Hostname: page-component-78c5997874-m6dg7 Total loading time: 0 Render date: 2024-11-03T00:20:47.917Z Has data issue: false hasContentIssue false
  • English
  • Français

Section II. Uncertainties in the public sector forecast

Published online by Cambridge University Press:  26 March 2020

Get access Rights & Permissions [Opens in a new window]

Extract

All forecasts are hedged around with uncertainties, and for some time we have been publishing error bands on our forecasts. The National Institute was the first mainstream forecaster to provide an indication of the chances that two key forecast variables, output growth and inflation, would be in particular ranges. The error ranges we used were calculated from past forecast errors and on the assumption (consistent with the data) that these errors were normally distributed. Shortly after we began publishing our probability bands, the Bank of England started to provide its own indication of the error range for its inflation forecasts and, some time later, for its projection of output growth. The Treasury, by contrast, provides no indication of the chance that the Government current account will be in any particular range although this is arguably the most important variable discussed in the Pre-budget Report.

Type
Articles
Information
National Institute Economic Review , Volume 183 , January 2003 , pp. 41 - 42
Copyright
Copyright © 2003 National Institute of Economic and Social Research

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

Footnotes

1

The work for this analysis was undertaken by Ray Barrell and lan Hurst, and a fuller version will be published in the near future.

References

Notes

2 For a number of variables it does publish the mean average percentage errors associated with past forecasts (see Table A8 of the Pre-Budget Report). If the errors are normally distributed, then there is a 60 per cent chance that the outturn will be within ±1 the mean average percentage error. We do find, however, that it is reported in one of the supplementary documents that the mean average percentage error associated with the overall public sector deficit is 1 per cent of GDP. It is not clear to which period this relates but, as it is published in November, we presume it applies to the next financial year. We infer that the standard deviation is 1.25 per cent of GDP and the 80 per cent and 95 per cent error bounds are 1.6 and 2.5 per cent of GDP respectively.

3 We run the model without a solvency feedback constraint for the first five years of our stochastic simulations, and then use one in order that the economy remains solvent.

4 The technique is discussed in Barrell, R., Dury, K. and Hurst, I. (2002), ‘International monetary policy co-ordination: an evaluation using a large econometric model’, Economic Modelling.

5 In ‘The uncertainty of budget projections’ in The Budget and Economic Outlook 2003-2012, Congressional Budget Office, Washington DC, the CBO calculate that four years into the forecast period the standard deviation of the budget forecast is around 3.4 per cent of GDP, and hence is virtually identical to that shown in chart 4.

6 Our current year estimate of a unit standard deviation is based on the assumption that there are two uncertain quarters, and hence may not use all available information for this year as the extra information is not seasonally adjusted.

7 The distribution of the debt stock is asymmetric because the dynamics of accumulation depend on the interest rate, and there is a lower floor on rates but not an upper one.