« Suite Francaise, part II | Main | D'oh! »
Can macromodels and simulations be trusted?
The Economist runs a good article. Excerpt:
But how plausible were the numbers? Twelve years on, economists have shown little inclination to go back and check. One exception is Timothy Kehoe, an economist at the University of Minnesota. In a paper published last year, he argued that the models “drastically underestimated” NAFTA's impact on trade flows (if not on jobs). The modellers assumed the trade pact would allow people to buy more of the goods for which they had already shown some appetite. In fact, the agreement set off an explosion in the exports of many products Mexico had scarcely traded before. Cars, for example, amounted to less than 1% of Mexico's exports to Canada before the agreement. By 1999, however, they accounted for more than 15%. The only comfort economists can draw from their efforts, Mr Kehoe writes, is that their predictions fared better than Mr Perot's. A low bar indeed.
Here is Kehoe's paper.
Posted by Tyler Cowen on July 16, 2006 at 01:31 AM in Economics | Permalink
Comments
A mathematical model expressed symbolically is a nicely concise and unambiguous representation, and therefore good for communicating with other specialists. The Economist is right that a representation like that is almost impossible for a non-specialist policymaker or a layman to wrap their heads around, though.
I think part of the answer here is software. The article makes an analogy with flight simulators, and says that a pilot training in one need not know the details of how it works. The flight simulator does a lot to help the pilot out, though--it presents a nice visual display that summarizes the elaborate calculations going on behind the scenes.
Where are the economic model simulators? One could build software that would take a symbolic model and give you knobs for each of the parameters and a graphical display of the interactions. This would make it a lot easier for a policymaker to "fly" a model, and impart some intuitive sense of how it operated. It might even make decent TV.
Well, one can hope.
Posted by: mitch at Jul 16, 2006 8:34:35 AM
As for your macroflight simulator: University of Zurich has (maybe had, new curriculum looks different I THINK) something like this for one of their introductory macroeconomics courses: http://www.carlostrub.ch/oec/makro/
Scroll down to KoSiMo (needs Java).
I was not a particular fan of it but that might be related to that awful book it goes together (and the fact that I dont like macro anyhow)...
Posted by: Random Econ Student at Jul 16, 2006 8:56:23 AM
Economics has always tried to lord it over the other social sciences with its air of precision and use of numbers with decimal parts which implied an ability to understand social phenomena and predict the future.
It seems to me that basically this is a fraud. The reality is that not only do the big models not model reality but even the base assumptions are not understood.
As an example, consider the supply and demand curve. As cost increases, demand goes down. (as mentioned in the article, no one can know the shape of the curve). The problem is that we don't even know what cost is. The only way to understand it is retroactively.
In a famous day care center example, the centers tried to decrease late pickup by parents They introduced a fine to increase the money price for late pickup. Unfortunately late pickups increased. RETROACTIVELY it was realized that the cost of late pickup included a large moral component that they had effectively reduced by adding a money fee.
This uncertainty is true for all of the assumptions on which these models are based.
Posted by: RobbL at Jul 16, 2006 9:38:13 AM
Is he referring to Ross Perot? Who says Ross Perot was wrong?
Come to Ohio and I will show you the damage.
Economists need to be better attached to reality.
Posted by: save_the_rustbelt at Jul 16, 2006 2:06:43 PM
Hey RobbL,
Are you forgetting about elasticity? Some parents are willing to pay in order to be late, so they'll continue to demand a certain quantity of day care.
Posted by: Brian at Jul 16, 2006 10:55:33 PM
There's a distinction between comparative static calculations and forecasts that needs an airing. The CGE estimates of the early 90s were comparative static calculations, NOT forecasts. Their estimates can only be evaluated after controlling for other sorts of economic change (e.g. technological change, other aspects of globalization, macroeconomic shocks, etc). One can argue about what other sources of economic change were important enough to be included in an assessment like Kehoe’s (much as one might argue about what right hand side variables to include in a regression). Kehoe’s assessment implicitly treats NAFTA as the only major event that might have affected the trade pattern.
And mitch: Policymakers are often given tools to “fly” the models, as you suggest. But most know they don’t really understand the models and prefer to turn them over to professional “pilots”. Thus we are back at square one, relying on the judgement of the modeller.
The biggest problem in communicating with the public is the reliance on point estimates rather than distributions of outcomes. Most politicians use model results as advertising for their product, and not for analytical purposes. One can almost always count on the media to oversimplify model results by reporting only the central estimate. This has been a notable failing of the media (and the blogosphere) in the social security debate, for example.
Posted by: anon at Jul 17, 2006 2:25:49 AM
The problem is that we don't even know what cost is. The only way to understand it is retroactively.
Posted by: שש בש באינטרנט at Jul 17, 2006 4:06:07 AM
Brian,
I think RobbL's point is that in the Day Care example, the day care thought, ah people are taking advantage of the fact that there's no charge for picking up kids late, so we'll charge people and they will pick up their kids on time, because they don't want to pay more. What the day care center apparently didn't realize is that the largest component in getting people to pick up their kids on time is not cost, but guilt, that is they feel guilty picking up their kids late when they're supposed to be their at 5 (or whatever) after all that's all you've paid for and I'm sure that the day care provider has something else they'd rather do then take care of your kids. So the guilt was a strong motivator to get people to pick up their kids on time. Putting a monetary fee on late pickup though assuaged that guilt, after all, if you pick up your kids late, hey you pay the fee, like a late movie. Nobody think's I take advantage of Blockbuster when I return a movie late and pay the fee (before no late fees), instead, it's just a term of the contract. The lesson, be careful before you assuage a non-monetary cost, with a monetary one, there are plenty of people that would just rather pay, and guilt is a very useful motivator.
Posted by: Joel B. at Jul 17, 2006 10:41:59 AM
Did anyone actually read the Kehoe paper that the Economist article was based on? Looks like Ross Perot was right. Check out Table 5 (http://www.econ.umn.edu/~tkehoe/papers/NAFTAevaluation.pdf). Mexico’s exports were predicted to rise 50.8%. They actually rose by 140.6% Imports rose 50.5% versus a prediction of 34%. Sure looks like a “big sucking sound”. Of course, it didn’t do Mexico much good. GDP growth from 1994 to 2005 was only 2.7% per year (Chile was 4.8%).
Worse, US farm exports devastated rural Mexico triggering a tidal wave of illegal immigration that is savaging much of America today. Any gains to America farmers (wheat and corn are global commodities) were trifles compared to the vast costs to our overall society.
The net is that NAFTA has been a disaster for both countries. AMLO, to his credit wanted to repeal many of the agricultural provisions. Of course, cheap labor elites in the US and crony capitalists in Mexico have done well… And they have their apologists...
Posted by: Peter Schaeffer at Jul 17, 2006 2:41:05 PM
Oh yes, I absolutely hate buying things for lower prices. I also deplore the opportunity for poor people to better their lives.
"Disaster" indeed. "Savaging" indeed.
What a crock of protectionist, xenophobic nonsense.
Posted by: Noah Yetter at Jul 17, 2006 3:13:16 PM
Noah Yetter,
When you start sending your children to schools dominated by illegals, you can protest the word “savaging”. When your income crashes because of Open Borders you can object to “disaster”. When you can’t afford a home because of runaway population growth, you can tell us how great mass immigration really is.
Wages for production workers have been falling since 1973. They are now down to the 1959 level. Our nation can pay for only 53% of its goods imports via exports. All the rest is debt.
Exactly who benefits from “lower prices” with real wages declining for 33 years? “Cheap” imports paid for via debt are a well trodden path to national ruin.
Try looking at the facts, not libertarian ideology.
Posted by: Peter Schaeffer at Jul 17, 2006 8:02:56 PM
genegenheid babe ^^^ bbw bildergalerien ^^^ orge gif ^^^ coy infermiera ^^^ erotique asiatique ^^^ froid amateur masturbation ^^^ cul quarante ^^^ papa chat ^^^ faustficken lesben ^^^ sexy geweitete muschi ^^^ teiniseksi ematin huora ^^^ penis pilavideot huora ^^^ anselig fillesak ^^^ mye geisha ^^^ varmere penetrering ^^^ kaldest overnaturlig kysk ^^^ geladarisa mouni irinika ^^^ agrios moro avnanismos ^^^ plus frais maman photos ^^^ mere fin ^^^ to pio drosero apofasistikos pateroulis ^^^ zesto apofasistikos pateroulis ^^^
Posted by: levan at Sep 11, 2006 3:30:11 AM
Last chaos Gold
Buy Last Chaos Gold
Cheap Last Chaos Gold
Posted by: aion kina at Mar 21, 2009 1:03:07 AM
