Economic Errors
That the discipline called economics is in a bit of a mess is beyond dispute. That is has taken this long for many of its leading lights to come down from their respective mountain tops and admit so is, frankly, a scandal, but not surprising given the enormous prestige and ego reinforcing power attaching to being an nationally renowned economist. The top players represent the apex of a deep and loyal band determined to defend their position stoutly in the face of periodic insurrection in the ranks. That way they can ensure the perks flow to the loyal troops. This much we all know, and our knowledge simply tells us that economics is much like any other trade. Insular, defensive, conservative, and self-regarding.
Those untethered to the need to be loyal to orthodoxy have the privilege and opportunity both to make fun of the naked emperor and to call for changes. Perhaps a new set of clothes is in order.
For fun we should all set our minds to deciding the key moments when the subject left its earthly orbit and began its steady trajectory towards the outer reaches of human invention to take its place alongside alchemy and astrology as one of our most intricate, well developed, and yet futile endeavors.
I know you all have suggestions.
The list is vast.
Was it the Walrasian impulse to describe an economy as a massive machine inevitably arriving at a nirvana like steady state? Was it earlier with Smith’s metaphorical invisible hand? Was it the so-called marginal revolution, when the classical economists sought to redefine the subject as cleansed of human groups and tensions? Was it Marx, who disrupted us all by pointing to those groups and tensions?
The list is, indeed, vast.
My own favorite, which I acknowledge is idiosyncratic, is the invention of revealed preferences.
For those of you blissfully unaware of what this is allow me to give a very restricted description.
Since economics is concerned with, or used to be, the way in which people exchange, buy, and sell stuff, it behooves us all to identify why they may be doing these things. The answer seems to be that they are trying to satisfy a set of needs. These needs may be very basic – as in survival. Or they may be very ornate – as in buying luxury goods. So economists need to find a way to describe this diverse impulse to acquire and to link it with a set of requirements. That set of requirements is called a set of “preferences”. And preferences are thought to express “utility” because they match the things we acquire against our needs. Phew. But that idea of utility turns out to be a slippery slope into mushy thinking. What do we mean by utility? More to the point: how do we measure it? Since we are trying to understand the machinery that matches our needs against our ability to satisfy those needs, we need to be able to measure and deploy utility as a key part of that machinery. The debates about how to do this lingered long in the early years of economics. The originator of the notion of utility, at least as known in economics, was Bentham, but he left the concept informal. And, as we know, economists after about 1870 wanted to describe everything in more formal or mathematical terms. So the Benthamite explanation wouldn’t do. His version of utility was far to vague, psychological, and intractable to mathematical manipulation.
Since economics was all about mathematics that had to change.
Enter Frank Ramsey and Paul Samuelson.
They liberated economics from the deeper psychological question of why people acquire what they do. That is where the intractability stems from. They realized also that the future of economics was mathematical modeling and thus required logical rigor. Internally consistent math was more important than understanding why people did what they did. Economists were more interested in the consequences of exchange than why exchange took place in the first place. So they needed to rid themselves of reasons which might be whimsical or prone to sweeps of fashion, and instead assume that what people did was based upon a rational process. We needed to strip out what ought, might, or could have happened, and replace it with only what did happen. Economics had to be positive not normative.
So a sleight of hand was needed.
Rather than dwell on the tangled roots of why people acquired stuff, economics contented itself with assuming that whatever they did they must have wanted to do. That is to say that the murkiness of actual preferences is too unstable and messy to model, but if we assume that what we observe people doing is a manifestation of what they wanted to do all along, then we can model that more easily.
Thus when you buy blue socks you are said to reveal your preference for blue socks.
It gets more complicated and jargon infested.
You are assumed to prefer blue socks at all times over those red socks you set out to buy but couldn’t because they were sold out. You see the reason you bought them is a messy subject. That you indisputably did buy those blue socks is not. We can model the latter and not the former. So we brush aside the complicated questions and move on swiftly secure in the knowledge we now know what your preferences are. After all you revealed them through your actions.
Now as a representation of human behavior this sucks. But it has the great advantage of freeing economics from psychology, which we all know is a pseudo science anyway. I am kidding. Unfortunately orthodox economists aren’t. Their revealed preference is for massive simplification in order to preserve their ability to model what they perceive to be more important: the machinery of the marketplace. No one denies that what goes on inside people’s heads is unimportant. It is just that were economics to accommodate that nearly infinite set of causes for action it would never be able to build those neat models it so values.
So we simplify.
So we eradicate humanity from the market system.
The hope within all this is that the outcomes of the now dehumanized models will be a good representation of the effects of what real people do in their awful and messy way. Economists are proud of calling these rough relationships “good first approximations”. That is they stop short of getting their hands dirty in behavioral details because they have convinced themselves they can arrive close enough to the right answer without so doing.
Newtonian physics is also a good approximation. So good that it stood the test of experiment for centuries until advances in its details required a more sophisticated approximation. That is what relativity and then quantum physics provided us. They are both more accurate and better descriptions of reality than that of Newton.
Unfortunately economics is stuck firmly in its Newtonian phase content with its rough approximations. One reason is the enormous success of the notion of revealed preferences and its removal of human life from the models subsequently produced. As more than one critic has suggested, orthodox economics has morphed into a study of cyborg activity rather than human activity.
As we debate ways to get economics back from its current unearthly orbit we need to realize that while the classical and neoclassical systems give us beautiful and in many ways excellent first approximations, they remain just that: approximations. We should not be content as being “approximators”. Especially when newer methods and insights are littered about us in other disciplines, many of which have moved great distances away from the machinery metaphor towards more complex and yet just as disciplined other metaphors.
If, as McCloskey argues, rhetoric matters, and I am a story teller by nature so I agree with her, then the choice of metaphor is crucial. Economics is still using an old and quite poor one. It needs to move on.
One reason it doesn’t is the power of revealed preferences to both eliminate the complications of human causation and to simplify the logic underlying the math of the models that produce those first approximations.
Approximations which, by the way and judging by the recent crisis, seem to be getting less useful by the day.
Hence the choice of real economics as the description for the new way is not a slur against the classics. It is simply a claim that progress, and the refining of those approximations, must now accommodate more real expressions of behavior. There are many ways to do this, and we cannot predict which will succeed. But what we do know is that parts of the machinery like revealed preferences do not help. They hinder.
That’s why I nominate them as my example of economic error.
What’s your nomination?
