« Musings on the Trade Release and Consumption Theory | Main | Trusting the birth/death model »

November 16, 2007

Assessing CBO Projections

CBO releases CBO's Economic Forecasting Record: 2007 Update [pdf].

See in particular pages 11-12, for a comparison against the administration and Blue Chip forecasts, in terms of mean error, mean absolute error, and RMSE.

Technorati Tags: , .

Posted by Menzie Chinn at November 16, 2007 05:24 AM

digg this | reddit

Trackback Pings

Comments

Figure 2 is pretty interesting. You can totally see the influence of HMOs in the 1990s.

Posted by: Buzzcut at November 16, 2007 06:15 AM

Menzie,

Is there any analysis of their initial unadjusted forecasts compared to their adjusted forecasts.

I have a friend who has developed a metric to predict the CPE 6 months in advance using gold and his results have been much better than the CBO and that is unadjusted numbers 6 months in advance.

Looking at the numbers reported it appears that their accruacy is worse than 10%. Do they report their error rate as a percentage? Did I just overlook it?

Posted by: Anonymous at November 19, 2007 05:13 AM