Work on my book-in-progress has been slowed by other commitments. Among other things I’m fighting privatisation proposals from a Queensland Labor government that seems to have learned entirely the wrong lessons from the global financial crisis. Here’s a section on the GFC and the failure of the micro-foundations approach to macroeconomics. As always, comments much appreciated
The obvious criterion of success or failure for a macroeconomic theoretical framework is that it should provide the basis for predicting, understanding and responding to macroeconomic crises. If that criterion is applied to the current crisis, the micro-foundations approach to macroeconomics has been a near-total failure.
The failure of the dominant stream in macroeconomics was comprehensive.
First, during the bubble years the dominant approach gave little or no warning of the impending crisis. Neither sophisticated DSGE models nor the more pragmatic but less elegant micro-based models employed by the central banks gave much, if any, warning of the impending crisis.
Second, the dominant approach encouraged a benign view of the developments that gave rise to the crisis such as the growth and globalisation of the financial sector and the associated global imbalances. The boosterism of Alan Greenspan was an egregious example, but it was typical of the majority viewpoint.
Third, even as the crisis developed over the course of 2007 and 2008, its seriousness was persistently underestimated. This was exacerbated by the political context in which supporters of the Republican Administration in the US, a group little concerned with reality at the best of times, sought to deny the existence of a recession in an election year.
Fourth, the near-consensus apparent during the Great Moderation collapsed with the onset of crisis, revealing that the split between Keynesian and New Classical views had never been resolved, but merely papered over.
Fourth, it offered little or no useful guidance on the policy and theoretical issues raised by the crisis. The result that the public policy debate has been driven mostly by economists from outside the micro-foundations school. Advocacy of policies of fiscal stimulus has come, to a large extent,from economists such as Paul Krugman and Brad DeLong whose primary field is not macroeconomics, but who retain a historical understanding informed by Keynesianism. The most effective criticism has come from finance theorists like John Cochrane and Eugene Fama, mostly notable as advocates of the efficient markets hypothesis. Arguments on both sides have been couched in terms familiar to economists of 1970 and earlier, with each side accusing the other of holding views that had been refuted by the 1930s
The roots of this failure may be traced in part to problems with the micro-foundations approach itself. Micro-foundations models take general equilibrium as the starting point – modest variations of the standard classical assumptions suggest that deviations from classical properties are also likely to be modest. Macro models calibrated to the Great Moderation encouraged this assumption, as well as exclusive focus on monetary policy based on Taylor rules, which proved unavailing. The collapse of this position forced economists to shift either to the left (back to older versions of Keynesianism) or to the right (to extreme versions of classicism).
However, the broader intellectual climate of market liberalism, in which thinking about macroeconomic issues was conditioned by the assumptions of the efficient markets hypothesis and the apparent lessons of the Great Moderation. Concerns about market imbalances could not easily be reconciled with the implications of the efficient financial markets hypothesis or with the triumphalism of the Great Moderation.
{ 17 comments }
NomadUK 11.05.09 at 12:15 pm
However, the broader intellectual climate of market liberalism, in which thinking about macroeconomic issues was conditioned by the assumptions of the efficient markets hypothesis and the apparent lessons of the Great Moderation.
I’m not an economist, but I do know that this isn’t a complete sentence.
Luis Enrique 11.05.09 at 12:41 pm
isn’t the problem micro-plus-rational-expectations-equilibrium, rather than micro per se? (I haven’t kept up with reading your stuff on this, so apologies if you have already covered this one) only isn’t the computable agent-based models stuff, advocated by the likes of Leijonhufvud, Howitt, Clower, Geanakopolos etc microfounded too?
Zamfir 11.05.09 at 1:20 pm
Nitpick: you have two Fourths and no Fifth
marcel 11.05.09 at 2:11 pm
The obvious criterion of success or failure for a macroeconomic theoretical framework is that it should provide the basis for predicting, understanding and responding to macroeconomic crises.
Isn’t this instead evidence in favor of the EMH (Efficient Macroeconomy Hypothesis)? The macroeconomy operates so efficiently that there no information would allow anyone to come up with systematically better predictions, understanding or responses to crises?
Phillip Hallam-Baker 11.05.09 at 2:14 pm
Hmm, quoting Krugman and DeLong… Well they are certainly the most visible economists on the blogosphere, and Krugman recently got the ‘Not being George W. Bush award’. But surely some others in the field worth mentioning?
Atrios had been predicting the collapse of the sub-prime market for years. And we now know that the Bush administration was expecting some collapse to occur and had been trying to avoid it.
When it comes down to it, the long term lesson of the fiscal crisis of 2008 should probably be the same as the long term lesson of Katrina: competence matters. I think it is probably a mistake to try to read anything more into the events. Similar sets of circumstances come up every 10 years or so, none has caused the same degree of chaos: competence does too matter.
Economics is not physics. If you think that you have an economic law that is true for all time and for all circumstances then you are a fool and you are wrong and we must make sure that on no account should you be allowed to manage our economic affairs.
All economic models, macro or macro require simplifications of reality to make them work. Often the ‘conclusions’ of the ideological theorists are clearly the consequence of the simplification assumptions, not reality. For example economists have no way to model the impact of technology, they simply assume that there will be precisely enough technological progress to fill in their smooth graphs. What stupidity! But what else can be done? Economists have no more ability to predict the future of technology than anyone else. Yet they build these models of the economy that must assume that the rate of technological progress is constant and then conclude that there is no necessary role, or only a minimal necessary role for government in economic affairs.
This ignores the obvious fact that governments are deeply involved in dealing with development of technology (the Internet is the result of millions of dollars of research money from numerous governments) and the economic impact of technology. What was the paramount concern of governments in the 1970s as the effects of containerized shipping were being felt on manufacture, is assumed away in the economic models.
This is of course very convenient for politicians who would rather get on with the business of government and not bother with any of that nasty practical stuff involving little people. They can sit in the big car and dream about reorganizing the borders of Mesopotamia in more orderly geometric fashion, jet off to conferences with other world leaders, inspect troops, etc.
At the end of the day there is no economic theory that can be perfect. But that is not to say that the root cause of the Bush administration failure was overdependence on economic theory. The real cause of the failure, and thus the sub-prime melt down was a failure of leadership. If you have an administration where the only people who are promoted are flatterers who tell you what you would like to hear, then disaster is inevitable.
Trying to adapt economic theory to deal with a political failure is futile. Stupidity does not fit into smooth curves any more than technological progress does.
If the Bush administration had been 1) competent and 2) ideologically capable of making a pre-emptive intervention in the economic sphere, the sub prime melt down would have been of major concern back when Atrios first started blogging about it and when Paul Krugman made Atrios into a big-blogger by telling them to go read what Atrios was writing on sub-Prime. competence back in 2004 would probably not have averted the crisis completely, but it could have avoided bringing the financial system to almost total meltdown. The fed could have easily popped the sub prime bubble early, yes there was a lot more to the crisis than just the sub-prime market, but popping that bubble would have popped the other real estate bubbles too.
Billikin 11.05.09 at 2:22 pm
“The obvious criterion of success or failure for a macroeconomic theoretical framework is that it should provide the basis for predicting, understanding and responding to macroeconomic crises.”
Don’t forget preventing them.
Contrary to what you hear a lot of these days, it is not necessary to predict a crisis to prevent it. Speed limits do not depend upon predicting any particular accident, but upon recognizing danger. We do not have to identify each accident that they prevent to know that they prevent accidents.
Kevin Donoghue 11.05.09 at 2:39 pm
The criticism of fiscal stimulus coming from Cochrane and Fama didn’t strike me as terribly effective. Strident, maybe. Krugman is a special case I think, since he has been berating the profession for more than a decade for the failure to take note of what was happening in Japan. Did economists with a strong interest in international problems, as a group, ever really take to the Panglossian stuff? I’ve no idea what a course in international macro looks like these days, but I find it hard to see how anybody could get much mileage out of those models when the topic for discussion is a currency crisis.
P O'Neill 11.05.09 at 2:53 pm
You’d get an interesting case study in the last 3 occupants of the position of IMF Chief Economist, although a case study that go somewhat against the idea that mainstream macro wasn’t sending warning signs. First, Raghuram Rajan who had written about the risks of concentration of wealth before going to the Fund and then was bashed at the now infamous 2005 symposium in praise of Alan Greenspan for warning that all was not as benign as it seemed
http://www.kansascityfed.org/publicat/sympos/2005/sym05prg.htm
Then Simon Johnson who didn’t get long enough in the job but since the crisis has been one the main public voices of scepticism about the priority given to large bank rescues over other macro considerations.
And now the current occupant, Olivier Blanchard, as good a Keynesian economist as any and very much pushing the fiscal stimulus debate at the level of policy.
But while there was a decent weight of “conventional” economic opinion sending warning signs, they were up against a broader political-economy tide that things were working just fine. And, to converge with another CT discussion, the bright young things who might previously have been doing some crisis models within the micro foundations framework were instead on Freakonomics-inspired excursions to figure out why the express check-out at the supermarket is always the slowest, or whatever.
John Quiggin 11.05.09 at 8:10 pm
@Enrique #2 I don’t make very clear here, but I will try to do so in redrafting that “microfoundations” is meant to refer to “standard neoclassical microfoundations, perhaps with a tweak or two”.
That said, I’m unconvinced that computable agent-based modelling has delivered on its promises so far. We need different microfoundations, but also a willingness to accept that macroeconomic aggregates may be related in ways that are hard to explain on the basis of rigorous microeconomic models. For example, it seems obvious (to me, at any rate) that the housing bubble was related, via excess liquidity and soft credit, to international imbalances like the US trade deficit, but hard to nail this down in a micro model.
Alex 11.05.09 at 8:54 pm
“The criticism of fiscal stimulus coming from Cochrane and Fama didn’t strike me as terribly effective.”
Ah, but John didn’t say that it was. He said that it was the “most effective criticism”. If someone said “Bob is the heaviest mouse”, that doesn’t imply that “Bob is heavy”.
Bruce Wilder 11.06.09 at 2:32 am
The missing “microfoundation” was rather obvious: financial markets. All that general equilibrium modeling to refine what it means for prices to be sticky did not engage with the functions and effectiveness of financial markets: money and marketable securities and debt and leverage and credit.
The macroeconomists were paying very little attention to the growing amount of debt. For all their alleged concern with “microfoundations”, no one thought to ask what was happening with Angelo Mozilo and Countrywide and the underwriting of mortgages and mortgage-backed securities.
I think, if you ask a macroeconomist (Mark Thoma?), he might say, well, I thought that, if innovation in the financial markets was having any effect, it was to make catastrophe LESS likely, because risk was being spread more widely by securitization.
The general view among macroeconomists, apparently, had narrowed in scope, as they refined their view of the management of monetary policy.
One of my personal obsessions has always been the role of the yield curve as a “leading indicator” of recessions. My personal view is that, if you are not an idiot, you recognize that the yield curve represents a fundamental mechanism for the economy, and so, it is not surprising, that inverting the yield curve, induces a recession, because it curtails the carry trade — the business of borrowing short and lending long, which is the traditional core of bank lending (and also the core of business lending using marketable securities, in our post-junk-bond era, if it comes to that). But, a popular view among economists has been that the yield curve is, somehow, not a reliable leading indicator, or that it’s remarkable performance as a leading indicator has been some kind of fluke or meaningless curiosity, or that financial innovation makes things “different this time” and it won’t mean what it meant last time.
Meanwhile, the financial economists were not paying any attention to financial markets, either. The “efficient market hypothesis” has transmigrated from a convenience of research methodology into a highly respected worldview for financial economists, one that relieved leading practioners from knowing much of any thing about the mechanics, of say, mortgage underwriting practices, or trading in credit default swaps.
As I see it, the EMH serves the ideological need to remain ignorant of practical details. Everyone in financial economics recognizes that EMH leaves the question of the degree to which financial markets are efficient, and the identification of features that make them more or less efficient, unaddressed, but the framing of these performance issues remains largely outside the ambit of leading theorists.
I suppose, one could say that the specialization among economists, combined with ideological blindness, to set up policy for failure. The significance of institutional developments in U.S. and global financial markets — deteriorating mortgage underwriting standards, the elaboration re-packaged mortgage-backed securities, and the development of the CDS markets — attracted the attention of neither macroeconomists nor financial economists.
Oddly, I’m not sure that leading practitioners in either field see recent events as indicting their fields. Macroeconomists are likely to confess that they haven’t modeled such large deviations and movements, but will protest that these events are infrequent, and so they have little data to work with. (Given the power of their favorite econometric techniques, and the frequency of these crises, they will have the definitive answer in a millennium or so.) And, just as oddly, the financial economists are likely to make a remarkably similar protest about the frequency of crisis: a market crash of this magnitude every 70 years of so is consistent with their notion of an efficient market engaged in a random walk. Nothing to see here, move on, folks.
I’ve read Cochrane’s polemics — not just his attack on Krugman, but his earlier piece attacking the Obama stimulus, as well. I don’t know how well they correspond with positions he might take, among professional colleagues — he seems to think there ought to be some sharp distinctions between what one says to the unwashed public (whom he is determined to hector with moral imperatives more than analysis of function), and professional peers, that leave me wondering what he actually thinks. Leaving that aside, the most “effective” criticism leveled against Krugman’s NY Times magazine piece was that Krugman appeared to fall short in his understanding of financial markets. It is the one substantive criticism that seemed to me to hit home.
I know the strong desire to make a point about the pointlessness of pursuing the holy grail of “microfoundations” and an equilibrium of equilibria. But, it does seem to me that looming in the foreground is this odd neglect of financial markets, as a functional institutional nexus for the (macro)economy.
Chris 11.06.09 at 2:20 pm
ISTM that one of the problems with economists revealed by the present crisis is that they seem to be unusually easy to pay to say whatever you want them to say, relative to other academic disciplines. (Or maybe there is just greater motive?)
This makes it easy to confuse a well-funded view with a well-founded view. After all, you can cite the Wall Street Journal and several published economists.
IMO, reality-based economists need to define themselves as a school of economics (cutting across subject areas — the primary crisis in economics is methodological, i.e. placing theoretical tractability above empirical verifiability) and police their boundaries, repeatedly pointing out when theoretical arguments are unsupported by data, and then it will become apparent pretty quickly that the armchair pontificators and their oversimplified models have all the scientific credibility of creationism or homeopathy. Currently, there’s just no reliable way for someone outside the field — and apparently many people inside it as well — to distinguish scientific economics from pseudoscientific economics.
The fact that micro scale theories and macro scale theories are difficult to reconcile (if they are) wouldn’t be any more of a fundamental problem than the fact that gravity and quantum mechanics are difficult to reconcile, if each of the component theories is well-confirmed by data collected at the appropriate scale. It can fairly be booted down the road as something that someone will eventually figure out in the future and win a Nobel for. Unsolved problems are an opportunity, not a crisis.
Shorter me: economics needs more chlorine in the meme pool. Consilience between different subject areas can wait.
Bruce Wilder 11.06.09 at 5:12 pm
I agree about the corruption problem, pointed out by Chris. Even that wouldn’t be a problem for the science, if such a political arms race added to the resources available and resulted in a competition for intellectual quality.
The corruption problem is compounded by the fact that the conservative, vested-interest positions are well-served by dumbing down.
JoB 11.06.09 at 5:20 pm
Methinks the plane is too heavy to allow lift-off. Or the runway too short. Or wings are in need of de-icing. Sorry.
9 – it seems exceedingly simple, there is a factor (let’s call it ‘d’ for distortion) that has a certain dependence on how the public views the macro-econmic climate. Pre-bubble it emphasizes everything to the positive, risk is taken because everything is seen in lights of making good bets. Post-bubble it emphasizes everything to the negative, people buy an excess of insurance thereby crashing anything that is not insurance related. There’s only one way to address the conundrum macroeconomically (enter Keynes) design the collective agencies that reduce public spending for certain ‘d’ and vice versa. On a long run we can educate the people increasingly as to not behaving as g*mblers & insurance agencies in the real economy (which a.o. requires, quite regardless of Keynes, the need for collective means of insurance for any and all goods that are – or whose withdrawal is – life-threatening (i.e. no choice for individuals where individuals have chosen not to want to choose – in my case: electricity & heating, I don’t see the friggin’ choice I have).
(at least, John, that’s how I understood you up to now)
Billikin 11.06.09 at 9:19 pm
Chris: “ISTM that one of the problems with economists revealed by the present crisis is that they seem to be unusually easy to pay to say whatever you want them to say, relative to other academic disciplines. (Or maybe there is just greater motive?)”
Hard sciences aside, I don’t think so. (OTOH, consider the Challenger disaster. Some engineers were quite willing to say what their bosses wanted to hear.)
Bruce Wilder: “I agree about the corruption problem, pointed out by Chris.”
OC, economist can convince themselves that corruption is rational. ;)
Chris 11.06.09 at 10:11 pm
@15: Corruption *is* rational, from a short-sighted individual perspective. Intellectual honesty is a collective action problem. The collective enterprise of a scientific field functions much better if everyone practices intellectual honesty (and can rely on each other’s data not being falsified, nor their conclusions overreaching the data), but for any individual it’s time-consuming and sometimes forces you to circumscribe your statements in a way that makes them less impressive, and the overall deterioration of the field is a function of everyone’s behavior, not just your own.
Maybe this explains why laissez-faire types are the most likely to tarnish the reputation of the field and impede its progress toward greater accuracy with negative externalities of intellectual dishonesty. (And why intellectual honesty is enforced by social norms, which are internalized as a higher cause for which it is proper to consider one’s own self interest of lesser importance.)
Kenny Easwaran 11.10.09 at 2:40 am
“The obvious criterion of success or failure for a macroeconomic theoretical framework is that it should provide the basis for predicting, understanding and responding to macroeconomic crises.â€
Maybe this is the case for a policy-maker, but from an academic perspective this seems wrong to me. We don’t judge other scientific theories by their applicability, and we shouldn’t judge macroeconomic theories solely on this bases either. If there’s a model that does do well with crises, but does badly in explaining ordinary year-to-year variation in the behavior of the economy, it seems to me that it wouldn’t be any good.
I would have thought that the benefit of having a micro-based macro-theory is that it helps give you some idea of the mechanisms by which various variables affect the behavior of other parts of the economy. Having a pure macro-theory with no micro-foundation may give you a theory that’s more accurate, but it might give you less understanding. In academia there should be room for both sorts of theories.
Of course, when actually using the theory for policy purposes we should make sure we know which type of theory we’re getting, and in those cases we should make sure that we’ve checked things both against a model that deals well with ordinary conditions and one that is robust to crises.
Comments on this entry are closed.