David Brooks can’t quite reconcile himself to the reality that deploying economists is, was, and always has been a means of providing intellectual cover for the practical politicians in Congress:
Economics achieved coherence as a science by amputating most of human nature. Now economists are starting with those parts of emotional life that they can count and model (the activities that make them economists). But once they’re in this terrain, they’ll surely find that the processes that make up the inner life are not amenable to the methodologies of social science. The moral and social yearnings of fully realized human beings are not reducible to universal laws and cannot be studied like physics.
Once this is accepted, economics would again become a subsection of history and moral philosophy. It will be a powerful language for analyzing certain sorts of activity. Economists will be able to describe how some people acted in some specific contexts. They will be able to draw out some suggestive lessons to keep in mind while thinking about other people and other contexts — just as historians, psychologists and novelists do.
At the end of Act V, economics will be realistic, but it will be an art, not a science.
The great physicist Ernest Rutherford once quipped that all science is either physics or stamp collecting. He was distinguishing between predictive sciences (like chemistry or physics) and descriptive sciences like anthropology. Economics is a science of human behavior and, like all sciences of human behavior, it is fundamentally descriptive rather than predictive.
This isn’t to say that the predictive models of economics are useless; they aren’t. They need to inform the decisions of voters, politicians, and policy makers in the same way that studying psychology and history should inform them or that predictive models of physics inform the work of architects. They characterize the limits on what can be achieved rather than defining it.
I call bullshit on this.
A couple points:
Given my original formal training in engineering I certainly am sympathetic to the notion that all the physical sciences ultimately boil down to “the ultimate science:” physics. However, as any engineer knows, there are all kinds of handy dandy “engineering factors” or “co-efficients” (approximations) that work perfectly well in describing real life situations. Equations bastardized with these approximations work, and are sufficiently predictive, in the real world. (As an aside, a group of physics professors could never run a plant. No chance.)
Milton Friedman once wrote (if I can find the reference I’ll post it) that although the social sciences (including economics) may not have the absolute rigor of physics, they have sufficient robustness to qualify as predictive sciences.
If memory serves, I think about a month ago I pointed out to Dave that his comment that “Newtonian physics don’t work” was true only in the absolutist sense. That the equations of general relativity have terms that become vanishingly small at speeds less than the speed of light – and become the Newtonian equations – seems to me to be the correct practical point of view if you live in the day to day world, and not in the nucleus of an atom. I think of economics in the same way.
As for Steve V’s comment, I’m with Steve. I’m not so sure it was the models that failed per se. They were aware of black swans. It was input data that was suspect.
Separately, I referenced awhile back a PBS Frontline documentary about a regulator who forsaw the issues. She was bludgeoned by various politicians and government types. It was a fascinating documentary. The time was 1997. And two of the bludgeoners were Larry Summers and Timmy Guithner.
Methinks our criticism would be best directed not at faulty economic models, but legislators and policymakers.
“I call bullshit on this.”
“I think our system sucks and it will collapse in the X number of years.” Ok, let’s exclude those types of people. I can predict that we will have another huge market crash in the next hundred years. That is not useful. Who then made useful accurate predictions? In order to do so they must have predicted the international nature of the crisis. Besides the incident Drew mentioned, there was a presentation in honor of Greenspan where Kohn shot down someone making dire predictions (Jackson Hole meeting?), name escapes me, but even that person, IIRC, was concentrated on domestic problems.
Many people called the housing bubble. Who called the international recession? I know Sumner tried to make list and it was very short, all with some qualifications. Drew is partially right about input I believe. The other was the ideological belief that companies, and the people running them, would always act in the best interests of the the company.
Steve
@Drew
” I’m not so sure it was the models that failed per se. … It was input data that was suspect.”
Was the data known to be suspect before the input to the model, or only after (the whole thing came tumbling down)?
Mankiw on the article,
http://gregmankiw.blogspot.com/2010/03/david-brooks-on-state-of-economics.html
It isn’t just that is that many of the models really weren’t models to try and spot financial crises. That Gaussian copula equation everyone loves to point too we never ever intended to be a harbinger of financial crises.
And predicting crises is somewhat self-defeating. Lets say you have a model that accurately predicted a crisis with sufficient warning. What would happen? You’d likely avert the crisis as that information entered into the calculus of many actors in the market.
I’d love to see one of these models that predicts market wide outcomes. I bet they don’t exist.
Yawn….
It isn’t just that some of the input data might have been wrong, but that people thought they had found Truth and failed to appreciate the view that all models are ultimately wrong, but some can be useful. Forgetting that and thinking that a given model is the Truth will lead down the primrose path to Hell.
Slammin’ Sammy –
I’m not sure, but it seems to me it matters not. All I am saying is that it does not seem to me that our recent problem with housing or stock market bubbles is the fault of the economics profession, or economic models.
Just go print out the 100 year Case-Shuler index, and the 100 year Dow to GDP ratios. You don’t need a PhD in economics, or econometrics to see the issue. The root source issue was poor public policy, followed by public mania. (If you want to follow up on this. I’ve posted a dozen times that the housing index took off like a rocket in 1996. Explore policy issues in that time frame if you want understanding. Focus on Bush if you want to sling mud.)
Put in laymans terms, I refused to buy the mega-million home in Naples, FL in recent years because it was intuitively obvious that we were in a bubble. Now, I’m swimming around like a shark. In fact, that’s where I am next week.
Buy low, sell high. You don’t need fancy models for that insight. And you don’t need to blame fancy models for lack of that insight.
“I’d love to see one of these models that predicts market wide outcomes. I bet they don’t exist.”
I think that’s right. Complex models of the inter-relationship of various securities prices and economic variables are not market-wide predictors. If I have an equation of how chemicals react in a chemical refinery under controlled conditions………I don’t necessarily have an equation of what happens if the refinery blows up.
BTW – if it wasn’t clear. “As for Steve V’s comment, I’m with Steve V..”
Drew and Steve, those are good and thought-provoking answers to my question. That Gaussian copula equation was the Li Formula, right? The uses to which that formula was put might not have been the uses for which its author fashioned it (I think he said as much), but the fact that otherwise intelligent people allowed themselves to be seduced by it is worth thinking about. What is it about economics that engenders all the passion and interest outside the discipline. I mean, folks don’t generally get all that excited about the latest theory in anthropology, except in a Gee Whiz kind of way. Economics is different. People fashion policy and bet fortunes (not necessarily their own) on economic theories and models. And, pace Drew, I think economists themselves do bear some responsibility for any misuse because of the apodictic manner in which the results are sometimes presented, the So-and-so Theorem, etc. A laymen might think, Well, if it’s a theorem. …
OT, but there’s too damn many people in Naples, Drew. Head on up to Sarasota (most beautiful city in the US). I’ll be there next month (well, on Anna Maria Island :)).
Correct.
I agree, one should always be careful of how a model is used. For example, could the gaussian copula be valid for a single person with little impact on the market? Maybe. But if everyone uses is it…what then? Maybe it no longer works.
Probably because it impacts our lives so directly? An intersting anthropology thoery is…interesting, but not likely to mean much beyond that. People who have power to implement policy either want:
1. Cover for their preferred world view.
2. Guidance on how best to craft policy.
Or even both which gets really messy.
Again, that is assuming one has found the Truth vs. a relative truth (i.e. if the assumptions hold then this follows logically). Yes, the person presenting the result should underscore that the result hinges on the assumptions inherent in the thoerem/result. But at the same time people should not be so dismissive when some economist says, “Well, all other things being equal….” Start making snide comments along those lines and pretty soon the economist might just say, “Well…oh nevermind,” as he then goes and adjusts his own investments accordingly.
A word in defense of slighted anthropology and in particular its effects on things economic: the notion that the Amazon is a pristine and natural environment is falling to archeologists and anthropologists. They’re discovering that to a surprisingly large degree the supposedly primeval forest is in fact an artifact, a result of millennia of human (Indian) cultivation.
The economic impact? These discoveries weaken resistance to exploitation of the Amazon’s resources. They suggest that the Amazon may be more manageable, even with a degree of farming, logging, etc…
Because someone has to defend anthropologists.
IMO, Math is the fundamental science, not physics.
Michael,
Who’s attacking anthropology? I don’t think anyone said it’s useless.
As an intelligence analyst, I’m quite aware that my profession is more a descriptive art/science than a predictive one, even though a primary function for intelligence is estimating the future. I think the problems I deal with have much in common with the problems economists face and I think I realize the limitations of what’s practicable and possible.
On the topic of models, I agree they are helpful but ultimately they cannot and should not substitute for human judgment. Like any tool, their utility is limited. In addition to models, which provide a false sense of precision, IMO, I would like to see economists adopt an indicator system similar to what intelligence uses to provide warning of high-impact events. Indicator systems work pretty well at this task.
Judgment is the product of experience and experience is the product of age. Allowing anything, whether ideology, models, or what have you, to substitute uncritically for judgment is a conceit fostered by the Baby Boomers as a means of giving themselves a competitive edge over their parents’ generation. I’ve seen the products in sales, marketing, product development, and manufacturing, nearly all bad.
I recall a refresher course I took a few years ago. The final was on practicum. I literally saw the youngsters weeping, unable to cope with real world problems. The oldsters just plodded on. They’d seen most of the problems before at one time or another.
Dave,
You’re right about age and experience to a certain extent, but age and experience also fosters the development of mindsets which make dealing with change more difficult. For example, as a junior intel analyst in the early 1990’s the Soviet/Russian military began sudden, drastic changes in their operations and known doctrine. In one case a certain kind of submarine did went and patrolled in an area no other sub of it’s type had ever been. That patrol was contrary to almost everything we knew about Soviet operations and doctrine. The old, experienced Soviet hands, despite pretty solid evidence the vessel was where it was, refused to believe it for several days until the evidence became incontrovertible. They could not conceive that the Soviets would do such a thing. This wasn’t just limited to few analysts – the senior theater intelligence organization persisted for two days in its analysis arguing that the submarine was where they expected it to be and not where it was. As a junior analyst fresh from my intel training I was incredulous that so many experienced and what I considered brilliant analysts would downplay and discount the evidence in front of them. For me it was clear where the sub was from the beginning. Had there been a war on that sub probably would have sunk and entire carrier battle group.
That early lesson made me value experience to a point, but also taught me of the dangers of developing mindsets which cloud judgment.
Since change is one of the things of which can be most certain, any behavior that discounts its likelihood is likely to prove problematic. My own experience is that rigidity is not solely a product of maturity.
Dave, would you accept the characterization of “predictive, but not deterministic” as a sort of middle-ground compromise?
It’s not that I think that economic modeling is useless, merely that I think it has limited applicability. Specifically, I think it’s incapable of fine-tuning.
Using principles of physics we can send rockets to the moon and planets and construct enormous machines to produce subatomic particles that may not have existed since the universe itself was created. Using principles of chemistry we can produce plastics and other materials in enormous quantity in reasonably efficient plants. The plants don’t randomly start producing marshmallow fluff or high explosives.
I think that today’s economic models are capable of revealing general trends but are inadequate for doing fine-tuning and that’s mostly what people want to do.
Additionally, there’s the problem that, if you pose the same basic question to a chemist or physicist, you’re pretty likely to get the same answer. That’s notoriously untrue in economics.
I think that economics is beyond the “flower—pretty” or throw-the-maiden-into-the-volcano-to-appease-the-volcano-god stage of development. However, it’s still about where, say, physics was 500 years ago, interwoven with politics and religion.
Zenpundit makes an important point:
“Additionally, there’s the problem that, if you pose the same basic question to a chemist or physicist, you’re pretty likely to get the same answer.”
Like global warming.