This forecastingerror was at least understandable since the dotcom boom was an exceptional episode, but the chancellor then started to bank on this buoyancy continuing.
Of course, the financial crisis explains this massive forecastingerror, but the study finds that relatively little was due to fiscal stimulus programmes in response to the downturn.
Many thanks to Mike Cowan (author of the Cowan LRA Forecasting Model) for pointing out an error in a prior version of this post, and for providing the numbers behind the chart.