Select Committee on Treasury Minutes of Evidence

Memorandum submitted by Professor Sheila Dow, University of Stirling


  1.  The forecast for RPIX shows a continued easing in the short-term, but less likelihood of a dip below 2.5% than the August Report predicted, in spite of the increase in repo rate earlier this month, and a rising trend towards the end of the forecast period. The risks are seen as roughly in balance around the central projection.

  2.  The forecast for GDP growth shows slightly stronger growth in the short-run than the August Report, with a surge of around one percentage point over the next two quarters, and then a more steady path close to 3% annual growth, with the balance of risks more on the negative side.


  3.  Both fan charts show smaller amplitude than in August, implying greater confidence in the central projections. But comparing GDP projections before and after the ONS revisions in the series is problematic, as is assessing how the relationship between GDP and inflation is altered by the revisions.

  4.  The revised GDP series indicates stronger growth in 1999 and 2000 than the earlier series, mainly due to an underestimate of growth in net trade. This means that 2003Q2 GDP is now shown as slightly higher in absolute terms than the earlier series. But the rate of GDP growth had been overstated in 2002-03; the picture is now one of fairly steady growth at around 2% over the last year.

  5.  The change in GDP series therefore poses problems not only for GDP projections, but also for analysing the influence of GDP growth on inflation. It is suggested in the Report (p 43) that greater flexibility on the supply side may explain why inflation was not higher during the period when the GDP data were underestimating aggregate demand, and therefore that "the factors which have been bearing down on inflation are likely to persist into the forecast period". Investment growth had been increasingly underestimated, suggesting that there is more capacity than was earlier thought, but the effect on inflationary pressure may be offset by higher GDP levels than previously recorded. The Bank will have recalibrated their core model to take account of the national accounts data revisions, but no indication is given in the Report of what that process has revealed about the relationship between output growth according to the new measure and inflation.

  6.  There are data series issues too with the inflation projection. If as expected the Chancellor changes from a 2.5% RPIX target to an HICP target, then that too will require recalibration of the Bank's models. RPIX inflation is slightly above target at 2.8% in Q3, while HICP inflation is currently only 1.4%, which is low (and low specifically when compared with an ECB target of 2%).

  7.  The current low level of HICP inflation underlines the proximity to deflation of non-house prices, and the risks that could pose were the economic recovery to be weaker than expected. Further, since house prices can be expected to follow a secular upward trend, the HICP series as presently constructed is likely to diverge increasingly from RPIX over time. Unless the target is changed to take account of this, monetary policy with an HICP target will be looser than with an RPIX target. (However, since there is the possibility that HICP could in due course change to include house prices, this issue may not arise.)

  8.  While the May 2003 Report provided detail on the difference between the two inflation series, it is disappointing that updated detail is not provided this quarter. There is a reference (p 35) to the fact that easing house price inflation will serve to erode the difference between the two series, but that is not the only source of difference. Indeed, house price inflation increased again in the third quarter, while the difference between RPIX and HICP was slightly reduced. Chart C on page 39 of the May Report shows that "other" factors accounted for RPIX and HICP at the end of 2002 being closer by 0.5% than would otherwise have been the case. Perhaps the Bank has developed an HICP forecast in anticipation of the change in target series. While it could be argued that the (as yet unannounced) precise specification of the target may itself impact on the outcome, it would have been helpful if an HICP forecast could have been provided in the meantime.

  9.  If indeed the target does change to HICP, then the Bank will be faced with basing its monetary policy on projections of a significantly revised GDP series and a completely different inflation series. This will make even more important the different types of evidence and analysis, and judgement, which go into the exercise, compared to the formal modelling framework.


  10.  The projections of aggregate demand, particularly household consumption and corporate investment, depend on developments in financial structure. The data shown in the Report are primarily aggregate data, showing financial ratios for UK averages, for example. Yet these averages mask wide disparities. Chart 1.10, for example, shows the rising ratio of average house price to average household disposable income. But the average house price masks wide disparities, even at the aggregate regional level. Average prices, according to the latest HBOS data for example, range from £82,068 in Scotland at one end of the spectrum to £223,460 in Greater London at the other end. Incomes similarly cover a wide range, but there is no indication as to whether the ratios between the two are similar.

  11.  When considering vulnerability to financial distress, which may need to be accommodated by sharp contractions in expenditure, it is important to identify those groups of households which are most vulnerable, such as those which are refinancing consumer debt in order to meet payments, and who are already paying high interest rates. Where there is also a low rate of employment, the vulnerability is exacerbated. The last Financial Stability Review, which addresses such issues directly, was published in June, since when there are signs that vulnerability for some groups may well have increased. The Bank have in the meantime commissioned a study into the composition of unsecured debt.

  12.  The Inflation Report shows that, in aggregate, unsecured debt as a percentage of household disposable income has roughly doubled since 1992; Chart 1.12 shows the ratio continuing to increase to over 22% in 2003. Of this total, outstanding credit card debt has more than trebled during the same period. Individual insolvencies are on an increasing trend (although as yet low in absolute numbers). Net secured debt has increased over the same period from around 75% of disposable income to 96% in 2003 Q2. Households have mostly been able to meet mortgage payments—mortgage arrears and repossessions have been steadily falling in recent years. But any significant reversal in house prices could change that situation. What is adequate security at high house prices can turn into negative equity.

  13.  There is therefore another vulnerable group, namely those who are heavily leveraged in the property market with a view to making capital gains—this group will be most evident in areas where house price growth has been fastest. This group is particularly vulnerable to falling property prices and rising interest rates. Even if others appear to be more financially secure, financial failure as a result of speculative activity at a time of weakening house price growth and rising interest rates prompts house sales by speculators which further weaken property prices and spread financial distress. The end result is more widespread cutbacks in consumption than would otherwise be the case, and reduced inflationary pressure. Financial stability is significant for inflation, and financial stability is best understood by studying disaggregated data.

  14.  While surveys of investment intentions have picked up (p 41), financial structure is still an issue for firms and may act as a drag on investment plans. Corporate debt is high, particularly relative to market valuation of capital, and some companies will be focusing on replenishing pension funds rather than investment. Corporate holdings of liquid assets has been increasing. While this can be regarded as a sign of financial health, it can also be taken as a sign that firms are too uncertain about the future to commit their financial resources to investment. Further, investment intentions are sensitive to household consumption, which as we have seen is itself vulnerable to financial structural conditions. Even though investment is shown to be higher in the new National Accounts measure than the old, there has still been a reversal of investment growth in the first half of 2003. Indeed, since investment has been shown to have been higher in recent years than it appeared in the earlier measure, there is likely to be more spare capacity now than earlier thought. Again, financial structure can have real consequences for investment planning and therefore also for inflationary pressure.

  15.  The composition of household and corporate financial structure is therefore particularly important when assessing incidence of financial vulnerability which cannot be detected from national averages. But it is more generally the case that composition of aggregate demand and supply are important for identifying pressures on costs and prices. Yet the Report offers no insights into regional composition of aggregates.

  16.  At the national level, the Taylor rule suggests that inflationary pressure builds up as the output gap (the difference between actual and potential output) becomes smaller. The output gap is difficult to pin down statistically. With that proviso, the Treasury study published in June, Assessment of Five Tests, shows that there can be marked differences in the deviations in regional output gaps from the UK average. In the 1990s, for Northern Ireland and the North-East, it is double the UK's deviation from the Eurozone in 1993-2002; the deviation was higher than average also for London, the East and Scotland. With less regional dispersion, the UK could have a higher GDP growth rate with less inflationary pressure.


  17.  There are indeed signs of improvement in aggregate demand, and of forces leading to higher price increases. But the risks attached to the Bank's forecasts seem higher than indicated in the Inflation Report. I have drawn attention here to two particular types of risk:

    —  The first type of risk is one of understanding the forces underpinning the new GDP series and its relation with the RPIX series, a risk which will be compounded by the expected need to switch to an HICP target.

    —  The second type of risk is posed by the vulnerability to financial instability posed by the level of indebtedness (secured and unsecured) and uncertainty about house price movements, and the potential for interdependence between financial instability and economic instability.

November 2003

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2004
Prepared 11 February 2004