We write about GDP forecasts all the time, but we realised that nobody has really been keeping track of those forecasts. So we decided to do it.
Looking at the forecasts over the years, a few things become apparent. Not surprisingly, predictions closer to the end of the year are narrower and closer to the actual GDP growth rate.
What is perhaps more surprising is how far off the mark early estimates can be. On average, estimates for the current year's real GDP growth that were made in the last quarter of the previous year were 3.3 percentage points off of the actual growth rate. The median mishit was 2.4 percentage points. Given that real GDP growth from 2012 to 2014 has been less than 4.5 per cent annually, that margin of error is huge.
Why are those forecasts so poor? It could be that these are simply volatile economic times, such that conditions at the start of the year can change dramatically by its end. The other factor is that we are looking at real GDP estimates, which require that forecasters predict not just economic growth, but also the rate of inflation. It's hard enough to peer into the future for one variable, much less do it for two.
The takeaway is to take early GDP growth forecasts with a ladleful of salt.