Comparing economics with physics… physics is a science and economics isn’t

I know am late on this but still people who have not read the speech should take a look. RBI Gov Subbarao reflects on the crisis and issues with economics. He says economics was seen as science before the crisis which was wrong. It is a social science and will remain one.

This was a particularly hard landing for the profession. The years before the crisis, in fact, saw economics as a subject gain impressively in clout and popularity. The price stability and macroeconomic stability that prevailed over an extended period – the Great Moderation – enhanced the standing of economics and gave economists an enviable halo; the increasing sophistication of financial markets where risk could seemingly be measured with precision of upto five decimal points gave economics the clout of prophesy; and the way economists were able to raise obscure questions such as why drug dealers continue to live with their mothers, what school teachers and sumo wrestlers have in common, and answered those questions with impressive insights, awed common folk. Economists were being sought out to pronounce on an ever growing number of issues and their opinions were being heard with regard and trust. And then the financial crisis came and crashed all this.

The sharp reversal in fortunes raises two questions: what went wrong and what can be done about it. Both questions are complex; the first because it has too many answers and the second because it has too few. What I propose to do in this first part of this lecture is to address these two questions.

He compares economics to Physics

Let me turn to the first question. What went wrong with economics? I am not an economist, only a practitioner of economics. I will, therefore, restrict myself to just the bigger maladies that, I think, afflict the discipline.

With the benefit of hindsight and by wide agreement, it now seems that by far the most egregious fault of economics, one that led it astray, has been to project it like an exact science. The charge is that economists suffered from ‘physics envy’ which led them to formulate elegant theories and models – using sophisticated mathematics with impressive quantitative finesse –  deluding themselves and the world at large that their models have more exactitude than they actually did.

Is Economics like Physics?

As I started thinking about this charge, I realized that there are indeed quite a few parallels between economics and physics. Let me explore these parallels briefly.

The theory of rational expectations says that wages and prices adjust instantaneously to new conditions because of perfect information just as Newtonian physics says that the gravitational configuration of the universe will change instantaneously in response to any infinitesimal change in the system, an inference that Einstein found troubling because it conflicted with his special theory of relativity.

 The centerpiece of Keynes’ theory is the existence of inescapable uncertainty about the future which implies that risk cannot be measured precisely beyond a point, and that taking uncertainty seriously has profound implications for how one applies economics. Look at the parallel in physics. The foundation of quantum mechanics is Heisenberg’s Uncertainty Principle which puts an irreducible limit on our ability to simultaneously determine the position and momentum of a particle.

Physicists know of ‘singularities’, or black holes if you will, where the laws of physics break down. In economics, the analogy would be Depression Economics. There is currently a fierce debate, especially in the US, about the quantum of fiscal stimulus and the timing of its withdrawal. Some economists, notably Krugman, have argued that the size of the stimulus should be much larger than what the models suggest simply because ‘in Depression Economics, the usual laws of economics do not apply’.

Being a student of IIT, he would know these similarities better. He then points to differences:

Striking as these comparisons are, I am sure, you have noticed an obvious flaw in this line of thinking. Similarity in a few laws does not mean similarity in the basic nature of the academic discipline. The fundamental difference between physics and economics is that physics deals with the physical universe which is governed by immutable laws, beyond the pale of human behaviour. Economics, in contrast, is a social science whose laws are influenced by human behaviour. Simply put, I cannot change the mass of an electron no matter how I behave but I can change the price of a derivative by my behaviour.

The laws of physics are universal in space and time. The laws of economics are very much a function of the context. Going back to the earlier example, the mass of an electron does not change whether we are in the world of Newton or of Einstein. But in the world of economics, how firms, households and governments behave is altered by the reigning economic ideology of the time. To give another example, there is nothing absolute, for example, about savings being equal to investment or supply equaling demand as maintained by classical economics but there is something absolute about energy lost being equal to energy gained as enunciated by classical physics.

In natural sciences, progress is a two way street. It can run from empirical findings to theory or the other way round. The famous Michelson-Morley experiment that found that the velocity of light is constant led to the theory of relativity – an example of progression from practice to theory. In the reverse direction, the ferocious search now under way for the Higgs Boson – the God particle – which has been predicted by quantum theory is an example of traversing from theory to practice. In economics, on the other hand, where the human dimension is paramount, the progression has necessarily to be one way, from empirical finding to theory. There is a joke that if something works in practice, economists run to see if it works in theory. Actually, I don’t see the joke; that is indeed the way it should be.

Karl Popper, by far the most influential philosopher of science of the twentieth century, propounded that a good theory is one that gives rise to falsifiable hypotheses. By this measure, Einstein’s General Theory was a good theory as it led to the hypothesis about the curvature of space under the force of gravity which indeed was verified by scientists from observations made during a solar eclipse from the West African islands of Sao Tome and Principe. Economics on the other hand cannot stand the scrutiny of the falsifiable hypothesis test since empirical results in economics are a function of the context.

The short point is that economics cannot lay claim to the immutability, universality, precision and exactitude of physics. Economics is a social science and its predictive power is at a fundamental level influenced by human behaviour and actions.

Fantastic stuff.

He then points to problems with Fama’s efficient market hypothesis and weaknesses in current economic models.

Moving on with what went wrong with economics, another flaw, actually one related to ‘physics envy’, is the obsession of economists with models so much so that they convinced themselves that if something cannot be modeled, it is not fit enough for academic pursuit. Indeed, with the benefit of hindsight, it is now possible to see that one of the basic causes of the crisis was that the models used by central banks, such as even the sophisticated Dynamic Stochastic General Equilibrium (DSGE) models, remained confined to the real sectors of the economy and did not capture the complexities of the financial markets. It is not surprising that economists missed seeing the crisis brewing in the underbelly of the financial sector.

Yes, it is possible to construct beautifully precise models but only if you assume that rational economic agents with perfect information are operating in free markets that always return to equilibrium. But none of these assumptions holds true in the real world; models of economists are mere abstractions of reality that are useful for understanding but woefully inadequate for prediction. That is why good economists are those who superimpose judgement on the predictions thrown up by models, a subject to which I will return in the second part of my lecture.

Of all the economic theories that came under attack after the crisis, the one that got the most grilling was the efficient market hypothesis put forward by Eugene Fama of the Chicago School. Its central tenet is that the price of a financial product captures all available information about it. The efficient market hypothesis did away with the unrealistic assumption of perfect information but it assumed perfect information about risk. An obvious inference of the theory is that risk is perfectly measureable, and if it can be measured perfectly it makes eminent sense to use that measurement in economic decisions.

Not surprisingly, the efficient market hypothesis spurred furious model building based on the assumption that the distribution of risk is captured by the Gaussian bell curve. The spectacularly, albeit briefly, successful Black-Scholes model for option pricing too was based on the normal distribution of risk and ignored the possibility of extreme events.

What can be done to improve things?

 The first thing is for economics to give up the pretence of being an exact science and striving for false precision. While there is value to models for furthering understanding of economic phenomena, economists should, however, be sensitive to the limitations of their models and use judgment in interpreting model results. And we clearly need to get back to emphasizing the importance of economists ‘getting their hands dirty’ with empirical work.

Then there is the crucial aspect of economic history. Much of economic thinking has been handicapped by economists not having a sense of economic history. ………If only training in economics had included a study of economic history, perhaps we could have avoided repeating history, never mind as a farce or as a tragedy.

Finally, economics, perhaps more than other social sciences, has suffered from ‘group think’. Group think is best illustrated by the simulated game of a beauty contest where prizes are awarded if your choice matches the aggregate choice of the group. Under this rule, you get rewarded not for original thinking but for mastering the art of thinking like others.

Excellent stuff from Subba.

This point about science and social science was made in an earlier post as well.

Advertisement