« Like a thunderstorm at O'Hare | Main | Boudreaux vs. Rodrik »
July 12, 2007
In which I contemplate a few of Max Sawicky's "Ten Boxes of Heterodoxy"
Max sounds his "barbaric yawp" over at the TPM Cafe.
Let's look at a few of his positions.
1. Supply and demand, 1. This celebrated and most basic economic model while in principle multidimensional in practice obscures anything interesting that affects market conditions. It bespeaks militant, ideologically-based reductionism. A good illustration is the minimum wage debate. In the usual S&D model, a MW can only reduce employment. Nothing else is logically possible.
We've been over this before. I am reminded of the old quote by Thomas Carlyle that I first heard when I was taking principles of economics, "Teach a parrot the terms 'supply and demand' and you've got an economist." Max's concern that the supply and demand model is taken to a reductionist extreme is nothing new. Unfortunately.
In principle it is multidimensional, but general equilibrium theory is hard to put into the format of a sound bite or an op-ed. The blog medium maybe does a little better at conveying the message because it allows for a more extended and thoughtful discussion.
My main caution about the supply and demand model in its simplest form is that it assumes homogeneous goods (or factors of production--to use the minimum wage example) and perfect competition. When the basic model is used in cases where those conditions are not satisfied, there is potential for mischief.
2. S&D, 2. The outcome in an S&D model in principle has no inherently attractive qualities, in and of itself, since it depends on the distribution of ability to pay. If Oliver Twist has no money to buy a crust of bread, his zero allotment is "efficient." The lack of any normative foundation is typically glossed over.
Or as I like to say, "No normative judgments, please; we're economists." Deirdre McCloskey as chided the profession with variations of that phrase as well, so that's where I picked that up. Max is right up to a point. The supply and demand model is value free in and of itself, but it can shed light on the effects of imposing alternative policies, and policies are generally based on some set of values or norms. It is true that economists often shy away from doing the latter. There are good reasons for being circumspect about recommending one policy over another, and indeed there is some value in specializing in the positive analysis. But we live in a political world, and our models can deal with that if we let them. There is no reason to suppress the normative discussion. The normative questions are the interesting ones--if we give up on those, our students will lose interest.
3. GDP. Add up all the Qs in the S&D models over the year ("final goods and services") and you get GDP. Solemn assurances that GDP is not synonymous with economic welfare fall easily by the wayside. More GDP (and less leisure time, less environmental quality, a less sustainable economic future) is always better. If terrorists knock down the Empire State Building, GDP could go up. More! Better! Comrade Stalin would approve.
See the broken windows fallacy.
7. Capital fundamentalism. As with reductionism of the S&D model, growth modeling zeroes in on private capital accumulation, even though a) other factors are demonstrably important and beg for attention; and b) private capital accumulation may be a consequence of other factors, rather than a cause and appropriate object for policy. Out of an obsession with this premise, the International Monetary Fund has screwed up a lot of countries too weak to ignore its advice.
Other factors are always important. If a model is simple enough to use, it necessarily leaves out many other factors. Those two fundamental facts are at the heart of more than just this criticism. In some research that a student of mine conducted we found that the relationships between growth, capital and other factors can be notoriously hard to identify. Yet, I think it is important to try. And yes, my confidence interval around theoretical growth models has increased once I started reading more of the empirical papers. Does that make me heterodox? Nah, probably not. But I know enough to know that some fundamental questions haven't been answered yet, and thus a little modesty is in order.
9. The unnatural rate of unemployment. Economists used to say it was 6.0, maybe 5.5. Lower would give rise to ruinous inflation. The huge social benefits of another couple of percentage points less unemployment were -- are -- implicitly discounted. Current rate is 4.5. 'Nuff said.
I think I know what he's getting at here. While I also am no fan of the natural rate concept of the old days, I come at it from a different angle. Max's statement here suggests that the social benefits to having a 4.5% unemployment rate were feasible back when economists thought the natural rate was 6%--and without the inflation we all feared back then. I'm not sure about that. The productivity gains of the 1990s did allow for lower unemployment and lower inflation.
When inflation and unemployment came down together in the 1990s, the Friedman-Phelps version of the natural rate gave way to the NAIRU (non-accelerating inflation rate of unemployment). It was not without its skeptics. Estimating the NAIRU is like finding your way around the room in the dark when someone keeps moving the furniture. It is easier to identify what it was in the past based on your experience, but there is no guarantee that it hasn't changed. Again, in what seems to be a common theme here, that would lead me to exercise caution in using the NAIRU to guide policy. Yet, the practical advantages in doing so carry a lot of weight, and so you use your past experience and hope that the furniture hasn't moved too much. And that means that occasionally you'll stub your toe.
But what mechanism would allow the policymaker to systematically do better?
10. "Power? You want the political science dept." Power looms over economic transactions, except in economic theory. Workers do not hire capitalists. Consumers do not choose merchants. Shareholders do not choose managers. Voters do not choose elected officials.
I have wrestled with this problem in my own models. It becomes difficult to produce clear predictions when you allow variations in power (e.g. bargaining power) to enter the model. It is similar to allowing preference shifts. Suddenly you can explain everything--and therefore you can explain nothing. Yet acknowledging the role of power (of various kinds) makes for a compelling narrative sometimes. It's something of a shame that it often needs to be excised from published research. It gives the impression that we don't care about it. That would, at least in my case, be a wrong impression. Yet at the same time, I believe it is appropriate to keep those power considerations separate from the rest of it, lest we end up with a model where anything goes.
I guess that makes me orthodox. Ultimately, I line up on the neoclassical free-market team more often than not, but I never stop asking many of the questions that Max does. And so I temper my free-market analysis with thoughts like: perfect competition is hard to find, rent-seeking in politics influences outcomes, real time estimates of the NAIRU should have a decent confidence interval, not all $1 transactions have the same social value, and power does matter even if it is hard to measure or even describe.
In some circumstances, those considerations are more important than others. But much of the time, they don't radically alter the main story that incentives matter--a story that the orthodoxy mostly gets right.
UPDATE: Max links back. Robert Waldmann also responds. Other blogs covering the topic include Greg Mankiw, Angry Bear and Cafe Hayek.
Posted by William Polley at July 12, 2007 12:36 AM
Trackback Pings
TrackBack URL for this entry:
http://www.williampolley.com/cgi-bin/mt-tb.cgi/771
Comments
"It becomes difficult to produce clear predictions when you allow variations in power (e.g. bargaining power) to enter the model. It is similar to allowing preference shifts. Suddenly you can explain everything--and therefore you can explain nothing. Yet acknowledging the role of power (of various kinds) makes for a compelling narrative sometimes. It's something of a shame that it often needs to be excised from published research. It gives the impression that we don't care about it. That would, at least in my case, be a wrong impression. Yet at the same time, I believe it is appropriate to keep those power considerations separate from the rest of it, lest we end up with a model where anything goes."
Well this gets to the crux of the problem. Most of us really don't care what ends up published in Econometrics. We do care when the Federal Reserve uses those results to raise rates anytime it looks like workers are making gains. We live in a world where real wage gains and drops in unemployment are seen as signs of potential inflation and so as dangers to be counteracted by monetary policy.
Excising power relationships from the literature may from one perspective be the only way to advance the field at large. But allowing policy perscriptions to be written on that basis while ignoring the fact that the actual results tilt to those with power is to my view willful blindness. Real world debates about Minimum Wage, Free Trade and Income Inequality are freely infused with argumentation drawn from what you admit is a simplified model. One that just happens to systematically favor the interests of holders of capital over wage earners.
It really doesn't matter whether your hearts are true, in the real world The Man is using you to keep us down.
I wrote a graduate paper on the Peasant's Revolt of 1381. Ostensibly it was a revolt against landowners, yet most of the actual physical violence was directed against lawyers. Generally this result had been dismissed by historians as essentially irrational. Yet on examination you can see that the economic position of the peasantry in regards to usage rights over common lands had been systematically reduced by a novel reinterpretation of land law over the two centuries prior. A reinterpretation that needless to say was carried out by lawyers. You can argue the rights and wrongs of the argumentation, viewed from the interior it seems perfectly consistent. Viewed from the outside the results were systematically in favor of landlords over tenants and the lawyers were blamed.
I am not suggesting that orthodox economists need to worry about peasants with torches, but a similar dynamic is happening here. Real world policy decisions on taxes and wages are being made in a way that from the outside seems systematically tilted to one side of the debate. Economists may think of themselves as Scientists, but in point of fact they are being used as Advocates. A defence based on "Hey I am just trying to get this article published in a peer reviewed journal" isn't going to suffice.
Posted by: Bruce Webb at July 13, 2007 12:19 PM
And if the point of public policy is to maximize GDP, then Kelo was a good decision. Thus begins another peasant's revolt--with a lot of "free market" economists leading the charge.
I agree with you to a point, at least insofar as Econ 101 models are carried to dangerous extremes by policy advocates. Furthermore, I would say that the way that the political/business establishment has hijacked the work of Adam Smith and twisted it into a greedy capitalist manifesto (which it is not) is part of the problem.
It is also unfortunate that the presence of those who advocate based on a stripped down Econ 101 version of a model provides a convenient target for vested interests on the other side. It's easy for anyone to poke holes in the simple "free market based" model as a description of most economic realities. This may lead some to accept uncritically the more interventionist position. Though I am sure you would say that this is the "free marketers'" own fault, do you not agree that it is an unfortunate state of affairs? By the same token, I find it unfortunate that some people drop "socialist" accusations at the drop of a hat.
The best path is, I think, in the middle. But that will be hard to get there if a qualified version of the orthodox model is pilloried by one side for being orthodox and by the other side for being qualified.
So what else is new? The middle usually gets trampled by both sides. This is not a 21st century phenomenon, nor it is unique to economics.
Nor will this problem, if we wish to call it that, go away.
And so it goes, as we retreat to our own sides to regroup before the next skirmish in the middle ground.
Posted by: William Polley at July 13, 2007 01:52 PM
"The best path is, I think, in the middle"
Thus thought the Dead Skunks and the Armadillos.
Me I am not quite so willing to give up the idea that the better solution to social problems lies in social solutions. And am not really that worried that you can't spell 'Socialism' without 'social'.
There is a good deal of evidence that 'Third Way' 'triangulation' leads to progressive policy becoming Road Kill. I am not willing to concede to Cato, I am not willing to concede that the New Deal and the Great Society were failures, which in the end is the presupposition of all 'middle of the road' arguments.
Posted by: Bruce Webb at July 13, 2007 11:01 PM
Nor am I willing to concede that the negative side effects of raising the minimum wage can be dismissed out of hand.
However, I am willing to admit that many social problems are best solved by social solutions. We may disagree about which ones, why, and how much intervention is needed, but I'm willing to listen to arguments to the contrary.
And I'm not worried about becoming road kill just for listening.
Posted by: William Polley at July 14, 2007 02:59 AM