In the pre-Budget forecast, we emphasised that the use of market expectations to derive the path of interest rates was potentially inconsistent. It may have biased the economic forecast upwards, with consequential effects on the fiscal forecasts. That inconsistency might have been present in both short and long-term interest rates.

It is very difficult to calculate the possible degree of bias, but a thought experiment might provide a guide to its upper limit. It would be normal to expect some monetary policy response to additional fiscal tightening, to help offset the impact on aggregate demand if that were necessary to meet the inflation target. Suppose that:

  • Markets had correctly predicted the degree of tightening in the Budget and the MPC’s reaction to it; and
  • The MPC was able to offset precisely the demand effects of the Budget measures, and aimed to do so.

If this were the case, the short-term interest rates used in the pre-Budget forecast, which assumed no further fiscal tightening, would be too low. To be consistent, they would instead be higher to the degree required to compensate for the smaller fiscal tightening. Applying those higher interest rates would mean that, under these strong assumptions, the output path in the pre-Budget forecast would be the same as in the Budget projections.

There is a similar argument for long-term interest rates. Long-term interest rates are determined in international financial markets, and can be affected by many factors including expectations of future monetary policy, the fiscal position and risk premia. Empirically, lower debt and deficits are associated with lower long-term bond yields, though the relationship is complicated and probably non-linear. The likely effect of the Budget on long-term interest rates can nevertheless be estimated by using regression estimates of the relationship between bond yields and government deficits, debt, the output gap and inflation. This calculation suggests that bond yields could have been around 30 basis points higher over the next three years without the measures in this Budget.

Taking these points together, if the market expected further fiscal tightening in the Budget, then following the path of fiscal policy in the pre-Budget forecast would imply higher interest rates than those used. If higher interest rates had been assumed in the pre-Budget forecast, the implied path of output would have been lower. In other words, the differences in the path of output between the two forecasts would be at least partly offset by the effect of the different interest rate paths.

These calculations are illustrative, highly uncertain and depend on a range of assumptions. For example, although there was a widespread expectation of further fiscal tightening in the Budget, it is not possible to establish what exactly the market was expecting. The MPC is also unlikely to be able to offset precisely the demand effects of the Budget. Consequently, were it possible to produce one, it is unlikely that an adjusted pre-Budget profile for output would be identical to the Budget profile. Nonetheless, this illustration helps explain why it is potentially misleading to interpret the difference between the pre-Budget and Budget forecasts as the economic impact of the Budget measures.