Ex-post theorizing is a tad easier than forecasting

by Russ Roberts on February 5, 2009

in Complexity and Emergence

A paraphrase of a quote from an economist (can't remember which one) heard at a recent conference:

Everyone can explain why the dollar rose or fell yesterday. No one can predict what it will do tomorrow.

Says it all.

View Comments Share Print Email

{ 12 comments }

1 Lee Kelly February 5, 2009 at 6:57 pm: Initial conditions. Even if all your economic theories are true, it is difficult to predict many important events in an economy. By the time you have collated enough information to get a decent picture of a situation (the initial conditions from which to derive predictions), the situation has passed and its consequences have already occurred.

That's just an unfortunate problem of trying to forecast large and complex systems. Backcasting is always easier, even when your theories are all true.
2 Ari February 5, 2009 at 7:13 pm: Reminds me of something I saw on Slashdot a while ago:

"In economics, the exam questions are the same every year. They just change the answers."

Economics is not only a dismal science, but often a dead wrong one.
3 Greg Ransom February 5, 2009 at 8:08 pm: Russ, you keep illustrating the confusion of economists when it comes to explanation and science.

A biologist cannot predict the precise course which adaptation and the origin of species will take over a "biological instance" — BUT a biologist CAN prospectively anticipate the GENERAL course of adaptation and speciation, and in the same way they can explain it after the fact.

Good explanation and good science doesn't have to be like simple 2 body Newtonian physics — even much of HARD science isn't like this (see for example the book _Chaos_ on non-linear phenomena and explanation in the hard sciences.)

Hayek has several important essays on this fact of explanation.

"Explanation of the principle" is a KEY part of good science — its at the core of Darwinian biology for example.

See, for example, F. A. Hayek, "Degrees of Explanation" and F. A. Hayek, "Scientism and the Study of Society" and F. A. Hayek, "The Theory of Complex Phenomena".

Understanding the nature of good explanations and good science is crucial to the advance of economics science — it's telling that economists haven't gone beyond kindergarten myths about science and good explanation.

Russ wrote:

"Everyone can explain why the dollar rose or fell yesterday. No one can predict what it will do tomorrow."
4 Martin Brock February 5, 2009 at 8:36 pm: It's called "market efficiency". I can't predict what the dollar will do tomorrow, because I don't know what will happen tomorrow that traders haven't incorporated into today's price. If I could know, other traders would know too, and today's price would already be different.
5 Martin Brock February 5, 2009 at 8:39 pm: Actually, market efficiency is one the most interesting economic principles I know. It only tells us that it can't tell us anything, but it tells us so very elegantly.
6 muirgeomuirgeo February 5, 2009 at 8:56 pm: FROM OCT 25 2006

Der Spiegel

America and the Dollar Illusion
By Gabor Steingart

The dollar is still the world's reserve currency, even though it hasn't deserved this status for a long time. The devaluation of the dollar can't be stopped — it can only be deferred. The result could be a world economic crisis.

FROM OCT 25 2006
7 Gil February 5, 2009 at 10:45 pm: Economics was invented to make meteorology look good. (Boom. Boom.)
8 Ray G February 5, 2009 at 11:12 pm: The crux of the matter is that economics is what it is precisely because no one can predict the results of so many disparate self-interests.

When I traded currencies, a major source of amusement was watching the online headlines change by the minute after a big announcement.

An economic announcement would come out, the charts would go wild for a few minutes and we'd do our thing (or not) and then Bloomberg et al would have an explanation instantly ready. Within a minute or two they'd have another reason, often times completely opposite of the first one given. A few more minutes and yet another reason would emerge. The bigger the move caused by the news, the more the reasons would change.
9 Hammer February 6, 2009 at 9:11 am: I think the purpose of macro-economics is to remind us that though we can know fairly well what will happen when we know all the initial conditions, knowing those is nearly impossible due to the scale and interconnectivity of the system.

The purpose of micro is to limit the scale and connections such that we can try to predict things.

I think the definition of economics as "applied philosophy" is a pretty good one.
10 Greg Ransom February 6, 2009 at 11:50 am: Let's say it again.

The 3 body problem doesn't make physics bad science ..
11 Kevin February 6, 2009 at 1:44 pm: Actually, you can't explain the past either. People weave narratives that fit, but they don't and can't know what the process has been.
12 Mark February 6, 2009 at 2:47 pm: "a biologist CAN prospectively anticipate the GENERAL course of adaptation and speciation,"

You mean SWAG according to fertile imagination?