Mathematics can be beguilingly elegant. It can also be dangerous when people mistake its elegance for truth.Bloomberg
Beware of Economists Peddling Elegant Models
Mark Buchanan
(h/t Ryan Harris in the comments)
Mathematics can be beguilingly elegant. It can also be dangerous when people mistake its elegance for truth.Bloomberg
10 comments:
The usage of the phrase elegant is giving too much respect for such stuff.
I've always felt that finding the simplest solution to a complex problem was elegance.
The simplest solution is the best. But often there is no simple solution, especially to complex problems. If you have a simple solution to a complex social problem then you are looking at a very narrow slice of reality which isn't realistic. If you go to the publications from the central banks of your choice anywhere in the world and work through a couple of the papers. You will find, they use some short hand notations to give themselves the appearance of soundness of argument in their models but don't be put off by it, and keep pushing and asking yourself at every paragraph, what are they doing wrong, criticize the flaws. At first you think, you surely are missing some finer point, these are intelligent people, they can't possibly be so stupid. But as it turns out, they really have all gone bat shit crazy and are pursuing simple minded abstractions appropriate to an adolescent but dressed up in pseudo-notation. You'll notice few provide sources of data upon which their analysis is based and those that do take quite a few liberties. Not that there aren't some good ideas mixed in. The people aren't idiots or morons by any stretch. They are just misguided in their approach and have been poorly educated. (Wikipedia up folks like Stiglitz and Lucas Jr to get an understanding of how economists see the logical problems and compare their analysis to reality if you find it interesting) I've had to carry these simplistic models into computer programs and it has given me a whole new disrespect for DSGE.
The economists have high profile debates over whether or not it matters that they are always wrong. They always conclude, it doesn't matter, everyone else just doesn't get it: Invisible clothes.
Lars Syll is doing good work lately and has knowledge and foundations to really take some big name economists apart. I hope and suspect in the near future, with his arsenal of knowledge, he will move from exploring the types flaws made to singling out bad papers. If even one or two articulate, logical scholars like him should be assigned to work at the editing desk BEFORE papers get published, the peer reviewed work would look very different. It is daunting though, imagine yourself going after a body of work by nobel prize winning economists. Dozens of other economists will circle their wagons because all their reputations are intertwined. Messy business admitting errors and taking responsibility for the enormous damage they have caused to the world.
"But often there is no simple solution"
The trouble with this statement...even though it may be true...is that it can be difficult to find the simple solution but that doesn't mean one doesn't exist.
In math and physics a simple relationship is often the foundation of entire bodies of work.
F=ma is the basis of mechanical physics...from that the rest of it can be derived.
The sectoral balances identity...a model of a closed system...is another example. All of monetary economics is constrained by the reality of that system....Spending=Income.
Trying to examine a complex economic system in it's entirety is practically impossible...breaking it down into smaller modules makes a solution, or at least an understanding of that particular part of the system, possible. Then the interaction between systems can be examined.
This is how engineers solve complex problems.
We are easily confused by the smoke and mirrors clouding every issue...largely by the cloak of conventional wisdom.
We need better skills for defining the problems. We seem to spend most of our time solving the wrong ones.
The simple equilibrium at the center of mmt, in the sectoral balance, appears simple. But put yourself in the place of a policy maker who was going to design forward looking recommendations to congress and the fed. How much spending? Where should rates be? How should it be balanced given the other sectors? If you had to try to calculate where spending and rates should be exactly -- you can't, it is incredibly complex and requires alot of assumptions and cluges. Hence, the simple solution is to have buffer stock of employed people to give space for error without creating misery by design. The solution however is not to pretend the problem is simple, assume away the complexity, and create a model that portends to give insight into the nature of relationship. You can work around complexity and keep it simple by using things like transforms and buffers and other tricks. You can't just create a pretend simple world.
Ryan,
Unemployment tells us that we aren't spending enough, inflation tells us we're spending too much.
We can make it a lot more complex than that but those are the problems that have to be solved. Our problem isn't lack of supply, it's lack of demand.
The problems we experiencing now are not from lack of resources...it's from lack of liquidity. That isn't complicated unless you choose to make it so.
I believe the MMT academic community has designed the policy approaches that can deliver the desired outcomes. We are deluged with so-called "conventional wisdom" that says otherwise, and we are paralyzed by it, unable to move forward.
We've already tried "their" solutions with resounding failure.
We have to do something based on our beliefs in what we know and make adjustments from there.
I agree, Paul, that MMT keeps it simple in those important ways. But at some point someone has to put a numerical number on what needs to be done NOW and six months from now. Unemployment, trade, capital flows and other data are collected slowly and congress moves glacially. So somewhere along the line someone has to make informed predictions and has to craft policy using the data collected to guide decisions and be able to answer what-ifs and what-thens. Those predictions, using a heterodox understanding of complex systems can be done much more accurately than the orthodox folks have been doing. The problem is they completely dominate government at all levels. From the central bank, to the congressional budget office, no one else has a seat. The point of all these articles and criticism is to get popular support among the businesses that drive politics into understanding that their businesses would be better off if they remove the orthodox modelers from policy making ventures.
Ryan: "The point of all these articles and criticism is to get popular support among the businesses that drive politics into understanding that their businesses would be better off if they remove the orthodox modelers from policy making ventures."
That's a really interesting point. We spend a lot of time talking about our elected representatives and maybe we ought to be targetting the business community. Afterall they are the entities that form "elite' opinion. If they can be enlightened, then they'll watch the mainstream news media in bewilderment as well.
And the business community will be the force for change. Show them how/why it's in their 'business interest' to get this stuff right. While their may be a small minority of plutocrats and oligarchs that, as Rodger Malcolm Michell stresses… want to widen "the gap"… I reckon most people, most businesses owners, if given the chance to understand, would support more progressive public policies if they thought it best for their business.
Ryan, the problem is decades of brainwashing.
The perception amongst business and corporate elites is that 'governments can't do anything right' or 'they are the problem, not part of the solution'. And from there, to all the market fundamentalism.
Obviously they are going to hire economists whose models say what they exactly want to hear, and push into public service positions the same cadre. And so we have a big giant circle jerk of business men, economists and politicians and policy makers singing the same song.
This campaign started from the big corporate world and and top-down penetrated every place from the public sphere. And as money flows, the problem is of difficult solution (specially when the corrupt status quo just reinforces the vision of the electorates of the different countries that the government is part of the problem).
Quite a conundrum.
Economics, ie "house-law", isn't like predicting the weather.
Most economists seem like libertarians to me, both left and right.
"No one is in charge... this is like the weather... we dont know what is going to happen... etc..."
WE humans have been set under authority over and are in complete control of these economic systems, perhaps a libertarian cannot even see this.
Real inputs (flood, drought, pestilence) can influence our decisions ex post, but our own future decisions made in response to these REAL unknowable inputs cannot be "modeled"... this is why we humans have been given intelligence...
This whole "modeling" thing is ALL absurd to me... we should be 'monitoring' not 'modeling'...
FD: I am NOT a libertarian ... rsp
Post a Comment