longer fathom anybody not understanding the main points of the theory as obvious.
Younotsneaky raises good points, namely that the theory of the second best is a result that can be demonstrated within a very neo-classical framework, and that any 'first-rate economist' would have understood that all along.
True, but let's not obscure how far things have moved in economics in the past 30 years, and how much we owe that to people like Dani Rodrik who built careers challenging many more narrow-minded economists.
Rodrik's message has been that we need to open our eyes to political economy, information asymmetries and incentives, non-convexities in production, and other issues that greatly complicate analysis and that render policymaking advice a much more delicate art than just saying 'let the market rip, baby.'
You have to transport yourself back nearly two decades to see the dragons that Rodrik had to slay, to understand what he was up against and why "an economist of the second best" is today an animal that we now recognize as an important, and indeed now dominant species.
Look back for instance at the years Rodrik took aim at the so called 'Washington Consensus" package of policy reforms most closely associated with the World Bank (under Anne Krueger as chief economist) and the other two members of the policy cartel, the IMF and US Treasury. Their main mantra was 'privatize, liberalize, and downsize the government' and don't look back. They had many good points to make, they were right about many things, but Rodrik hit on target because he pointed out so clearly that such a broad un-sequenced and un-coordinated packages of reforms failed to "consider.. how one choice affects another and how the feedback mechanisms work .." Along the way Rodrik has also been one of those who have argued that building a strong state is also very frequently a necessary first step if you are to progress in the task of building and extending markets, and making sure that they are competitive.
In short, Rodrik's vision is one that we (trained economists) can all now agree on as essentially right, but let's not take away from the real intellectual battles that people like Rodrik had to fight and win in the 80s and early 90s.
Dani Rodrik fought from the fringes and now he's in the mainstream. But that's as much proof of far the field has moved as much as that 'he was just like us all along.'
In the interest of precision, the Growth Diagnostics framework entails looking not for biggest direct distortion (i.e., the largest "tax"), but for the distortion with the costliest effects. The latter requires taking a stand on how the economy works (or, more precisely, doesn't work), second-best interactions and all. Please re-read pp. 62-63 of the book, where the distinction is made explicitly.
I say, burn all labels. Let's judge policy proposals on their own merits rather than the labels attached to those making them. (Idealistic and infantile of me, I know.)
Okbut,
I don't disagree with any of that. You're completely right. Also, you should have an Okbut Blog.
Dani,
You're also right. "Biggest" means distortions "largest shadow price" not "largest gap between social and marginal benefits".
Gabriel,
Yup, I also agree (some thing's wrong with me today). Part of the point is how much we all really agree on stuff and drawing these distinctions is mostly semantics.
Post, post!
What do you do if you don't have any reliable estimates whatsoever about the shadow prices of distortions in terms of welfare (or even any particular sense of what the wedges between private and social value are, as opposed to a general identification that things aren't working well)? This might be a quite common problem.
Well, yes. And the rest of the book (past chapter 2) is meant to convince that those shadow prices can be reasonably identified and ordered. You might remain unconvinced and that's fine.
And any reform strategy will have roughly the same problem.
But past chapter 2, it's all about the costs and benefits of policy changes in terms of welfare, (gu)estimated directly. The wedges between social and private returns don't show up anywhere else in the book, and indeed a lot of the rest of the book is about things which are really quite hard to fit into these terms.
Getting specific, one pretty common proposal is the creation of a securities law framework. You can force this into the tau-mu model by declaring that "the absence of a stable regulatory framework" to impose an additional cost on investment, declare this a "distortion" and estimate the value of getting rid of it. But this clearly makes the "First Best Economist" triumphalism more questionable - a set of insider trading laws *is* a restriction on behaviour, it *is* a distortion of market activity itself and there's no way of getting round that. A lot of chapters 4, 5 and 6 are about this sort of thing.
Ok, but that just seems to indicate that you believe that chapters >2 don't really convince about the lambdas and slip into talking about taus because that's easier for an economist to do. Which is fine, in fact I sort of sympathize with that view (for other reasons) but that's a different criticism.
As far the specific example, I don't think I understand. I guess one could say that regulating a monopoly or a polluting firm is "distorting things" relative to leaving it alone, even when such a policy lines up the social and private marginal benefits properly. But that's not really the way that economist think about it. Distortions can come about through the market or through the government. But they are "differences between private and social marginal benefits" (wherever these comes from) and NOT "restrictions on market activity" (though of course they could be). Which is at least part of what I was trying to point out over at CT before we got to arguing about the maths. So introducing securities law would be removing a distortion, not creating a new one.
So introducing securities law would be removing a distortion, not creating a new one.
But for that definition of "removing a distortion" (which would presumably also apply to introducing antitrust law under the right conditions, or even to land reform), then your caricature of "second best economists" doesn't go through, does it? Most of the "second best" recommendations can be dressed up as "removing distortions" in this regard, at the cost of quite a big departure from ordinary usage.
But the one which can't is industrial policy. There, fairly and squarely, it does fall into your version of "second best economics". There's a wedge between private and social investment returns, and the industrial policy solution is, usually, to try and offset it with a government-led investment program financed by taxation (ie, another wedge). Dani Rodrik does very definitely take a position on whether industrial policy can be in some cases desirable, and it's very much not in the direction of your "first best economics". He's specifically and definitely, in the right conditions, in favour of government-owned, taxpayer-financed institutions making investments in industrial capital. The more I think about it, the more I think I should have emphasised this, because I don't see how it can be fitted into the Chapter 2 model at all except at the expense of robbing the tau parameters of any empirical content.
I think you're mostly right about industrial policy - which is why I like that part the least and also why I think Dani takes pains to stretch the definition of "industrial policy" beyond the way it's traditionally used - 5yr plans and 5yr plans lite - to include things like securities law or what have you.
And no, I don't think that most traditional "second best policies" - price and wage controls, import licenses, high tariffs, government directing which industry is to produce how much and with how much investment - all of which HAVE been justified on 2nd Best grounds before - can be recast as "removing distortions".
I do think they're different specie of a creature called "adding a new distortion in hopes it will offset another distortion".
Or, as a mock slogan from 1960's Poland had it "Socialism heroically conquers problems that other systems never encounter!"
(maybe it sounds better in Polish)
A very simple comment: in a developed economy, the most important policy consideration may be that adequate checks and balances between competing interests in a situation be or are already established -- the hidden hand can take it from there.
Post a Comment
<< Home