Forecasting & power

“Where does this end?” asks the Economist of Libya. The answer, of course, is that nobody knows.
Which raises an important point. It’s well known that economic forecasts go horribly wrong, at least when they are needed. But could it be that this is not a failure of economics, but rather a reflection of the fact that human affairs are inherently unforecastable?
AFAIK, other social sciences - politics, sociology, anthropology are no better that forecasting than economics. Political science failed to predict the collapse (pdf) of centrally planned economies in 1989, the colour revolutions in the old Soviet bloc in the mid-00s and the Arab uprisings this year. All of which corroborates Philip Tetlock’s finding (pdf) that, in political forecasting “experts only occasionally exceeded chance predictive accuracy.”
There might be a reason why the social sciences don’t get forecasts right. It’s not to do with individual incompetence - though I don‘t deny there‘s that too - but rather with the nature of the disciplines. Quite simply, there are no - or very few - immutable laws of human behaviour. Instead, as Jon Elster has written:

[The social sciences] can isolate tendencies, propensities and mechanisms and show that they have implications for behaviour that are often surprising and counterintuitive. What they are more rarely able to do is state necessary and sufficient conditions under which the various mechanisms are switched on. (Nuts and Bolts for the Social Sciences, p9)

This means that forecasting is often impossible.
Why, then, do people think otherwise? One reason is that the urge to think we live in a predictable world is a powerful one. 
There is, though, another reason. As Elster says, knowing about mechanisms gives us understanding, whereas forecasting gives us control. And this is, of course, what our rulers - politicians, the media  or business “leaders” - want. Social scientists make forecasts not because they know the future, but simply because they are asked; they are simply meeting a demand.
But the thing is that our rulers’ claim to legitimacy is founded in large part upon the notion that they do have, or at least have access to, the ability to predict and control. For example, the government justifies its spending restraint by appealing to a forecast - that without such restraint, bond yields would rise sharply. As Alasdair MacIntyre wrote:

The realm of managerial expertise is one in which what purport to be objectively-grounded claims function in fact as expressions of arbitrary, but disguised, will and preference  (After Virtue, p107)

And herein lies the reason why it’s so convenient to mock economists for their forecasting failures. Doing this distracts us from the fact that human affairs are unforecastable - a fact which challenges our rulers’ claim to legitimacy.

Comments

[Alexandre Delaigue]

In the libyan case, it's not a forecasting problem. It's about defining the political goals of action. "war is uncertain" is a cliché. What is necessary is to know what will you make consider that the goals of war have been achieved - which means, defining the goals of war.
France and UK haven't done that. the UN resolution states that war is necessary to stop violence against civilians : but the only way to make it stop is to see Gadhafi leave. Which is not a stated goal.

[Shuggy]

"they are simply meeting a demand."

This apologia for pontificating economists doesn't quite work for me. Yes people ask them what they think is going to happen - but they wouldn't answer if they didn't think they knew, would they?

[rjw]

yes, but ... some things in the future ARE more likely than others, and we can be more certain about some things than others. I'm pretty certain that 'TEAM A won't win the premier league this year' is a very plausible claim, for some team A. Which leaves a whole host of questions too hard for me to think about right now.

[Alex]

"And herein lies the reason why it's so convenient to mock economists for their forecasting failures. Doing this distracts us from the fact that human affairs are unforecastable"

Um, surely if "human affairs are unforecastable", then those economists making forecasts SHOULD be mocked?

[Glenn]

Hey the future's uncertain but it must be planned for. Therefore we all need to take a view of the future as best we can. This is where forecasts come in.

When we take a loan or mortgage out, we're making a forecast of our future income and ability to pay the mortgage, aren't we? when we take out insurance we're thinking there's a tiny chance we might drop dead or die of some horrible disease.

Everybody forecasts. The big deal if some unforeseen calamitous or fortuitous event happens.

Economists forecasts should be take for what they are - which is a guestimate based on past trends. They are only as good as the data available, and they cannot predict shocks or big events which are game-changing.

For most studies of economic forecasting - 50% of the errors occur due to bad quality data about the past (Sveriges Reichsbank).

So if you moan about economists not knowing about the future - well, they only half know about the present, and some of that will be inaccurate due to bad quality data!