Economic Policy

Forecasting failure

  • Published: Oct 16, 2012
  • Author: Tom MacInnes
  • Category: Economic Policy

The Office for Budget Responsibility today published their “Forecast Evaluation Report” – an explanation of why their predictions for economic growth have been wrong. It’s fascinating, salutary reading, and it is admirably open about the difficulties of making economic forecasts. 

The top line is that the OBR made very accurate predictions of public sector net borrowing, and much less accurate predictions about the overall state of the economy. They write

“In our June 2010 forecast we over-predicted PSNB in 2010-11 by £7.4 billion, a relatively small error even at that short time horizon. We under-predicted revenues by £1.2 billion and over-predicted spending by £6.2 billion”

It's worth noting in this extract that their main error here was due to departments underspending. On the economy as a whole, they say

“Following the Coalition’s first Budget in June 2010 we forecast that the recovery would be slower than its predecessors, but nowhere near as slow as it has been. We forecast that GDP would rise by 5.7 per cent from the first quarter of 2010 to the second quarter of 2012, but the latest data suggest it has grown by only 0.9 per cent.”

This is a pretty big difference. The OBR thought the economy would grow steadily every quarter from June 2010. In fact, the economy grew much slower than predicted in the second half of 2010 and shrank in 2011, as their graph shows. 

obr_graph.pngEven in March 2011, the OBR was predicting strong growth, before revising the projections down sharply last autumn. The gap between their estimates in March and November of last year accounts for the majority of the difference between their 2010 prediction and the actual figures for 2012. The predictions from last November, which were far more pessimistic than earlier estimates, have stuck reasonably closely to the actual figures. The OBR offer a range of reasons why growth may have been slower than they anticipated – errors in the GDP statistics, high inflation leading to low consumption, for instance. They also suggest that the level of “fiscal consolidation” – cuts, broadly – may have itself had a more negative effect on growth more than they expected it to. This has long been the view held by the opponents of austerity ( eg the TUC), and a view to which the IMF appear to have come round.The OBR is independent of the Treasury, and it is this independence which gives its predictions such profile and status. But being independent is not the same as being value-free. Given that the historical evidence is not unequivocal, the assumption that cutting spending will have a limited effect on growth is as much a political one as an economic one. It is only when the predictions are reassessed retrospectively that such assumptions become clear. 

We would also point to the huge private sector surplus, which shows how unwilling companies are to invest. There is no mention of this surplus in the OBR report. This is also indicative of a particular worldview, namely that there is only one deficit that matters, and that’s the government’s.

There are two points to make here. Firstly, this is not an accusation of bias – we are all biased. In our work, we present the figures honestly and accurately but behind that there’s a whole set of values and beliefs around what we think matters. The OBR’s forecasts may be independent in origin but they are to be used politically so it is vital to understand their assumptions.

Secondly, projections are always a best guess. The OBR themselves say

“We have been at pains to point out that there is enormous uncertainty around any economic forecast and that policymakers and others need to recognise this when taking decisions based on them”

Obviously that’s correct, and, again, their honesty and transparency is laudable. But the OBR are not some small think tank on the edges of the discussion –they are hugely important and their statements carry real weight. Either the predictions have to get a lot better, or they should be treated as best guesses, like so many other economic projections. 

blog comments powered by Disqus