The Myth of the Model
Most people don’t notice it, but “model” may be the most dangerous word in the English language right now. Models justify a lot of the bad policies that have been, or soon will be, foisted on us. For example, what was used to justify the fiscal policy of the big “stimulus”? That’s right. And as I wrote this, “experts” were using models to gear us up for another one.
More than a year after the original “stimulus,” not only are economists nowhere near consensus about its effects but few if any of the models used to justify it have turned out to be right. Obamanomic adviser Christina Romer, for example, has come under heavy criticism because her team’s plan has performed abysmally. The model behind the plan predicted unemployment would peak at 8.3 percent. It exceeded 10 percent before dropping back slightly. In defending her plan she appealed to counterfactuals—that is, how bad things could have been without it. That her team failed to reach its rosy targets, she says, “prevents people from focusing on the positive impact.” But did Romer ever consider the possibility that her model was just wrong?
When it comes to prediction and explanation, macroeconomic models are often just as bad after the fact as before it. There are just as many debates raging about the effects of the “stimulus” as explanations of the crisis used to justify it. Consensus consistently eludes us. Almost all the arguments presuppose models. There are Keynesian models, “new” Keynesian models, and unbranded models proffered by leading economic lights like Harvard’s Robert Barro. Comparative analyses of these positions offer little except further evidence that, as Stanford’s John Taylor writes, “[T]here is no consensus.”
But why? These people aren’t stupid. I’d like to suggest in nontechnical terms why the problem might be with the models themselves.
“Mainstream macroeconomics is ‘hydraulic,’” writes recovering macroeconomist Arnold Kling at EconLog. “There is something called ‘aggregate demand’ which you adjust by pumping in fiscal and monetary expansion. I wish to reject this whole concept of macroeconomics.” I think he’s right. In fact, as I’ve argued in these pages before, economies are not pumps to be primed, but economic ecosystems. Economists are thus notoriously bad at predicting, much less planning, economies. That’s why next-generation economics must focus on the fundamentals—namely, the rules that give rise to successful entrepreneurship and sustainable growth. Not aggregates. Not models. Rules.
“Institutions form the incentive structure of a society and the political and economic institutions, in consequence, are the underlying determinant of economic performance,” said Douglass North in his Nobel Prize lecture. And he’s not alone in the wilderness. James Buchanan, another Nobel Prize winner, says economists are simply asking the wrong questions:
How do markets work? Standing alone, this is an inappropriate and unanswerable question. It must be replaced by the question: How do markets work under this or that set of constitutional and institutional constraints? Economists’ scientific expertise can be brought to bear on the predicted effects of alternative sets of constraints. The relevant question is not how this or that outcome may be put in place through possible collective or political action. The question becomes instead, how can this or that set of constraints be predicted to operate so as to allow the generation of an order that meets certain criteria of desirability? The difference between the two methodological stances may appear minor, but much ill-advised effort might be avoided if economists would recognize the limits of their own discipline. [“Economists Have No Clothes.”]
Buchanan not only laments the poor framing of economic questions, but also says that even those attempting postmortems of the 2008-09 financial crisis have been “embarrassed by their inability to offer ‘scientific’ explanations.” Those who’ve dared to come forward have offered little more than warmed-over Keynesian nostrums. Just think of folks like Romer, Paul Krugman, and Joseph Stiglitz—all of whom have recommended various versions of doubling down on the failed policy. At the end of their yellow-brick road? A curtain, behind which lies a model, behind which lies an agenda.
So what do all these macroeconomic models have in common?
- They’re rendered either in impenetrable math or with sophisticated computers, requiring a lot of popular (and political) faith.
- Politicians and policy wizards hide behind this impenetrability, both to evade public scrutiny and to secure their status as elites.
- Models vaguely resemble the real-world phenomena they’re meant to explain but often fail to track with reality when the evidence comes in.
- They’re meant to model complex systems, but such systems resist modeling. Complexity makes things inherently hard to predict and forecast.
- They’re used by people who fancy themselves planners—not just predictors or describers—of complex phenomena.
None of this is to argue we should do away with measurement. Temperature, blood pressure, and other indicators are all useful data for telling whether you’re healthy. But using proxy measures to determine an economy’s health is a far cry from using models to reduce dynamic systems to steady-state snapshots.
What does this mean for economics as a discipline? I think it’s time we admit many economists are just soothsayers. They keep their jobs for a host of reasons that have less to do with accuracy and more to do with politics and obscurantism. Indeed, where do you find them but in bureaucracies—those great shelters from reality’s storms? Governments and universities are places where big brains go to be grand and weave speculative webs for the benefit of the few.
And yet “ideas have consequences.” Bureaucracies are power centers. So we have a big job ahead of us. We’ve got to do a seemingly contradictory thing and make the very idea of complex systems simple. How best to say it? Economists aren’t oracles? Soothsaying is not science? Ecosystems can’t be designed?
“The very term ‘model’ is a pretentious borrowing of the architect’s or engineer’s replica, down-to-scale of something physical,” says Barron’s economics editor, Gene Epstein. “These are not models at all, but just equations that link various numbers, maybe occasionally shedding light, but often not.”
Take this as a throwing down of the gauntlet. Macroeconomic wizards owe us more than the circular justifications for cushy jobs.
Likewise, we have to explain that a scientist’s model, while useful in limited circumstances, is little better than a crystal ball for predicting big phenomena like markets and climate. It is an offshoot of what F. A. Hayek called the “pretence of knowledge.” In other words, modeling is a form of scientism, which is “decidedly unscientific in the true sense of the word, since it involves a mechanical and uncritical application of habits of thought to fields different from those in which they have been formed.” A model is thus a cognitive shortcut for both the wonk and the journalist, the latter of whom wants to peg his story to something authoritative the wonk has to provide. At the receiving end of this wonk-writer alliance are the rest of us—with little besides common sense as a shield. And I don’t mean this as populism. It is rather a defense against scientism launched from the turf of Austrian economics.
Complex phenomena can be counterintuitive. Sometimes they require scrutiny by experts to make sense of them. Notwithstanding their expertise, experts are just as often wrong as right. Can we base policy decisions on what amounts to coin flips? Models are a means of making the most fragile of hypotheses seem strong and substantive. But the only thing we can really predict is that they’ll eventually shatter.
Pingback by “What Sort of Despotism Democratic Nations Have to Fear” | The Freeman | Ideas On Liberty on 20 May 2010:
[...] It’s not what we don’t know that gets us into trouble. It’s what we think we can model but can’t. Max Borders explains. [...]
Comment by Loren Crawford on 25 May 2010:
It is as Doyle’s Sherlock Holmes explained: “We twist facts to suit theories instead of theories to suit facts.”
Pingback by Worth a Read | Taking Hayek Seriously on 26 May 2010:
[...] a Read May 26, 2010 | Posted by Greg Ransom “The Myth of the Model” by Max [...]
Comment by Samuel Clemens on 27 May 2010:
The challenge is not that models don’t work, it is that models don’t work everywhere. Complex systems require a complex set of models which may each be mutually contradictory.
Pingback by Recomendaciones « intelib on 31 May 2010:
[...] Recomendaciones The Myth of the Model, by Max Borders [...]
Comment by N. Joseph Potts on 2 June 2010:
The use of models for bamboozling the hapless masses is by no means limited to economics. Very much the same powermongers who collude with economists collude also with various (also government-funded) natural scientists to promote the hoax that government can help with “global warming.”
People who think government does a good job of “managing” economics, of course, are pleased also to support government management of the weather.
Comment by lrshultis on 6 June 2010:
Models are hypotheses as to how things work. A model must satisfy all the known evidence and any predictions of the model must correspond to any future evidence. If it does not do either of those conditions, then it must be discarded. If it satisfies those conditions without fail, it becomes a theory, but may be discarded or used under certain strict constraints when it no longer corresponds to evidence. For example,the Newtonian theory of mechanics is a theory which is valid under certain restraints on velocity and gravity.
As with any mathematics–the science which studies all possible relations and is used in concept formation in the differentiation among mental percepts–one should never reify a model or one will find oneself in a fantasy world causing all sorts of mischief.
Comment by Daniel Shapiro on 8 June 2010:
Can you model human behavior in aggregate? Impossible! I can’t even model my own behavior. If economics is a derivative of human nature then economics cannot be coherently modeled. If politics is coercion (observably true), then this can be modeled (or extrapolated) such that over time a prison-like society emerges wherein indivdual choices are dictated entirely by an increasing quantity of political edicts.
Comment by Richard W. Fulmer on 8 June 2010:
If you assume that the purpose of economic models is to reflect reality then you are correct, they are worthless. This assumption, however, must be accompanied by an explanation of why sane, intelligent people continue to use such flawed tools decade after decade even though they offer little in the way of explanatory or predictive power.
On the other hand, if you assume that the models’ purpose is to provide economists’ political patrons with a rationale for spending more tax dollars, then they work extremely well, and no further explanation is required to account for their continued use.
Comment by Steven Hankin on 8 June 2010:
Could it be said that modeling of macro-economic matters is tantamount to treating macro economics as a true science. A true science, of course, means that experiments can be performed in which no constant can be assigned to all the variables, other than the one (or ones) that we wish to test. In the case of macro-economics problems, the factors, that are in play, may be nearly unlimited and even many of the ones, that we can identify, can not be adequately expressed as a numerical constant. Hence, it appears that most macro-economic problems are not really susceptible to the scientific inductive approach and therein lies the problem with making predictions–based upon modeling– with respect to most macro-economic problems,
Comment by Steven Hankin on 8 June 2010:
Please read this revised comment, instead of my previous one. I meant to say “experiments in which constants can be assigned” instead of “experiments in which no constants can be assigned”
Could it be said that modeling of macro-economic matters is tantamount to treating macro economics as a true science. A true science, of course, means that experiments can be performed in which constants can be assigned to all the variables, other than the one (or ones) that we wish to test. In the case of macro-economics problems, the factors, that are in play, may be nearly unlimited and even many of the ones, that we can identify, can not be adequately expressed as a numerical constant. Hence, it appears that most macro-economic problems are not really susceptible to the scientific inductive approach and therein lies the problem with making predictions–based upon modeling– with respect to most macro-economic problems,
Comment by Tom on 10 June 2010:
In my field (systems engineering) we have a saying, “All models are wrong, some models are useful.” The obvious implication is that some models are NOT useful.
Comment by Drik on 13 June 2010:
Ecosystems CAN be designed. They are all the time. Witness Bioshpere. And they will even function for a short while. But after a while the simplification breaks down and if you stay in it, you die.
Prediction based on complex systems works almost as well in economics as it does in weather. Chaos theory predicts that in a sufficiently complex system, any iterative model’s prediction ends up as accurate as a guess.
Our ability to model the complexity is much like the new man of Shelley’s Frankenstien. A crude representation of the original. And after a while it accidentlally kills you.
Comment by Mo on 14 June 2010:
All models are wrong, some are useful.
Comment by William Anderson on 16 June 2010:
Excellent article, and I think this is one of the most important articles in the Freeman I have seen in a while. For all of their “complexity,” the macro models are utterly crude, for they ultimately depend upon throwing everything — all economic relationships — into a single line called “aggregate demand,” and that in and of itself is meaningless, economically speaking.
Yet, I have seen economists like Paul Krugman claim that because these models use sophisticated math techniques, any criticism of them by people who cannot do all of the math is illegitimate. There is a more important thing to remember about mathematical economic models: garbage in, garbage out.
Pingback by The Myth of the Model :The Thinker on 18 June 2010:
[...] This Freeman Online article by Max Borders posits that the word model “may be the most dangerous word in the English language right now.” Models justify a lot of the bad policies that have been, or soon will be, foisted on us. For example, what was used to justify the fiscal policy of the big “stimulus”? That’s right. And as I wrote this, “experts” were using models to gear us up for another one. [...]
Pingback by There a LOT more to “Road to Serfdom” | Taking Hayek Seriously on 21 June 2010:
[...] to the success of competitive market actors. The idea that we can aggregate economic data – like macroeconomic data – in order to say something meaningful about how things should to be coordinated is, well, a [...]
Comment by Chuck Logan on 16 September 2011:
Like all other forms of human action, economies are based on, and evolve through, trial and error. Trial and error cannot be modeled, and it works best when it is interfered with least.
Comment by Richard A. on 12 October 2011:
Models are useless at predicting reality for the simple reason that there is no scientific way to construct a measure of any kind of a system that is based on subjective value. Money adds a common factor to all exchanges and just because it’s in base ten units people start thinking it can be added and subtracted, multiplied and divided, and graphed! It must be graphed so we can draw tangents and figure out the area under the curve. Of course money may as well be, and once was for some economies, sea shells. When people realize that all the key econometrics and aggregates models use are basically calculating in wampum, sometimes to three or more decimal points, they’ll realize how meaningless these models are. But then why still use them?
Because people need tools to ‘manage’ the economy, and the models are all that’s available. The alternative is that the economy doesn’t need a manager, and even if it did elites with models would be no more informed than any other random citizen. This possibility calls into question the entire political apparatus in every country, and the elites they employ as apologists for their failed policies. That is why otherwise intelligent people continue to use models. Consciously or not, they know their livelihoods depend on it.
Comment by England Patriots Season tickets on 4 February 2012:
I just want to inform you that I am just all new in order to blogging and definitely loved youre website. Probably Im going to bookmark your website . You possess outstanding posts. Kudos with regard to discussing around your blog.
Comment by a3778739 on 5 February 2012:
I’ve said that least 3778739 times. SCK was here
Comment by vacation rentals in siesta key on 5 February 2012:
You have mentioned very interesting points! ps decent site.
Comment by Karisa Cawein on 6 February 2012:
you’ve got an important blog here! would you wish to make some invite posts on my weblog?
Comment by InterDinosaurs.co.uk on 6 February 2012:
Does your blog have a contact page? I’m having problems locating it but, I’d like to shoot you an email. I’ve got some ideas for your blog you might be interested in hearing. Either way, great website and I look forward to seeing it expand over time.
Comment by london photocopier on 6 February 2012:
I would like to thnkx for the efforts you’ve put in writing this website. I am hoping the same high-grade site post from you in the upcoming also. In fact your creative writing skills has encouraged me to get my own site now. Actually the blogging is spreading its wings fast. Your write up is a good example of it.
Comment by medway electricians on 6 February 2012:
As I site possessor I believe the content matter here is rattling excellent , appreciate it for your hard work. You should keep it up forever! Best of luck.
Comment by Pawnbrokers on 7 February 2012:
The subsequent time I learn a blog, I hope that it doesnt disappoint me as a good deal as this one. I mean, I know it was my option to read, nevertheless I genuinely thought youd have something attention-grabbing to say. All I hear is a bunch of whining about something which you possibly can repair when you werent too busy on the lookout for attention.
Comment by kent seo company on 7 February 2012:
I precisely wished to thank you so much once again. I do not know the things that I could possibly have tried without the type of information revealed by you relating to this situation. This has been an absolute alarming scenario in my position, but taking note of your professional style you treated that took me to weep for delight. Extremely thankful for your support and in addition have high hopes you are aware of an amazing job you are always doing teaching others through the use of your web blog. Most likely you have never come across all of us.
Comment by luty 6 on 7 February 2012:
F*ckin’ remarkable things here. I am very glad to see your post. Thanks a lot and i’m looking forward to contact you. Will you please drop me a mail?
Comment by Carlton Kjellberg on 7 February 2012:
some genuinely good posts on this site, thanks for contribution.
Comment by bad breath remedies on 7 February 2012:
Hi just wanted to give you a brief heads up and let you know a few of the pictures aren’t loading properly. I’m not sure why but I think its a linking issue. I’ve tried it in two different browsers and both show the same results.
Comment by office photocopier on 7 February 2012:
Thank you a lot for sharing this with all people you really realize what you are speaking approximately! Bookmarked. Please additionally visit my website =). We can have a hyperlink change agreement between us!
Comment by fotografia slubna bydgoszcz on 7 February 2012:
i fucking know how to do this it has a godly little magic secret that comes in the box. its called an instruction manual. DUHHH
Comment by adwords management in india on 7 February 2012:
The video card (sometimes referred to as the GPU) is an additional vital component in any gaming rig, as it handles almost all with the graphics for your video game titles. A person with the issues with video cards – in particular the center and minimal end ones – is that they tend to come to be obsolete faster than the other components of a gaming pc. Commonly, acquiring a higher conclusion video card when you’re upgrading or building your gaming rig is vital as it presents you breathing area in advance of it’s time to upgrade once more.
Comment by office partitions auckland on 7 February 2012:
thanks !! very useful submit! http://www.ofis.getlisted.co.nz/office-partitions-auckland
Comment by sexfotka on 7 February 2012:
The first time I heard Rush Limbaugh for the radio in 1991, I heard him referring towards women as “Feminazis” not to mention his disregarded anything she has had to say seeing that that day. Others should do a similar. He is, simply put, a nut! sexfotka
Comment by free pantyhose clips on 8 February 2012:
Loving the information on this site, you’ve done excellent job on the posts .
Comment by vending machines in school on 21 March 2012:
There is noticeably a bundle to realize about this. I assume you created various good points in capabilities also.
Pingback by Insights from Author, Max Borders – Hamilton Society of Wall Street on 28 March 2012:
[...] the Library of Economics and Liberty 4. "Cash for Clunkers" – another piece for The American. 5. "The Myth of the Model" – in which Max criticizes macroeconomics (especially Keynesian [...]