I wonder how many of them pulled numbers out of their behind and how many have done proper forecasting. I bet the answer would be embarrassing. The interpretation of the journalist also doesn't help because it hides the diversity of opinions.
For example, go here:
http://hosted.ap.org/specials/interactives/_business/ap_economic_survey/index.html
Click broader economy and then in "Choose and economist" pick "view all responses". There's even a 5% "prediction" for the last quarter of 2010 made in January 2011 and not a lot of clustering around the mean.
I think that the Fed's Summary of Economic Projections which should soon be out with the upcoming FOMC meeting will be a more serious base to discuss this, compared to this particular survey.
Originally posted by KazetNagorraWell, they weren't asked for posteriors (or intervals) but for a point forecast. The press wants quantitative point forecasts and qualitative description of posteriors (uncertain environment=high posterior variance, risk of double dip-skewed distribution with a fat tail, etc.).
These predictions precise to a tenth of a percent always amuse me. Not a snowball's chance in hell that economists can predict fourth quarter growth this accurately.
Originally posted by PalynkaAt his press conference, Bernanke unveiled projections on inflation, growth and unemployment which were all within the ranges of what the survey gave.
I wonder how many of them pulled numbers out of their behind and how many have done proper forecasting. I bet the answer would be embarrassing. The interpretation of the journalist also doesn't help because it hides the diversity of opinions.
For example, go here:
http://hosted.ap.org/specials/interactives/_business/ap_economic_survey/index.html
Click b ...[text shortened]... eeting will be a more serious base to discuss this, compared to this particular survey.
Originally posted by KazetNagorraPredictions are always +/-; you have a range in mind, so you predict something in the middle of that range. If you're close, that's tantamount to being "right."
These predictions precise to a tenth of a percent always amuse me. Not a snowball's chance in hell that economists can predict fourth quarter growth this accurately.
That's true whether you're predicting economic growth, the weather or number of wins a baseball team is going to achieve this year.
Originally posted by sh76The range is actually of great importance.
Predictions are always +/-; you have a range in mind, so you predict something in the middle of that range. If you're close, that's tantamount to being "right."
That's true whether you're predicting economic growth, the weather or number of wins a baseball team is going to achieve this year.