Skip to content

Very cute

Yet economists say there is a good chance the IMF’s numbers are already out of date.

“The problem with the IMF forecasts is that they tend to be quite out of date when they come out,” says Andrew Goodwin, chief UK economist of Oxford Economics.

“If you think about forecasters like ourselves, we’re constantly updating our forecasts based on the data that’s coming out. The IMF generally has to draw a line quite a long way before the time they’re released and take the data as it was then.”

Well, yes. As an argument not to take the free stuff from government but buy our private sector produce instead that works very well. The problem here is that macroeconomic forecasts are near always total toss anyway, making private and later no more use than public and earlier.

10 thoughts on “Very cute”

  1. Michael van der Riet

    But somehow the government forecast known as the Stern Review was not a total loss; indeed it was and, despite being even more out of date than anything from the IMF, is solid gold.

  2. Riffing on Milton Friedman I wonder if any economist has ever sold his research to someone spending his own money.

  3. Is the IMF any worse than the OBR? Perhaps the rule should be that all offspring of George Osborne should be destroyed

  4. Guido has a good chart comparing IMF forecasts to what a really happened

    They are consistently wrong and invariably pessimistic

    It’s almost as if the anti-brexit cabal there was deliberately putting out misinformation

    In their defence they are equally useless at predicting the performance of most other countries

    It does make you wonder why anyone takes any notice

  5. People have a fundamental need to control things, or at least think that things can be controlled. People fear the unknown, so any prediction with any credibility (real, slight, vague, spurious or Neil Ferguson) will be seized on because the alternative is to say we just don’t know.

  6. Bloke in North Dorset


    Neil Ferguson as a unit of measure, I like it:

    “That forecast was out by a couple of Fergusons” gets the meaning across splendidly.

  7. @BiND

    Ah, a Ferguson is a larger measure of wrongness than a Cable (12 out of the last 2 recessions level of wrong) or an Abbott (where numbers exist in a Schrodinger’s cat-like state, wrong in all 11 dimensions, but only specifically wrong when uttered).

  8. I’m not sure about the “Ferguson” as a measure of forecast inaccuracy: I think that, like the Farad, it is so large compared to the range of ordinary error that we will never see “a Ferguson” – instead, we will be dealing with µFg and maybe (for egregious errors bordering on rank incompetence) we may see the rare mFg in real life.
    Even at that, it’s a useful tool.

  9. @dcardno

    I think you are correct – we need to calibrate the Ferguson against something us mere mortals might guess wrong (could be a picoFerguson), then scale up until we get to disaster levels 🙂

Leave a Reply

Your email address will not be published. Required fields are marked *