Simple model mania

from Peter Radford

Economics, as we all know, is a bit of a shambles. There’s plenty of good stuff sitting about, and there’s plenty of bad stuff. On any given day we can witness two Nobel prize winners disagree and put forward totally contradictory theories to support radically different sets of policy advice. One example at the moment is the continual warnings we gat about the imminence of high inflation. One side says we are doomed because all the money the Fed has been tolerating in the economy will cause prices to rise – this is a variant of the old “too much money chasing too few goods” argument. The other side says this won’t happen because the money is all idle in bank reserves and will stay there as long as the economy sputters and interest rates stay stuck at or near zero – no one wants to do anything with all that money, so it is irrelevant and certainly won’t end up chasing too few goods. So far the evidence supports the latter, but the former don’t stop, and they don’t revise their ideas either. The two sides just shout at each other. The public has a right to be bewildered. 

Milton Friedman has always been associated with the “too much money chasing too few goods” idea. He is well known for saying that inflation is always and everywhere a monetary phenomenon. But that’s because he wanted it to be. It fitted his view of the world.

One of the reasons economics is a shambles is that economists have an obsession with simple models. This is not new. It goes all the way back to the subjects roots. Quesnay’s obsession was a model that showed the flows of activity through three sectors of the economy. As if everything could be reduced to just three such sectors, and as if each sector was predictable, stable, and no one went off the script prescribed by Quesnay. This was in the mid 1700′s just before Adam Smith inadvertently shoved economics lurching off into deep unreality by mentioning – just once mind you – something about the effects of an invisible hand.

This is s brilliant metaphor. It is an awful basis for a model. Ever since he wrote about it economist have been obsessing over it. They see order everywhere on the surface and disorder everywhere beneath. They ask: how come? And, as they say the rest is history. What’s worse is that Smith made this remark right as the real sciences – let’s not dignify economics with that moniker – were careening off into a deterministic, mechanical, and ultra certain cul-de-sac. Well not really a true cul-de-sac. Newtonian physics remains very useful for some explanations and predictions. It just isn’t complete and it needed gentle correction, which is what Einstein did.

The problem is that economics never had its Einstein. It has stayed stuck in the past. It works relentlessly on adding ever more epicycles to its rational, reductionist, and ultra-certain view in order to accommodate the inconvenience of empirical contradiction. It stays simple. Very simple. This is so it can break things apart and study the pieces. Then it hopes that the bits just modeled behave nicely when they are thrown back into the pool.

Which they rarely do. The world is vastly complex and not amenable to simple modeling. Things may mimic the model for a while, but sooner or later the relationship changes and a new simple model is called for. There aren’t many, if any, truly lasting phenomena that economics explains well. And it can’t predict for toffee. Not that prediction is the be all and end all. It would nice, however, to get things right more consistently.

Let’s get back to Milton Friedman.

He has an undeserved reputation. He was very good at self promotion, and managed to have the good fortune to possess a simple model to present as an alternative to the Keynesian view losing its gloss in the late 1970′s. He was in the right place at the right time.

One of his ideas, the one for which he won the Nobel prize, is the Permanent Income Hypothesis. In a nutshell the PIH says that consumers make decisions about how much to spend or save based upon their estimate of their lifetime income, and not their current income. This is because people generally earn more later in life, and, knowing this, are willing to overspend – i.e. borrow – when they are young, and underspend – i.e. save or pay down debt – when they are older. This is important because it helps explain why consumption patterns are less volatile than incomes. People keep spending despite a drop in temporary income.

An alternative idea to explain consumer habits is the Relative Income Hypothesis first put forward by James Dusenberry back in the late 1940′s. It has a distinctly Veblen like flavor. It says that consumers are influenced, not by their lifetime incomes, but by what others are spending. People, it argues, want to keep up with the Jones’s.

Now. Friedman had to reject RIH. That’s why he invented PIH.

Why did he have to?

Because one implication of RIH is that consumers are influenced by social and institutional factors. This is the Veblen flavor coming through. It implies that consumers are not behaving purely along rational lines, and, worse, that economics has to take into account social and institutional stuff Friedman and his followers want to chuck out. This is so offensive to the simple world view Friedman built all his thinking up from that it had to be rejected and replaced with a hypothesis in line with that simple world view.

In my view neither PIH nor RIH is correct in and of itself. There are decisions, home buying for instance, where PIH looks pretty good. There are other’s, fashion and iPods, where RIH seems to come into play. The point being that the obsession to reduce everything to simple ideas that can be put into simple models, condemns economics to misunderstand, and therefore not explain, reality. People are far too nuanced and unpredictable to be shoved completely into either PIH or RIH. They bounce about. The truth is that there are likely plenty more explanations as well. Perhaps there are twenty hypotheses each explaining a little bit. But, since that makes modeling complicated, economists stay well away from reality.

So scared are economists of abandoning their simple views and simple model mania that they would, apparently, prefer to become irrelevant rather than change.

One last example, one that might shock. Growth.

We live at the latter end of an extraordinary event. Humanity has just gone through an unprecedented period of growth. Over two hundred years and counting. In the context of our entire history this is so abnormal as to scream for an explanation. Average incomes and wealth did not change much for centuries. Thousands of years of near subsistence for the vast majority of our ancestors, and a not much better life even for the kings, queens and sundry elites. At least by our standards. Then a sudden rush to wealth. Why?

Given this outrageous anomaly you would think economics would have a water tight explanation for growth. It doesn’t. It really doesn’t. Textbook economics cannot explain, from within its set of ideas, how that rush to wealth came about. It cannot explain the Industrial Revolution. It cannot explain why it began in England. It cannot explain its consequences. And it is hopeless at explaining whether growth is the new normal. Robert Solow had a shot at it in 1957. He wrote a paper that is regarded as ‘seminal’ by economists. It is relatively straightforward in its thinking, but better at its math. So of course it is famous. The problem with Solow’s theory is that it fails to explain well over half of all growth. About two thirds of growth seems to be caused by something that falls outside of the basic parameters of economics. It just sits there as a residual. Solow’s Residual. Yes, this is the economy we are talking about.

Let’s not be too harsh. Solow didn’t think he would get an explanation, he wanted to highlight to extent of the impact of technology and other stuff that sat outside economic theory at the time. Plus people like Romer and Lucas have embellished and expanded on Solow. Romer, for instance, had the temerity to make technological innovation an economic factor inside the economy and not something that dropped from the skies onto it. Clever that. Whoever would have thought that businesses innovate to try to make a buck? Certainly not economists. Well, not all anyway. And not the sort that write the basic textbooks.

Keeping technology outside and not inside the economy is another example of what happens when you allow your simple view and your desire to keep your models simple, drive your thinking.

By the way, that 50% to 60% of all growth that Solow couldn’t explain is still a bit of a mystery. But now it has a clever sounding name: Total Factor Productivity.

Here’s what I think: until we drop the obsession with simplicity, and until we embrace the complexity of reality we won’t resolve this conundrum. By that I mean that the causes of Total Factor Productivity are just as much social, institutional, cultural, and political as they are economic. That is to say that economics needs to change to incorporate those things alongside complex ideas, or it needs to admit it cannot explain the economy.

Then again much of economic theory has little to do with the economy and a lot to do with itself. So, maybe, we are stuck with simple model mania.

That’s embarrassing. We have to ask sociologists, political theorists, and historians what’s going on in the economy?

What are economists for? Wall Street model building?

Simple