I normally blog about issues related to finance or distribution, but this time I thought I might try something somewhat different. I’m not particularly expert at questions of economic methodology, but found this paper to appear in the latest Journal of Economic Methodology by my very smart colleague at UMass Boston Peter Spiegler to be really interesting.
Over the last decade or so we have all had to endure (I use the term advisedly) what may be termed the ‘economics made fun’ (EMF) genre of books. The bestseller Freakonomics for example has had enormous success, selling over 5 million copies to date and launching a whole cottage industry of books (some better and some worse that blended pop culture, armchair theorizing and quantitative economic methods to come up with surprising and ‘fun’ (counterintuitive or shocking) conclusions.
And then there was the inevitable backlash as economists took to criticizing pop-econ for its oversimplification and its propensity to aim for entertainment rather than popularization of economics ideas and sometimes downright bizarre conclusions (“Do comics make cats dogs?” as the cartoon suggests is just about right). The fundamental claim made by these criticisms is that there is a no connection between what economists do in their serious work and the overstretched just-so stories told in the EMF canon.
Peter is having none of it. As his abstract suggests
Many critics of the Economics-Made-Fun (“EMF”) genre have charged that it contains very little actual economics. As such, it would seem that criticisms of EMF do not apply to economics more broadly. In this paper I take a contrary view, arguing that, in fact, at a deep conceptual level the engine of EMF analyses is precisely the engine of mainstream economics. Specifically, I argue that both EMF and mainstream economics rest on a conceptual foundation known as the Principal of the Substitution of Similars (“PSS”). Understanding how PSS leads EMF practitioners to make claims well beyond what is warranted by their analysis also offers insight into how PSS can put economists in danger of overestimating the power and scope of their analyses. I explore the consequences of such problems through examples of economic analysis of the U.S. housing market in the lead-up to the recent financial crisis.
The target here is something wider—viz. the propensity of economists to graft assumptions about behavior on objects of inquiry without assessing whether these are either defensible or useful in understanding their actual behavior. We all of course have our own favorite outlandish examples of models that seem to make some very strong assumptions about the behavior of certain subjects in order to render it legible in economese: models of torturers as maximizing information gathering subject to a retribution constraint, theories about self-interested populations choosing between ‘cuddly’ and ‘cut-throat’ capitalism by comparing long term welfare between them and so on. Peter suggests we must ask a critical question prior to such analyses in order for economic theorizing to be illuminating: how plausible is the assumption that the object of inquiry is capable of behaving in the way that the models suggest? In many cases it will no doubt be plausible, but there are equally many cases in which it will not be the case. The fact that economists sometimes pay no attention to such a question suggests what may be termed an ontological complacency in which the behavior of the object is presumed rather than discovered.
Peter locates this propensity in a particular dubious philosophical maneuver due to Jevons—the principle of substitution of similars (PSS)—that is embedded deep within much of economic thinking. While Jevons’ ideas about utility and its mathematization are relatively familiar, his conceptions about logic are less widely known. Put very simply the PSS suggests in Jevons’ words that “ Whatever is true of a thing is true of its like”. The practical implications of this principle hinge on just how ‘alike’ two things need to be in order to have this ‘capacity of mutual replacement’ (again, Jevons’ term). For reasons Peter explains in his paper, Jevons’ beliefs about the nature of the social world led him to set the bar quite low, and this made the message of the PSS something akin to ‘if two things are alike even just in superficial ways, they are alike all the way down’. If one takes the principle seriously, then there is a deep connection between the behavior of Sumo Wrestlers and School Teachers (to use Dubner and Levitt’s canonical example) since both have personal incentives, one can comfortably proceed from that front and ignore other information about their behavior.
But the same pattern of ontological complacency is present, Peter claims, in more `serious’ academic economics. His (surprising) example is of the academic literature on the housing crisis. Post facto, one can reasonably claim that the legal and institutional changes in the US had transformed financial markets transactions from facilitating the functioning of the home ownership market to a token in a rigged Casino. But for the most part that this was missed because the highly relevant classes of information available on the actual functioning of the markets from news media, lawyers and even anthropologists were not adequately examined and absorbed. As a result, Peter claims, there was a high profile failure in assessing the housing bubble.
So what is to be done? Peter’s project is to pursue a change in economic methodology in which a set of standards are established to assess how plausible it is to assume that, say terrorists are maximizing a particular expected utility function, or that new products on the market behave like the old ones. As he puts it:
Doing so will require methods that are not currently in the standard toolkit of economics. Specifically, because what is needed is a set of standards for assessing the plausibility of one’s ontological presumptions in the study of a particular set of social phenomena, the assessment standard must include the ability to entertain alternative ontologies and to understand how to read ontologies of social contexts from the perspective of the individuals that constitute that context.
How likely is it that this can be developed? How convincing is Peter’s case? I’d be curious to know what readers think.
Triple Crisis Welcomes Your Comments. Please Share Your Thoughts Below.
I discovered the tendency to ignore information unless it was expressed in proper economics two decades ago. In a previous, and much more local, housing market crisis. Community activists in the Bronx were claiming the people were buying apartment buildings, flipping them to each other to inflate their prices, and increasing the rents because of the increased costs. Every economist who heard this (and I include myself) laughed, and thought “David Riccardo – corn isn’t high because rent is high, …” and pretty ignored it. Of course, it turned out that there was a rampant property flipping scheme happening, in which sales to straw buyers created comps to use when financing the next bogus purchase, with the intent of borrowing far more than the properties were actually worth, pocketing the proceeds, and sticking FHA and Freddie Mac with the losses. If you don’t intend to hold the property for long, there is no reason NOT to jack up the rents, since people will stick around and pay for a little while, and it helps support the bogus appraisals. The community groups had the chain of causality more or less, but the raw facts completely right. I learned from that experience not to dismiss facts just because they are couched in a logical chain that doesn’t fit my priors. I’m not sure that a lot of my colleagues have learned this.
A very telling anecdote–sadly, all too familiar–and one that reveals the two prongs of the problem. On the one hand, there is–to use Austin’s language–the dismissal of facts just because they are couched in a logical chain that does not fit one’s priors. A sin of commission. On the other hand, there is the failure to search for facts except in the manner suggested by one’s priors. A sin of omission. A truly empirical science (and I use that term in its broadest sense, to encompass everything from physics to anthropology) simply cannot allow these sins to be a part of common practice. The discipline’s standards must require the scientist to demonstrate not only that her model is accurate under the assumption that the world conforms with her priors, but also, and crucially, that her priors are plausible. Economics has econometrics for the first task, but nothing for the second. Certainly, many responsible economists DO seek to establish the plausibility of the theoretical lens through which they describe economic reality and “test” their theories. But it is not a requirement, and so analyses that are not grounded in this way can and do regularly pass through the filter of peer review into our journals and thereby into the body of “economic knowledge.” This porousness helps to explain how we could possibly have had hundreds of papers written about the macroeconomy, the housing market, and the mortgage and mortgage derivative markets in the five years or so before the crisis hit, with virtually no recognition of the idiosyncrasies of these markets during that time period and the imminent crisis that their particular configuration would bring about. The information was there–and being reported on minutely in the financial press–but not in a form that conformed with economists’ priors. As I argue in the paper that Arjun refers to in his post, it is time for economics as a discipline to start taking its empiricism more seriously. This means learning how to let empirical reality tell us what its meaning is, rather than approach it from entirely the oposite perspective. In this, we have something to learn from the new generation of anthropologists of economics and finance (mentioned in the paper) who are already doing important work in this area.
I’m curious whether Austin has any similar anecdotes about the housing or mortgage or mortgage derivatives markets during the mid-2000s. If so, it would be very instructive to hear about them.
And yet this goes against the very way we learn economics: entire classes focusing exclusively on theory without ever taking a look at any data. We come out of school talking about fictional models. We want to fit the data to conform to them. I had a well respected macro professor tell me that the commodity boom could not explain the dollar increase in oil imports worldwide since countries facing higher prices should substitute away from using oil. I love the field, but it’s hard to bare sometimes.
cottage industries are really needed for a very bad economy since they help boost the local economy.^
Take a look at all of the most recently released blog post at our personal webpage
<img src="http://www.healthmedicinelab.com/concussion-test/ “>
[…] Genre” del profesor Peter Spiegler de la Universidad Masschusetts Boston. Lo descubrimos gracias a este post de Arjun Jayadev en Triple Crisis, donde se resume la esencia del trabajo de Spiegler: “Many critics of the Economics-Made-Fun […]