Monday, February 14, 2011

Precise Forecasting - Who Actually Engages In It?

I'm constantly pricked to read the writings of those who would not agree with what I think or do.  Austrians, or Hayekians, or whatever you wish to call them, are one such category of people.  I think it's got to be a good thing to read what people who disagree with you have to say.

This is an interesting article on the 10 Hayekian Insights for Trying Economic Times by Bruce Caldwell, not least because it mostly talks common sense.  I often find this about economists who devote their allegiance to Hayek - I find it hard to believe he is really the only proponent of so many of these ideas.  But I haven't got any decent grounding in economic history to really make this case.

The case I will make relates to point 5 about precise forecasts being ruled out because the economy is complicated.  Duh.  This is the entire basis upon which econometrics functions: Economic variables are modelled as random variables, and hence we attach probabilities, not certainties, to possible outcomes.  That's why at the basic level we have 95% confidence intervals, why we attach confidence bands to our forecasts and are always looking to estimate standard error bounds for anything we plot.

If some foolish economist decides to make a precise forecast, then he is just that.  But I strongly suspect Caldwell means any economist who says that their forecast for growth is 0.5%.  This, to all too many Austrian-type libertarian economists, is too much.  Yet if 0.5% is the most likely figure for growth that the economst thinks, well sound bites are going to come out saying it's 0.5% that is the forecast.  But every forecast has a confidence bound, if only you bother to read the paper or article it comes from...


Post a Comment

<< Home