The economic soothsayers possess no crystal ball; There are not many easily falsifiable theories in macroeconomics, so some prefer the micro side

January 9, 2014 4:40 pm

The economic soothsayers possess no crystal ball

By Samuel Brittan

There are not many easily falsifiable theories in macroeconomics, so some prefer the micro side

It is with some relief that I find my normal column date falls outside the season for new year forecasts. These are either banal or inspired guesses. The advocates of economic soothsaying tend to say that all life involves forecasts – for example, that the ground will not crumble under your feet when you step outside the door. But that is not really the point at issue. What is under discussion is whether theworld economy, or any particular economy, will continue to recover at recent moderate growth rates, fizzle out or soar to unsustainable proportions. My guess (not, I stress, my forecast) is that it will continue to recover, but that the risk of excessive expansion will occur in 2015, which – not entirely a coincidence – is election year in the UK and the run-up to a presidential election year in the US.In saying this I am not contradicting Abraham Lincoln who is purported to have said: “You cannot fool all the people all the time.” For in the same statement he conceded that you could fool some of the people all of the time and all of the people some of the time.

There is a serous problem here. Most philosophers of science trace the scientific status of a theory to its predictive ability, or, in some versions, its resistance to attempts at falsification. How then to distinguish between bogus prophecy and genuinely falsifiable predictions?

Let me try a simple example. To assert that in stated conditions water will boil at 100C is a genuinely falsifiable scientific prediction. To state that these conditions will occur is a more hazardous attempt, easily put at risk by unexpected weather or atmospheric conditions. There are not many – or perhaps any – easily falsifiable theories in macroeconomics, which is why some teachers of the subject prefer the micro side.

Suppose the US upturn fizzles out? Will this be due to too little stimulus, the adverse confidence effects of too much stimulus or to political factors of various kinds? You can be sure that these and many other explanations will compete in the arena.

In the peculiar jargon of social science, many predictions are framed in terms of probability distributions that are said to have “fat tails”. In plain English: they may be badly wrong.

Think of the “great moderation” with which the 21st century began. Forecasters expected it to continue, only changing their minds in 2008 when the “great recession” began. But not everything is gloom and doom. The political leaders of the developed world combined to administer a collective stimulus, which prevented that recession from developing into a depression. Alas, as soon as the immediate danger was over, finance ministers and central bankers fell back on the mantra of balanced budgets and credit restriction, which fortunately they have not been able to achieve.

Nothing that I have said so far is an argument for fatalism. When I first entered economic journalism a favourite topic of discussion was whether there would be a recession in the US next year. It would be unfair to me to say that I neither knew nor cared. But I knew that I did not know; if compelled at gunshot to make a guess, I would look at the frequency of recessions (however defined) in the past and extrapolate from there.

I do not like to end my articles in a nihilistic way. There is an alternative approach to putting too much reliance on highly fallible forecasts. It is what might be called the contingency or toolkit approach. Instead of asking: “What will happen?” we should ask: “What shall we do if?” To turn away from domestic navel-gazing, instead of idly speculating on whether the Strait of Hormuz may be closed to oil tankers, it is more interesting to work out what we might do if they are – “we” in this case being both the western world as a whole and individual countries within it.

Of course, intelligent contingency planning would involve looking at alternative hypotheses on how long the Strait were to be closed, why and to whom, as well as assumptions about oil stocks in different parts of the world. The number of possibilities is endless and only a few can be examined thoroughly, which is one of many reasons why contingency planning will always be as much an art as a science.

At the end of the day, there is a fundamental difference between the toolkit and the prophetic approach to economics – and no doubt many other subjects. The choice between the two is in part, inevitably, a matter of temperament. We need both.

Do not get me wrong. I enjoy a wallow in the prophetic utterances of Spengler, Toynbee, et al – though less so those of Marx, which have been done to death.