Economics and climatology: Why perfect markets are like a dishpan

Brian Ferguson at A Canadian Econoview makes an interesting point:

Ever wonder why so many economists are sceptical about man-made global warming? It's because we've had a lot of humbling experience with just how quickly large scale computer models can go very badly wrong. Remember when we had inflation and unemployment under control through Keynesian fine-tuning?

There are some interesting parallels between the atmospheric sciences (climatology and meteorology) and economics: they study highly complex systems, the available data are non-experimental, and their results have significant policy implications. Not surprisingly, both rely heavily on models as tools for understanding what is going on and why. Of course, models need not be based on a computer simulation, or even mathematics. As Paul Krugman notes, you can put one together using some ordinary household appliances:

Dave Fultz was a meteorological theorist at the University of Chicago, who asked the following question: what factors are essential to generating the complexity of actual weather? Is it a process that depends on the full complexity of the world -- the interaction of ocean currents and the atmosphere, the locations of mountain ranges, the alternation of the seasons, and so on -- or does the basic pattern of weather, for all its complexity, have simple roots?

He was able to show the essential simplicity of the weather's causes with a "model" that consisted of a dish-pan filled with water, placed on a slowly rotating turntable, with an electric heating element bent around the outside of the pan. Aluminum flakes were suspended in the water, so that a camera perched overhead and rotating with the pan could take pictures of the pattern of flow.

The setup was designed to reproduce two features of the global weather pattern: the temperature differential between the poles and the equator, and the Coriolis force that results from the Earth's spin. Everything else -- all the rich detail of the actual planet -- was suppressed. And yet the dish-pan exhibited an unmistakable resemblance to actual weather patterns: a steady flow near the rim evidently corresponding to the trade winds, constantly shifting eddies reminiscent of temperate-zone storm systems, even a rapidly moving ribbon of water that looked like the recently discovered jet stream.

What did one learn from the dish-pan? It was not telling an entirely true story: the Earth is not flat, air is not water, the real world has oceans and mountain ranges and for that matter two hemispheres. The unrealism of Fultz's model world was dictated by what he was able to or could be bothered to build -- in effect, by the limitations of his modeling technique. Nonetheless, the model did convey a powerful insight into why the weather system behaves the way it does.

The important point is that any kind of model of a complex system -- a physical model, a computer simulation, or a pencil-and-paper mathematical representation -- amounts to pretty much the same kind of procedure. You make a set of clearly untrue simplifications to get the system down to something you can handle; those simplifications are dictated partly by guesses about what is important, partly by the modeling techniques available. And the end result, if the model is a good one, is an improved insight into why the vastly more complex real system behaves the way it does.

Popular usage notwithstanding, climatologists have not 'proven' a link between human production of greenhouse gases and global warming. Rather, according to the models that - in their best judgment - best reproduce the main features of the available climate data, the most likely explanation for the increases in the average temperatures we observe is the warming effect of greenhouse gases. That doesn't mean there isn't a more likely explanation - just that no-one has yet been able to think of one.

Although the idea of making important - and possibly very costly - policy decisions based on a modeling exercise doesn't seem to bother many people when atmospheric scientists do it, economists have a harder time getting away with it:

When it comes to physical science, few people have problems with this idea. When we turn to social science, however, the whole issue of modeling begins to raise people's hackles. Suddenly the idea of representing the relevant system through a set of simplifications that are dictated at least in part by the available techniques becomes highly objectionable. Everyone accepts that it was reasonable for Fultz to represent the Earth, at least for a first pass, with a flat dish, because that was what was practical. But what do you think about the decision of most economists between 1820 and 1970 to represent the economy as a set of perfectly competitive markets, because a model of perfect competition was what they knew how to build? It's essentially the same thing, but it raises howls of indignation.

Why is our attitude so different when we come to social science? There are some discreditable reasons: like Victorians offended by the suggestion that they were descended from apes, some humanists imagine that their dignity is threatened when human society is represented as the moral equivalent of a dish on a turntable. Also, the most vociferous critics of economic models are often politically motivated. They have very strong ideas about what they want to believe; their convictions are essentially driven by values rather than analysis, but when an analysis threatens those beliefs they prefer to attack its assumptions rather than examine the basis for their own beliefs...

The problem is that there is no alternative to models. We all think in simplified models, all the time. The sophisticated thing to do is not to pretend to stop, but to be self-conscious -- to be aware that your models are maps rather than reality.

There are many intelligent writers on economics who are able to convince themselves -- and sometimes large numbers of other people as well -- that they have found a way to transcend the narrowing effect of model-building. Invariably they are fooling themselves. If you look at the writing of anyone who claims to be able to write about social issues without stooping to restrictive modeling, you will find that his insights are based essentially on the use of metaphor. And metaphor is, of course, a kind of heuristic modeling technique.

In fact, we are all builders and purveyors of unrealistic simplifications. Some of us are self-aware: we use our models as metaphors. Others, including people who are indisputably brilliant and seemingly sophisticated, are sleepwalkers: they unconsciously use metaphors as models.

This doesn't mean that conclusions based on models are always right; quite the opposite. Models are approximations, and approximations are 'wrong' pretty much by definition. The problem is that it's hard to know if the model is wrong in a way that significantly affects its usefulness for policy analysis. Indeed, many advances in economic thought have occurred because a model-based policy went wrong in an unexpected way. A good economist will always make a clear distinction between a model and reality.

Since I personally know nothing about climatology, I'm willing to defer to the judgment of those who do. But there's always the risk that climatologists - being human - have gotten it wrong. Greenhouse gases may not turn out to be that big of a deal. On the other hand, they may be even more of a problem than even the worst of the model-based scenarios would predict.

When it comes to climate change policy, the most important question may be: Do we feel lucky?

Comments

[Dave From Battlefield]

Something very fundamental is missing from both sides of this question:

Knowing what we know about the human health impacts of pollution and emissions, how can it be right to dump this stuff into the air (and let's not forget the water).

And expand on your point, that I am reading as "we don't know what we don't know," then it should follow that we should be as clean as possible.

Trying to reduce this to just scientific or economic arguments is based on an incorrect premise that it might be okay to continue as is.

Dave From Battlefield.

[MikeM]

Climate science including predictions about increasing global temperature does depend on models but also depends increasingly on very large volumes of measurable data. Atmospheric CO2 concentration has been estimated for the past 800,000 years from air trapped in progressively deposited Antarctic snow, according to results released last September after analysis of a 3.2 Km drilling core. These results are consistent with other ice cores extracted in the Arctic.

Concentration of CO2 is now higher than at any point in that time. Denialists don't have a story as to how this change must have zero effect on climate.

It is true that climatologists have not 'proven' a link between rising gas concentration and rising temperature, any more than physicists and astronomers have 'proven' that Einstein's General Relativity Theory is the ultimate explanation for gravity. Science doesn't do such proofs.

However climatologists have been very energetic in devising and testing theories about causes of global warming that are either complementary to, or might supersede, an explanation based on increasing CO2 concentration. Last year it was established that the vigour of the Sun's sunspot activity has an effect on global temperature, although insufficient to explain recent changes. There is current speculation that cosmic ray concentrations have an, as yet obscure, effect on climate.

But there are no outstanding hypotheses that have any likelihood of displacing the major role of CO2 concentration. Nor are there any serious doubts about the evidence that the planet is becoming rapidly warmer.

[Andy Imboden]

First, a plea: please don't confuse climatology with weather prediction. Models are used in both fields; both fields study the atmosphere, but a distinction should be made.

Fultz's experiment that Krugman describes was used to simulate a slice of the atmosphere, how eddies moved, and other dynamic features of the atmosphere. That kind of stuff does not matter in climatology.

From the climatologist's perspective, you do not need models to tell you if the earth's atmosphere is warming. You do not need models to tell you that CO2 concentrations have been increasing since the industrial revolution. You can measure that kind of stuff. Where models are necessary are to answer the question is this a big deal or not?

Economists that are skeptical about global warming are ignoring evidence. Economists that are skeptical about how much we should spend on mitigating global warming have their heads in the right place.

[ajay]

I think a more relevant question might be: how many climatologists are sceptical about anthropogenic climate change? After all, they use models just as much as economists do; and, more to the point, they use climate models a lot more than economists do, and they know an awful lot more about the climate. The answer is: virtually none.

Just because economic models are bad at predicting the future of the economy, we should not assume that all models are bad at predicting all futures.

Another relevant question might be: why, when they are ready to accept the overwhelming scientific consensus on most other issues, do economists - and others - feel they know better on the issue of climate change?

[Tom Graff]

Its unfair to use anecdotal evidence of models (or empiricals) being incorrect and extrapolate that to say that the use of modeling generally is suspect. Regardless of the field of which we are speaking.

If we are to approach problems scientifically, we must accept that there are few absolutes in science. We cannot allow our lack of certainty, however, to paralyze us. Were there is uncertainty, we are relegated to finding the best evidence we can, to be as sure as is reasonable. But once we come to our best conclusion, we must be willing to act where appropriate.

[Valuethinker]

The analogy is entirely false.

Climate models are based on real physical principles, which can be tested in the lab. The underlying science of global warming is simply this: CO2 and water vapour have a known ability to block infra red radiation.

It would be very surprising if the CO2 increase we observe in the atmosphere did *not* lead to an increase in temperature.

We know how much solar energy hits the planet in a year. We know how much heat energy is radiated back out. The difference is being absorbed by the earth, and the oceans-- and we have several thousand thermometers out there to measure that.

Economics isn't like that. You can increase the money supply, it may or may not cause inflation. There isn't a physical, lab demonstrable theory, which would tell you either way. There's just equations.

To make analogies between the two, is to set equivalent economics and the laws of physics and chemistry, which is of course nonsense.

As to other causes of why the planet might be heating up that we haven't accounted for, the IPCC and other scientific bodies have gone to mind numbing detail trying to work that out. For example, we are currently spending $10m to rule out (or in) the effects of cosmic rays, although we have no real scientific basis to assume cosmic rays affect the climate.

We've also spent a lot of time ruling in/out other plausible factors. One lead candidate is solar flux, but for the period we have accurate observations (less than a century) there has been no meaningful flux in solar energy reaching the earth, yet the planet has been on a marked heating trend. We do think it is a long run factor, though.

Other factors, such as SO2 pollution, black carbon particles in the atmosphere, have proven to be more important. But again, you can show chemically that their effects are limited in time span, whereas you can show with radiochemistry that CO2 lasts in the atmosphere for up to 100 years.

The question is to what extent the planet will heat as the result of adding CO2 and other greenhouse gases into the atmosphere.

There, the uncertainty is as much on the upside as the downside, because of biological feedback effects (plants lose their ability to sequester CO2 as temperatures rise, and permafrost melting releases CH4).

The typical economist (see Lomborg) does things like take the median temperature increase forecast, work out possible costs, and then conclude the costs of global warming are much overstated, or at least not worth the abatement costs.

Which is a distortion of the science. What the science says is, we don't fully know, but it is very unlikely to be less than 1.4 degrees centigrade (on a Business As Usual case for increasing CO2 concentrations), but could be as much as 5 centigrade or higher.

It's not reasonable to look at just the median case, when the damage at the high case (5 degrees or higher) is so large, and so irreversible.

[Bruce Wilder]

It is possible, and may even be useful, to distinguish between analytical modeling and synthetic modeling. The first is a logical exercise, with the objective of finding a minimum set of necessary and sufficient factors for a particular logical result, which can be termed valid. The second is an attempt to determine what is happening with the case of a particular, imperfectly observed set of phenomena, with the objective of making accurate statements and predictions, which are "true".

Krugman fails to make this distinction, and it is, perhaps, telling, because economics, in general (as an academic profession) has had a tendency to fail in making this distinction.

A "theory" in the scientific sense is not a map of the world. At best, a good theory, by identifying the "necessary and sufficient" helps to identify what should be mapped -- what the correct variables and parameters are: not their values, but their unobservable relationships.

Imagining that the analytical model of perfect competition was (to use the word from the elementary economics Lipsey Steiner text I used in the 1970's) "robust" as a description of the world was a colossal intellectual error, an act of breathtaking arrogance and ignorance.

At best, the perfect competition model provides a good framework for defining and understanding some essential concepts, which is what a theory should do. But, failing to understand the need to do rigorous synthetic study, before making generalizations about the way the real economy worked, was an error. What was wrong with economics between 1820 and 1970 was the failure to develop a method of synthetic modeling and testing, whether it be case study, or experimentation or some other method of examining and testing the way the world works.

Particularly egregious is this passage from Krugman: "the most vociferous critics of economic models are often politically motivated. They have very strong ideas about what they want to believe; their convictions are essentially driven by values rather than analysis, but when an analysis threatens those beliefs they prefer to attack its assumptions rather than examine the basis for their own beliefs...". I see little in this, which can be defended. Does Krugman think the proponents of economic models are not politically motivated? He needs to attend AEA meetings while awake and sober. Attacking the assumptions behind a logical analysis would seem to be a logical thing to do; would he really prefer that critics engage in emotional navel-gazing? Samuelson's Principles text looks like nothing so much as a variation on Euclid's Geometry; attacking assumptions is exactly the appropriate thing.

"The problem is that it's hard to know if the model is wrong in a way that significantly affects its usefulness for policy analysis. Indeed, many advances in economic thought have occurred because a model-based policy went wrong in an unexpected way. A good economist will always make a clear distinction between a model and reality."

There are so many things wrong in those three sentences, that it is hard to know whether it is even worthwhile to criticize. An analytical or conceptual model is not "an approximation" at all; approximation has no meaning in such a context. "Wrong" and "useful" as equivalents would make a pragmatist squirm; praise for a policy model proved useful, because it was wrong, would make any sane person run screaming from the room. A good scientist's job is not to make a clear distinction between model and reality, or, as in Krugman's view, to fashion his self-consciousness into sophistication; the idea behind science is to use models, to figure out what is happening in the world, to figure out what the relationships are among observable variables, with the hope of being able to understand and (gasp!) control, to some extent, the working of the world. The economist's job is develop critical methods for distinguishing the extent to which a synthetic model represents the working of a real system or process, and to develop analytical models, that identify necessary and sufficient conceptual relationships.

I will give a simple example of an analytical failure in economics: the economic theory of production, which posits that output is a function of factor inputs. Oh, wait, output is not a mathematical function of inputs -- that would be logical error! Let's reformulate: maximum output is a function of inputs -- now, we are merely incoherant, since "maximum" can hardly be defined, without a reference to a scheme of control, but, of course, control is not an input, and would entail a concept of error, and there is no theory of error. And, what about energy as a distinct input? That would work, except energy consumption implies pollution as a necessary co-product, and besides, in our conceptual scheme of ultimate factor inputs, energy should be traceable back to "land" as an extraction . . . A critical examination, all at the level of pure analysis and theory, exposes this central element of economic theory as conceptual crapola. No empirical work required; economists, observing more than a century of industrial revolution, who could come up with a theory of production, which took no conceptual account of management, engineering or fossil-fuel consumption, but instead joyfully assumed the first two away and ignored the Second Law of Thermodynamics to avoid the third -- I'm sorry, but some intellectual disability, which went way beyond the limitations of available modeling techniques was at work.

I am not going to bother with an example of an error in synthetic modeling. I'll wait for Greg Mankiw to come out in favor of raising the minimum wage, assuming he's fully reconciled to even having a minimum wage.

[Christine]

Speaking as an economist whose father was a meteorologist (not a climatologist, was a greenhouse sceptic but since his opinions sadly ceased to be updated over a decade ago I suspect this is rather immaterial), I like thinking about weather/climate and economics as somewhat analogous. But it really can be taken a bit far - for instance, no-one goes around trying to say we should re-create the world into something resembling the dish-pan, but many people do go around suggesting (often for rather weak reasons) that we should try to ensure increased competition in markets, precisely because of the existence of models of perfect competition. More importantly, the analogy probably tells you absolutely nothing about the reliability of a particular climate model.

Mixing up the general point on the analogy with the specific issue of economics and global warming: I think one of the reasons people get upset about economists making policy prescriptions on GHG is that economists really do first have to take some stand on what the science says - because we can't calculate costs and benefits without having done that - but then almost immediately we'll get complaints that we don't know enough about climatology to do that, or we're ignoring the 5% catastrophe scenario, or whatever. Not to mention the usual reluctance of lots of non-economists to accept a cost-benefit type analysis as valid under any circumstances. It isn't an easy situation to be in, but someone needs to think about possible policy options, after all.

Also Valuethinker: do economists get any bonus points for having Nicholas Stern among our ranks? What about Greg Mankiw? John Quiggin? Do we just always have to hear whining about Bjorn Lomborg or Ross McKitrick??