Chapter I. The home economy.

Anderton, Bob ; Britton, Andrew ; Soteri, Soterios 等

The forecasts were prepared by Bob Anderton, Andrew Britton, and Soterios Soteri, but they draw on the work of the whole team engaged in macroeconomic analysis and modelbuilding at the Institute. Parts One and Two of the chapter were written by Andrew Britton, Part Three by Bob Anderton. We start our forecast this time with an exchange rate 91/2per cent higher in the third quarter of this year than we anticipated in May. The main reason for the appreciation of sterling was probably a reassessment by the market of the likely course of UK monetary policy over the next year or two. The likelihood of an early and significant fall in interest rates has receded, as forecasts of the underlying rate of inflation have been revised up. At the same time the commitment to join the exchange-rate mechanism has become firmer, and the Chancellor has been interpreted as wishing to 'talk the rate up'.

The higher starting rate of sterling itself has important implications for our short-term forecast. Moreover we share the changed view of the future course of monetary policy which we attribute to 'the market'. We now assume that base rates stay at 15 per cent until the second quarter of next year, failing to just 14 per cent by the fourth quarter.

If uncovered interest parity holds, then the existence of an interest differential in favour of sterling implies the expectation of sterling depreciation in the future. Applying that theory, and the principle of consistent expectations, to our forecasts, we now expect a faster rate of exchange-rate depreciation next year than we have previously. This would be consistent with membership of the exchange-rate mechanism, but with bands wide enough to permit some depreciation next year without a realignment. Looking further ahead our forecasts imply, as before, that downward realignments of sterling are allowed until 1997, when the UK is assumed to join an economic and monetary union.

Table A shows to what extent the change in our short-term forecasts can be explained simply in terms of different assumptions about the exchange rate and interest rates. The first section shows some features of our May forecasts. The second section shows those forecasts reworked with the exchange-rate and interest-rate assumptions we have now adopted. This can then be compared with the figures in the third section, which are taken from our new forecasts.

The higher level of the exchange rate and of interest rates next year is enough on its own to explain the lower growth rate of output we are now expecting for next year. The variant shows the current account deficit slightly improved, initially by the J-curve effect of exchange appreciation, subsequently by the lower level of activity. It also shows the rate of inflation next year markedly reduced, although this last effect must be seriously exaggerated in present circumstances.

Our new forecast also takes account of other recent developments and other information which has become available since May. The level of demand and activity in the first half of this year was higher than we expected. Consumer spending was still growing significantly, although the fall in retail sales in June may indicate that the trend has now flattened off. The buoyancy of import volume, impetus from the latest oil price rise. (Were it not for the rise in the exchange rate our forecast of earnings increase would have been raised considerably this time.)

The outturn for the current balance so far this year shows a deficit at an annual rate of 21 8 billion. We expect a small improvement in the latter half of the year, thanks to better terms of trade (higher world oil prices and a higher exchange rate). Next year the trade balance is expected to deteriorate a little as the volume effects of the exchange-rate appreciation make themselves felt.

The outlook for the medium term is based on our assessment of the effects of full EMS membership, followed by membership of an EMU in the late 1990s. inflation slows down to 'Franco-German' rates, with interest rates also coming gradually into line. The rate of growth picks up to a sustainable 2 1/2 per cent, with unemployment little changed at about 1 1/2 million. The current account of the balance of payments continues in substantial deficit for some years, gradually reducing towards the end of the decade, but that position should be financial in the policy context we assume. The surplus on the public sector increases again in the mid-1990s as lower interest rates ease the cost of debt service, but eventually the accounts move back to balance (or indeed a slight deficit if privatisation comes to an end). To understand the present problems of the UK economy it is necessary to look back at least three years and account for the very rapid growth of demand which took place from 1987 to 1989. The acceleration of inflation and the deterioration of the balance of payments have both been attributed to the boom which got under way at that time. in this section we conduct a brief historical investigation asking four questions. Could the boom have been prevented by the use of conventional instruments of fiscal and monetary policy? if so, why was it not prevented? Why was the need for action not anticipated? And could a similar sequence of events occur again?

A retrospective policy variant

We have used the institute's macroeconomic model to estimate what would have been the outcome for the economy if a different set of policies had been adopted from the beginning of 1987. The changes are made to the normal instruments of monetary and fiscal policy, that is to the level of interest rates, to the rate of income tax and the amount of public spending on goods and services. We have not assumed the introduction or re-introduction of any direct form of credit control or regulation.

In the retrospective policy variant we assume that the level of interest rates was held constant at 12.4 per cent continuously from 1987 to the end of 1991. The average level of interest rates in this period is therefore very similar to what has actually occurred (or is forecast to occur in the next 18 months). This 'counterfactual' case contrasts with an actual fall in interest rates in 1987, a steep actual rise in the latter half of 1988, continuing to the present, and with a small forecast fall during next year.

We assume in the variant that fiscal policy was progressively tightened in 1987, 1988, 1989, and 1990 by raising income tax instead of cutting it and by actual falls in the volume of public expenditure. The income tax increases (compared to the base run) are equivalent to rises of about 7p in the basic rate (with a direct revenue effect of El 2 billion a year) by 1989. The reduction in public spending is equivalent to about 8 per cent of public authorities' consumption by 1990. (It is assumed that these fiscal policy changes are reversed between 1992 and 1994). Had policy changes of this magnitude been introduced in the late 1980s they would doubtless have been regarded as very severe indeed. There can be few, if any, precedents for such a sharp deflation.

Before presenting the results of this variant, three points should be made about the methodology. The first is a general point about the treatment of expectations. In using the model to create a 'counterfactual history' we have to assume that the same expectational errors were made (by firms, unions etc) in the variant as in reality. Since these expectational errors cannot be identified we do not know how plausible an assumption this is.

The second point concerns the exchange rate, which has been assumed unchanged in the variant. This should be correct as an approximation. Neither a temporary tightening of fiscal policy, nor a change in the profile of interest rates with the average level unchanged, would necessarily result in a very different outcome for the exchange rate. The problem of using the model to recalculate more exactly the exchange-rate path is that the results depend crucially on the extent to which the policy changes were anticipated.

The third point concerns the behaviour of consumers. They too may make forecasts of future policy changes, and such forecasts may influence their spending. The announcement of tax cuts in the Budgets of 1987 and 1988 may have contributed to the consumer boom by raising expectations of tax cuts to come in the future. Similarly the much tougher fiscal stance shown in the variant might also have been extrapolated in expectations of further tightening to come. (Conversely some might argue that the more tax cuts there are now, the less tax cuts there are still to come.) Our consumption function is not forwardlooking in this way, so we cannot take any effects of this kind into account.

The results of the simulation are shown in table C. The level of output is reduced by about 1 per cent in 1987 and by a further 1.3 per cent in 1988. This reduction is smaller than the forecasting errors made at the time, despite the large scale of the deflationary measures assumed in the variant. Even in the variant unemployment falls in the late 1980s, but it levels off at about 2 million, significantly higher than the level actually reached this year.

The effect on the balance of payment is substantial. There is still a deficit on the current account after 1987, but it is much smaller than that actually observed, and by 1991 the account is almost in balance. It is likely that one consequence of this better balance of payments outcome would have been a higher level of the exchange rate, but that effect is not included in the variant shown.

The effect on the underlying rate of inflation is best judged by the figures for the deflator of GDP at factor cost. The effect of the assumed deflationary measures is felt mainly in 1988 and 1989. Later on the effect on the price level wears away, so that the rate of inflation is actually higher in the variant. The effects on the underlying retail price index are similar, but there are additional changes resulting from smoothing out the path of interest rates in the variant.

It is, perhaps, disappointing to find that the large scale demand deflation assumed from 1987 to 1989 does not do more to reduce inflation subsequently. In this context the fixed exchange-rate assumption is important, since one of the main reasons for the acceleration of inflation in 1990 is the fall in the exchange rate in 1989. It is possible that a more deflationary fiscal policy would have been associated with a higher exchange rate, even though interest rates were held constant.

Explaining past policies

The counterfactual simulation shows that it would have been possible to reduce the scale of the boom, and to prevent the balance of payments moving into such large deficit by taking deflationary measures in the Budgets of 1987 and 1988. In fact both these Budgets were expansionary, and in addition interest rates were reduced during 1987 and the first half of 1988. At the time the institute supported interest-rate reductions, but argued against tax cuts (National Institute Economic Review February 1988, p5). No one was advising the Government to embark on a deflation on the scale illustrated in the variant.

It is not clear to what extent macroeconomic policy was being conducted at this time with a view to managing domestic demand. Earlier in the 1980s such an aim was sometimes quite explicitly denied. It was asserted that policy was governed by objectives for monetary growth or considerations of fiscal stability in the medium term. By the end of the 1980S however the growth of 'money GDP', that is total output at current prices, had become the focus of attention. (When the targets for 'money GDP' were overshot this was excused on the grounds that the growth was real' rather than nominal'.) What was actually done by policy makers between 1987 and 1988 can be best explained however in terms of their preoccupation with the exchange rate and with the PSBR. Throughout 1987 and 1988 the sterling exchange-rate index was generally strong. Between the fourth quarters of 1986 and 1987 it rose by 8.9 per cent, and by a further 4.3 per cent by the fourth quarter of 1988. Moreover world interest rates were changing relatively little, and generally lower from those in the UK following the stock market crash of October 1987. External financial variables had been the most common trigger for rises in the level of UK interest rates throughout the 1970s and 1980s. There was on this occasion no need to tighten monetary policy in order to protect sterling. On the contrary the official reserves were rising strongly. This was the period when sterling was trying to 'shadow' the D-Mark, as if the UK were a full member of the EMS. it might be expected nevertheless that domestic policy, either fiscal or monetary, would have been tightened if the domestic rate of inflation had been seen to be high or accelerating. During 1987 however the average increase in the retail prices index was only 4.2 per cent, and it was little higher at 4.9 per cent in 1988. (Real interest rates were actually, very high in any historial calculation.) The forecast rate of inflation was 4 per cent in the Financial Statements of both the 1987 and 1988 budgets. The Institute's forecasts were only a little more pessimistic; indeed we remarked in the Review on how slight had been the impact of rising demand on the price level (National Institute Economic Review, February 1988 page 7), explaining that unemployment was still high, productivity was rising fast, and import prices were falling. The delay before the inflationary threat became manifest was one reason for the delay in the policy response.

An earlier symptom of excess demand was the deterioration in the trade figures. The Institute was amongst the first to draw attention to this, forecasting a serious balance of payments problem as early as the summer of 1986 (National Institute Economic Review, August 1986, page 5). In the event the trade figures improved briefly in the early months of 1987, before deteriorating very sharply by the end of that year. The line taken by the government was that the balance of payments presented no problem: it was the counterpart to a private sector financial deficit, and would therefore be self-correcting.

Believers in the 'political business cycle' will argue that action was not taken to reduce demand in 1987 because of the approach of a general election. Some may even suppose that the boom was deliberately engineered to suit the political timetable. The evidence on this occasion does not fit that theory well. The 1988 Budget was expansionary even though the election was over. The main rise in interest rates did not come until the latter half of 1988. A cynic would have predicted expansion in early 1987, with the brake sharply applied by the end of the same year. A more important influence on policy was the size of the PSBR, itself. We may argue that the scale of tax measures at the time of the Budget should be judged in relation to the state of the economy as a whole. In practice the state of the public sector accounts tends to attract more attention. The PSBR expected at the time of the 1987 Budget was 'only' 1 per cent of GDP; but at the time of the 1988 Budget the accounts were in surplus. The government, and many commentators said, and still say, that fiscal policy was very tight. By implication that meant that it could not be tightened further. A variety of explanations can be offered for the policies adopted in 1987, independent of the economic forecasts being made at the time. However, no explanation of the late 1980s would be complete without a post-mortem on the unusually large errors made at that time in forecasting the growth of demand and output.

Forecasting errors

Table D shows the forecasts of GDP growth published between February and June 1988 by the Treasury and 18 independent forecasters. For the year 1988 the convergence of views is remarkable. The range extends only from 2.8 per cent to 3.6 per cent; no-one at all took a significantly different view from the official Treasury forecasters. In the event everyone in the table was wrong. The herd instinct did not serve us well. (For the second year shown, the range was wider at 0.8 per cent to 3.3 per cent, and the average forecast, 2.0 per cent, was very close to the mark).

Table E shows the official forecasts of both March 1987 and March 1988 in more detail. The errors in the components of demand were even larger than the errors on GDP. The errors in GDP are about twice the average error from previous forecasts; the errors on consumption and investment in 1988 are three times the previous average. As the table shows some of the discrepancies can be explained by inaccuracies in the data available to the forecasters for the recent past. Thus in March 1987 the forecasters were told that consumers' expenditure had increased by 41/2 per cent in 1986; we are now told that the rise had been 51/2 per cent. This does not explain however why the forecasters expected the growth rate of spending to slow down in 1987; in fact it accelerated. A similar mistake was made by the forecasters in March 1988: again they were misled about the re growth of spending; again they compounded the error by expecting a deceleration instead of even faster growth in the next year. Recent statements have suggested that the policy errors of the late 1980s were partly the result of poor statistics(1). No doubt it would have been helpful if the coverage, especially of the expenditure data, had been better. It would be wrong however to see this as the main reason for the errors in forecasting output growth.

Table F shows successive estimates of output growth based on the three measures of GDP. The revisions to the output measure have been negligibly small. This is the estimate preferred by the Institute as an indication of annual growth rates, and used by us in forecasting throughout the 1980s. We cannot attribute the errors we made in our forecasts in the late 1980s to first estimates of the expenditure or income measures, since we largely discounted them at the time. Yet we made much the same errors as did the official forecasters. Like them we saw the economy slowing down, when in fact it was about to speed up even more.

Some of the factors which raised spending in the late 1980s cannot be accounted for even after the event. These show up as 'residuals' in our present equations for consumer spending, credit expansion and private investment. Taken together these 'residuals' raise the level of consumer spending by 2.4 per cent in 1988, of investment by 3.4 per cent and of GDP by 1.2 per cent. Further work is needed before we can explain in full the circumstances which led to the boom of 1987 and 1988. in the last-three years we have made progress by modelling better the influence of credit availability and financial wealth on consumer spending. The next stage may be to deal explicitly with consumer expectations. Could it happen again ? Having reviewed the operation of monetary policy at the beginning of this year, the authorities concluded that a reform of the method of control was not appropriate. Direct control of credit was again ruled out as an option. it also seems that little change is to be expected in the way that fiscal policy is conducted. The question arises whether another inflationary boom in the future could be avoided or corrected any more successfully than that of the late 1980s.

The possibility of such conditions emerging in the forecast period must be taken seriously. In the medium term, we expect interest rates to fall sharply, raising asset values and reducing the cost of credit. We hope that our model now takes account of the full effects this would have on spending by persons and by companies, but the period since credit was deregulated is too short for us to be confident that this is the case. Asset values could again be inflated in the future to an unknown extent, for example, by optimism aroused by the closer integration of Europe. The approach of another general election will also affect sentiment in a way that we cannot now foresee. The possibility of a boom beginning in say 1992 certainly cannot be ruled out.

It is to be hoped that next time, with improved statistics, better econometric models, and heightened vigilance, the forecasters, (ourselves included) will spot the boom in its early stages. It will however be more difficult next time for the authorities to respond quickly to such a warning if it is given to them. Once the UK is a full member of the EMS, interest rates will have to be guided by the prospects for the exchange rate. it will be fiscal policy (by default) which will have to cope with disturbances to domestic demand. As things stand, with Budgets just once a year, that response cannot easily be a timely one.

PART THREE. THE FORECAST IN DETAIL

Forecasts of expenditure and output (table 1) We begin this forecast with a higher exchange rate than we were expecting in the May Review and higher than expected outturns for some expenditure categories (such as consumers' expenditure and stockbuilding). The major effect of these two factors is to raise our forecast of demand and output for this year, with a correspondingly lower forecast for the growth of output next year. Most of this year's forecast of 2 per cent output growth occurs in the first half of the year and the path is fairly flat thereafter. Our new assumption of 15 per cent base rates persisting until mid-1991 adds to the depressing effect of a weaker export performance next year (due to the loss in competitiveness) and limits growth in output to around 2 per cent for 1991.

Previously we expected a fairly abrupt slowdown in consumers' expenditure this year, but given the recent path of real incomes we now believe that a gentler deceleration in consumption will continue into next year. Our forecast profile for investment has also changed in response to the recent higher than expected level of demand and our new path for interest rates; although the sum of predicted investment growth over this year and next is the same as our May forecast, it seems more probable that there will only be a slight fall in investment growth, of around 1 per cent this year, with an equally subdued recovery next year. The large de-stocking we were expecting has not materialised (de-stocking seems to be decreasing instead of increasing) and we are now assuming only mild de-stocking this year.

The net contribution to output from the balance of payments is now lower in this forecast for both 1990 and 1991. The higher general demand this year has increased imports (we, like most other commentators, had expected the high import intensity categories such as investment and stockbuilding to experience a larger downturn) and the deterioration in competitiveness, resulting from both a higher exchange rate and a higher than expected inflation rate, will have a particularly damaging effect upon exports next year. Personal income and expenditure (table 2) Although there are various strong indications that the growth of consumers' expenditure is decelerating from the rapid growth rates of recent years (for instance the large fall in retail sales in June which was only marginally offset by a mainly seasonal increase in July), the general impression is that consumption is still showing some resilience. Recent Gallup surveys of consumer confidence have shown a recovery in all components of the consumer confidence index, although the level of the index is still hovering around that recorded at the trough of the 80/1 recession. The recent data on consumer credit and bank lending give a stronger impression of a consumption slowdown but it has been suggested that the lower rate of borrowing may be connected with non-payment of the poll tax. The figures for the first quarter of this year for total consumption are actually quite high although expenditure on durables has obviously weakened substantially (UK new car registrations are substantially down on last year).

Given the high level of interest rates combined with the low level of consumer confidence resulting from falling house prices and rising unemployment we are expecting subdued real consumption for the rest of this year. Real disposable income will be boosted by the high rate of wage inflation and some fiscal measures such as the continued reform of national insurance contributions and independent taxation. Conversely, but to a lesser extent, fiscal drag and lower growth in dividends and overseas property income will have a negative impact upon income this year. Given the high consumption outturn and the growth of disposable income we expect the growth in consumers' expenditure to be around 3 per cent this year. The continuation of high interest rates through 1991, and the depressing effect of increased unemployment and decelerating wage inflation upon disposable income, will probably moderate consumers' expenditure further next year.

Fixed investment and stockbuilding (tables 3 and 4) The latest CSO survey of investment intentions for manufacturing industries indicated that real manufacturing investment would increase by around 1 per cent this year followed by a further larger increase next year. The survey also revealed that investment in some of the UK's more successful export industries-such as chemicals and electrical engineering-was expected to experience even larger increases. At the time of the survey the UK effective exchange rate was approximately 7 per cent lower than its current level and this may explain why the July CBI industrial trends survey gives a far more pessimistic view of the investment intentions of manufacturing industry. The July CBI survey shows that growth in export orders has now come to a standstill whereas in previous surveys it seemed that growth in export orders sustained demand for output even though home demand was weak. The survey also reveals that capacity expansion as a reason for investment has decreased recently which may be related to the bleaker export prospects. There are also signs that downward pressure on profitability is affecting investment, as 25 per cent of the firms surveyed by the CBI indicated that a shortage of internal finance will inhibit their investment plans in the year ahead. A larger proportion of firms are also citing the cost of finance as a constraint on investment which may be a result of firms resorting to more expensive external finance as internal funds dry up.

Our forecast seems consistent with most of the main findings of the CBI survey. We are forecasting a fall in investment of 1 per cent this year (with falls occurring mainly in manufacturing and construction). In addition, the higher level of the exchange rate in our forecast restrains export growth resulting in a further fall in manufacturing investment next year. The combination of a higher exchange rate, high interest rates and slow growth next year will probably restrict total investment growth to around 1 per cent in 1991.

Recent stockbuilding figures show that substantially less de-stocking took place in the first quarter of this year than the previous quarter. In addition, the July CBI survey shows that stocks have been run down less than predicted by the April survey. Although a further period of de-stocking is anticipated by the CBI survey respondents, the larger firms (over 5,000 employees) now expect a stable trend in stockbuilding. It may be the case that these larger firms are much more adept at maintaining stocks at more efficient lower levels by using computerised stocks systems (this seems consistent with the rapid decline in the stock/output ratio in recent years). Therefore, perhaps less de-stocking is now appropriate when demand slows down as stocks are already kept at minimum levels. Although there is downward pressure on profit margins the relatively high level of profitability compared to previous downturns may also partly explain why de-stocking is fairly restrained so far. Given these factors and our upwardly-revised forecast for output growth this year we now expect only mild de-stocking during 1990 with some subdued positive stockbuilding through 1991.

Balance of payments (tables 5 and 6)

Since our May Review the current balance deficit for 1989 has been revised down from E!21 billion to 219 billion. The balance of payments data for 1989 also reveal a large balancing item of 215 billion which is virtually the size of the deficit itself. This implies that either capital flows are under recorded or that the current account deficit was actually much lower in 1989. Furthermore, the current balance for last year would actually be in surplus if the substantial unrealised capital gains on UK net overseas assets were recorded in the current account.

Given these data and conceptual problems associated with the balance of payments statistics it is not surprising that some economists argue that the UK current balance deficit is not a problem. However, the recorded current account deficit for the first six months of this year is already C9 billion and it will take more than data revisions to make this disappear. This large deficit has persisted in the early part of this year partly because domestic demand has not decelerated as much as expected. In particular, the highly import intensive categories such as investment and stockbuilding are slowing down at a fairly gentle pace. Import volumes have therefore persisted at a high level in the first half of the year. Export volumes, especially manufactured goods, have been growing strongly but the latest figures suggest a recent deceleration which supports the findings of the CBI survey that export orders and optimism are losing their buoyancy. The UK oil trade balance has been disappointing as oil production has experienced an extremely muted recovery (there may be further production disruptions caused by industrial unrest but a higher oil price may encourage a quicker recovery of UK oil production). Given this rather bleak state of the UK current balance the recent large increase in the UK exchange rate seems inappropriate, as it will hinder an improvement in the trade deficit next year. Partly because of these factors we have revised upwards our forecast for the current account deficit to around 21 8 billion for this year and we do not expect much of an improvement in 1991.

Although some of our more pessimistic trade forecast originates from visible trade a large part of the poorer performance occurs in invisible trade. Recent data for 1989 show that the overseas earnings of 'the City' have fallen for the third year in a row. The banking sector was in international deficit for the first time since 1979. This is partly caused by the very high rates of interest paid by UK residents borrowing foreign currency (presumably partly to pay for the trade deficit itself) which has resulted in a large increase in net interest payments paid overseas. With high interest rates expected in the UK for some time there is little scope for an improvement here. The recent surge upwards in sterling results in downward revaluations of our assets abroad. Consequently, the already low level of net income received from interest, profits and dividends will become even lower in sterling terms. Although government transfers abroad should decrease slightly on last year we are forecasting very low invisibles surpluses for this year and next. Net exports of services do not grow much next year as the higher exchange rate worsens UK competitiveness.

The latest data for the capital account indicate that direct investment into the UK has remained at a high level in the first quarter of this year with the two banks Morgan Grenfell and the Yorkshire Bank being purchased by Deutsche Bank and the National Australia Bank respectively. UK portfolio investment overseas also exerted a positive influence upon the basic balance (structural flows plus the current balance) as actual portfolio disinvestment occurred in the first quarter. It is partly for this reason that our forecasts for portfolio investment outflows for this year and next have been revised downwards. However, other factors discouraging portfolio investment abroad are the high exchange rate, which reduces the sterling returns on assets held overseas, and our assumption of a longer period of high interest rates in contrast to our previous forecast of rapid cuts in interest rates. However, the expected basic balance deficits for this year and 1991 are still similar to our previous forecast, as the smaller net portfolio outflows have been somewhat offset by larger current balance deficits.

Output and the labour market (tables 7 and 8) Unemployment has increased over the last quarter. There have also been recent falls in job vacancies and it seems that the slowdown in domestic demand is now affecting employment growth. These aggregate figures tend to obscure some interesting facts; unemployment is mainly rising in the south east, it is actually continuing to fall in many other regions; female unemployment is still falling, which means that male unemployment is increasing faster than total unemployment; employment growth is still occurring on a substantial scale (115,000 new jobs were created in the first quarter of this year). Latest data show that the number of overtime hours worked is declining, which provides further evidence of a loosening in labour market conditions. in our forecast we expect unemployment to increase at a faster rate in the second half of this year and it will probably peak some time next year.

According to the CSO, manufacturing output is still on an upward trend well into the second quarter of this year. This is surprising given the demand slowdown but it does seem that the stronger manufacturing output growth is concentrated in the UK's more recently successful export industries such as the chemicals industry. Non-manufacturing output is experiencing a sharper slowdown especially in the energy sector which has been subdued by oil production problems and the warm weather. The combination of manufacturing employment falls and output growth has helped prevent manufacturing productivity declining too much and latest data show that output per head is growing at over 2 per cent at an annual rate (but for the whole economy output per head is actually falling). A danger here is that manufacturing wage settlements will incorporate this productivity growth and other sectors will follow suit while achieving inferior productivity growth rates.

Our forecast sees manufacturing output growth falling from the third quarter of 1990 as the recent growth in manufacturing output does not seem sustainable or consistent with other data series (such as the large falls in manufacturing capacity utilisation cited by CBI survey respondents). The higher level of sterling is a major reason for the expected decline in manufacturing output and as this depresses manufacturing exports next year we also expect manufacturing output to be subdued through 1991. Construction and distribution are two other sectors for which we expect only modest growth over this year and next. Growth in real GDP excluding oil for 1990 and 1991 is only slightly different to total GDP as we are now forecasting a slower recovery in oil output in the near future.

Wage and price inflation (tables 2,9 and 10)

After the impact of recent events upon the price of oil, the UK is now facing the prospect of the RPI peaking at an annual rate that could be closer to 11 per cent than 10 per cent. Less than six months ago the Chancellor was forecasting that the RPI would be decelerating to around 71/4per cent by the fourth quarter of this year. We are now forecasting retail price inflation to be around 3 percentage points higher than that in the fourth quarter. Ourforecast of retail price inflation of just under 8 per cent by the fourth quarter of next year is higher than previously partly because underlying inflation is higher and also because interest rates are not falling as rapidly.

Since the end of last year, sterling has risen and UK manufactures have been faced with falling input prices (fuels and materials prices). Yet manufacturing output prices have shown continual monthly rises throughout this year. Rising unit labour costs, perhaps combined with a reluctance to allow profit margins to decline must therefore account for this discrepancy between input and output prices. Profit margins should fall as demand slows down but output prices may also be reflecting higher inflationary expectations which will be shared by wage negotiators. Average earnings are currently rising at per cent (although manufacturing earnings are only increasing by 91/4per cent). The latest data also indicate a slight reduction in overtime hours in manufacturing which, if it continues, will reduce wage drift (the difference between wage settlements and average earnings). We are forecasting a 91/2per cent increase in average earnings for this year with a slight deceleration next year. This slowdown in wage inflation in 1991 should result from lower productivity related payments and an increase in the excess supply of labour which will manifest itself as an increase in unemployment next year. However, our average earnings forecast could prove too conservative if inflationary expectations are spiraling upwards. Furthermore, some large companies will be implementing second stages of pay deals based upon a certain percentage above inflation (for instance, some Ford workers' pay rises will be the inflation rate plus 2 1/2per cent).

The consumer price index will be subject to a discontinuity from the second quarter of this year because of the treatment of the poll tax in the National Accounts. Domestic rates were previously classified as a tax on housing services and part of nominal consumers' expenditure whereas the poll tax is treated as a deduction from income. As the volume of housing services consumed will be the same, the volume of consumer expenditure does not change. Consequently, the implied consumers' expenditure deflator will fall, but this obviously does not represent any change in underlying inflation. The main tables in the Review now reflect these changes.

Public sector finance (table 11)

Optimism concerning the size of future public sector debt repayments has substantially diminished since the 1989 Budget. At that time, Nigel Lawson predicted a PSDR of 14 billion pounds for the 89/90 fiscal year whereas the actual debt repayment was only 98 billion. Three main factors account for this over prediction of the public sector surplus:

(1) Revenue from national insurance contributions was depressed by the unexpectedly high take up of private pension plans.

(2) Spending by local authorities was boosted by the changeover to the poll tax and additional capital expenditure was brought forward to avoid new regulations introduced in the next financial year.

(3) Slower output growth adversely affected retail sales and company profitability which reduced revenue from expenditure taxes and corporation tax.

Therefore, forecasts for this financial year's PSDR begin from a lower level and the latest Treasury forecast now expects a debt repayment of around C7 billion for 90/1. However, this projection now seems optimistic as recent data show a 6 1/2 billion pounds public sector deficit for the first quarter of this fiscal year. Delayed poll tax payments are partly responsible for this-forcing local authorities to borrow at very high interest rates-and the financial situation should improve as the year progresses and more arrears are paid. But higher public spending is also a substantial cause of the deficit and it remains to be seen whether the imposition of cash limits will restrain spending to a greater extent in the latter part of this financial year. There are several other factors which will have a negative impact upon the PSDR this year; independent taxation of husband and wife and the continued reform of national insurance contributions will further depress revenue; rapid inflation feeding into public sector pay deals will increase expenditure; a further slowdown in output will also have a positive impact upon net expenditure as tax revenues increase less rapidly and current grants increase as unemployment begins to rise. Given these factors and allowing for privatisation receipts from Powergen etc we expect a public sector debt repayment of around 24.5 billion for this financial year.

Public spending for the 91/2 financial year will be especially difficult to control as it is a pre-election year. Ministers have already made bids for 21 7 billion more than the current target figures for next year. However, given our predicted recovery in output (and hence revenues) in the latter part of 1991 and the governments assumed privatisation proceeds of C5 billion, we expect a higher PSDR for next year of approximately E8 billion.

Medium-Term Forecast

Once again, but to a greater extent this time, the balance of payments tends to dominate the medium-term forecast. We envisage fairly large and persistent current balance deficits for the next few years with a substantial improvement only occurring in the mid-1 990s. With fairly slow progress towards convergence to lower interest rates and beginning from a high level for the exchange rate the deficit tends to perpetuate itself. Large UK deficits require capital inflows which, in turn, require higher relative returns (i.e. high UK interest rates) which decrease UK net property income from overseas. This makes the current account deficit worse, thus requiring greater capital inflows etc. The high level of the exchange rate not only exacerbates the visible trade deficit, but it also reduces IPD income paid in foreign currency from overseas. Throughout the medium term we have export volumes of goods and services growing more rapidly than imports. But starting as we do from a high level of imports relative to exports, the differential between export and import growth rates would have to be much larger to deliver a trade surplus.

Given this outlook for the balance of payments we expect real GDP growth in the medium term to be constrained to about 2 1/2 per cent a year. Consumer spending over this same period should grow at a more sedate pace (slightly below the growth rate of GDP) as the initial surge effects of financial liberalisation comes to an end. However, tight fiscal policy will be necessary to produce sufficient total savings to keep the current balance deficit under control. The lack of tax cuts and limited growth in public spending will restrain demand and also encourage a more sustainable path for consumers' expenditure. Given that a tight fiscal policy is necessary, since the scope for monetary policy is limited within the ERM, and the probability that there will be considerable public expenditure costs in areas such as defence, we expect public sector debt repayments of around 1 1/2 per cent of GDP for some time to come.

Our forecast produces an inflation rate that gradually converges towards other European rates. However, this is a slow process and the UK inflation rate is always above that of our European partners, right up until the time of European monetary union around 1997. This inflation differential is associated with the depreciation of sterling, which continues even though the UK is assumed to participate in the exchange-rate mechanism. The general path of decelerating UK inflation throughout the medium term indicates that, at around 2 1/2 per cent GDP growth, the UK is operating below its sustainable level of activity (in other words, unemployment is greater than the NAIRU). Average earnings growth is also forecast to move downwards towards European rates, which is consistent with wage discipline being exerted from membership of the ERM.

NOTES

(1) Monetary policy in the second half of the 1980s' lecture by the Governor of the Bank of England at the University of Durham, reprinted in the Bank of England Quarterly Bulletin, May 1990. See also 'Official Statistics in the late 1980s' by J. Hibbard, Treasury Bulletin, Summer 1990.