Friday, January 18, 2013

One Cheer on Accuracy

The Congressional Budget Office released a report which test the accuracy of their forecasts.  As someone who has opened many speeches with the line "Economists have projected 11 of the last 3 recessions" I read it with interest.

So how did the CBO grade itself.   Not surprisingly, pretty well.   As you can see from the chart, their forecasts are pretty much in line with what the Administration (through OMB) and the Blue Chip group (50 economists from across the country who do this kind of thing for a living) project.

Their 2 year forecasts are out of line with what happened on output by 1.4% and on inflation by .8%.   On 5 year forecasts their accuracy is even a bit better - 1.2% for output and .6% for inflation.   That is pretty good.   Both sets of forecasts from CBO have a slight tendency to overestimate wages and interest rates.  

The report also presents estimates of the sources of the errors.   What is amazing about the differences between what the CBO projected and what actually happened in all these categories is that the intensity of the errors is small by any account.

That being said there are two conclusions one should draw from this periodic exercise.   First, just because the CBO and the other major forecasters are pretty good at doing what we ask them to do does not mean that the way their forecasts are used is pretty good.   A whole bunch of economic pundits take these carefully drawn forecasts and misrepresent the results in minor and major ways.    Second, and most importantly, a forecast is just that, not a reality but the best estimate at the time.   Conditions change quickly and frequently in uncertain ways.   So while this report confirms that the CBO does a pretty objective job of projecting what might happen next, forecasts are not certainty.


No comments: