Chapter I. The home economy.

Anderton, Bob ; Britton, Andrew ; Gregg, Paul 等

CHAPTER 1. THE HOME ECONOMY

PART ONE. RECENT DEVELOPMENTS AND SUMMARY OF THE MAIN FORECAST (table A)

The probability has increased that the UK will become a full member of the EMS before the next general election. The issue is by no means settled, but full membership now seems the right assumption to make for the forecast. The precise timing is difficult to foresee: on the one hand the present economic situation in this country makes an immediate move difficult, on the other hand the Government might be loath to make the move in the run-up to the election. Fortunately the choice of the exact date is not very material to the forecast. We have assumed the fourth quarter of the year is the date of entry. A more important question concerns the terms on which the UK joins: whether sterling joins at the market exchange rate of the day and the width of the band within which it can fluctuate.

Given the recent experience of inflation in this country, and the continuing deficit on the current account the relationship of sterling with other currencies in the exchange rate mechanism must be an uneasy one, at least at the outset. The market could therefore put considerable downward pressure on sterling later this year, especially if the deutschemark continues to be strong relative to the dollar and the yen. A `realignment' to remove the pressures at the point when sterling enters the ERM cannot be ruled out, although it is not the most likely outcome. The most likely outcome would be that sterling remains relatively strong, and enters the ERM at the rate ruling in the market. Our forecast for the sterling-deutschemark rate in the fourth quarter of the year is 2.72. We assume that this is the rate at which sterling enters the ERM with a margin of 2 1/2 per cent.

As Chapter II explains, we expect the transition from the present EMS to an economic and monetary union to take many years. In the meantime exchange rates within Europe will not be `irrevocably' fixed and occasional realignments will take place. Since we cannot forecast the timing of these events we assume a steady trend in relative exchange-rate movements rather than discontinuous jumps. On balance we expect sterling to be one of the weaker currencies even in the medium term. Our assumption therefore is that sterling depreciates against the deutschemark to the same extent as does the Italian lira.

The implication of this assumption, together with the other exchange-rate forecasts set out in Chapter II is that sterling will depreciate slightly in effective exchange-rate terms.

Initially the depreciation is about 2 per cent a year, slowing down to about 1 per cent a year towards the end of the decade. This is a significantly slower rate of depreciation in the medium term than we assumed in our November forecasts. It is partly for this reason that we now show a considerably lower rate of inflation in the medium term. By the mid-1990s, on the new exchange rate assumptions, we would expect the increase in the consumer prices index to be about 2 1/2 or 3 per cent, compared with 6 per cent this year.

Entry to the ERM is also crucial to our forecast of UK interest rates. Currently, 15 per cent short-term interest rates in this country compare with rates of 8 1/2 per cent in Germany, 10 3/4 per cent in France and 13 3/4 per cent in Italy. We would expect these gaps to narrow significantly once sterling has established its place in the ERM. Indeed the first step may be taken even before then: we are forecasting a cut in UK interest rates of 1 percentage point by the third quarter of the year. Over the next few years our forecasts show a marked fall in UK interest rates, to an average of 11.3 per cent in 1991 and to 7 1/2 per cent by 1995.

Our assumptions for fiscal policy for the short term are partly based on the hints dropped by Treasury ministers before the pre-Budget period of purdah began. We assume a `neutral' Budget this year: that is to say no new changes in taxation relative to an indexed base, although the effects of changes already announced in previous Budgets will be to reduce revenue, from income tax and from National Insurance contributions next financial year.

Looking further ahead the fiscal room for manoeuvre is constrained by the need to improve the current account of the balance of payments. The strength of the real exchange rate and the fall in UK interest rates make this task more difficult. Hence fiscal policy is assumed to be tight throughout the forecast period, with no tax cuts and the growth of public consumption limited to 1 to 1 1/2 per cent a year in volume terms. The financial balance of the public sector is forecast to remain in surplus for the foreseeable future.

A continuing surplus for the public sector would be a most unusual development by historical experience, or in comparison with experience in other countries. It arises from the equally unusual behaviour of the personal sector in the UK, which seems to be running a persistent financial deficit financed by a high level of borrowing. If interest rates come down, as we expect they will for the reason already given, then it is unlikely that households will have any strong incentive to save more or to cut back their borrowing. Hence the need for a fiscal surplus will remain.

Against this policy background the main uncertainty is how fast the UK economy can grow. It has been clear for some time that a much slower rate of growth, or even a recession, is unavoidable in the short term, to make up for the boom of 1988. Recent developments confirm that this slowdown is now happening.

The growth of GDP in 1989 was probably about 2 1/2 per cent, with non-oil GDP up by nearly 3 1/2 per cent. This is a good deal slower than the two previous years, but still above the rate which we think is sustainable in the long run. Looking back over ten years the average annual growth rate of non-oil GDP was 2 1/2 per cent a year. Within that period there was an acceleration, but it may have been largely cyclical in character, and has involved an element of catching up after the poor performance of the 1970s. On the other hand the stimulus of competition within the single European market would argue for more sustained improvement during the 1990s. In our forecasts we have not made any special adjustments to the supply-side to take account of any of these arguments, as they seem on balance to cancel each other out. The average growth of non-oil output implied over the next decade would be about 2 1/2 per cent a year. This is consistent with a gradual move back to current account balance in the late 1990s.

The immediate prospect is for growth of only 1.4 per cent in GDP this year, or 0.8 per cent excluding oil production. Compared with our forecasts published last November, we have revised up the growth of consumer spending, revised down the level of investment, and reduced the scale of de-stocking. The net effect of these changes on domestic demand and on output is small.

In the main tables for this chapter we show 1991 in detail for the first time. This year, now appearing for the first time over the short-term horizon, shows the early stages of a cyclical recovery, with growth of about 2 1/2 per cent fourth quarter to fourth quarter The pattern of expenditure changes is a familiar one for this phase of the cycle, with the end of the rundown in stocks coming during the year, and with a recovery in investment. The falls in interest rates, although they are associated with entry into the ERM, come at a time in the cycle when interest rates have often fallen on previous occassions. So they serve to reinforce the normal pattern of recovery.

Unemployment is expected to level off in the first half of this year at about 1.6 millions and then to rise very slowly. This assumes no further changes in the criteria for eligibility to benefit. If the criteria were tightened further it is possible that the claimant count could fall further.

Since our November forecasts the trade figures have improved more rapidly than we expected, and for the immediate future our forecasts of the trade deficit have been revised down significantly.

Recent developments in wages and prices have not been so encouraging. The r.p.i. for the fourth quarter of last year (7.6 per cent up on a year earlier) was a little higher than we expected. Wage settlements have accelerated from a rate which was already fast relative to the rate of price increase given the slackening of demand in the labour market. In Part Two below we describe a `wage variant' on the main forecast, in which the pressure for higher real wage growth persists. For the main forecast however we assume average earnings growth of about 8 1/2 per cent between the fourth quarters of this year and next, compared with nearly 10 1/2 per cent over the same period of the preceding year. This is consistent with a forecast of r.p.i. growth rather higher than that in our last Review: between the fourth quarters of 1989 and 1990 the increase is expected to be about 6.8 per cent.

Looking further ahead the rate of inflation comes down sharply, partly because of the exchange rate path already described and partly because the reductions in interest rates have a direct effect on the r.p.i. This relatively rapid `disinflation' is consistent with the forward-looking character of the wage and price equations of our model. The longer-term implications of full EMS membership are assumed to be understood and given weight in wage bargaining and in price setting. In the variant described below this `rationality' assumption is retained, but the pressure for real wage increases is assumed to be stronger nevertheless. [Tabular Data Omitted]

PART TWO. THE CONSEQUENCES OF MORE WAGE PRESSURE (table B)

The size of recent wage settlements has been a source of anxiety either on the grounds that inflation will be higher as a result, or that unemployment will be higher, or both. In this section we explore the possible consequences of wage pressure, first on the assumption that the exchange rate and interest rates are fixed, and secondly on the assumption that the exchange rate is free. In both cases the Institute's domestic model has been used, and the variation from the main forecast case arises from adding a residual of 1/2 per cent to the equation for average earnings throughout the forecast period. The dynamic characteristics of the model are such that the size of this shock to earnings is magnified considerably, as the results in table B show.

The fixed exchange-rate variant can be seen as an interpretation of the effects of wage pressure in one country within a monetary union. The nominal exchange rate is unchanged, and hence the real exchange rate appreciates, making UK goods less competitive. In terms of relative wages the change amounts to 8 per cent by 1992. With nominal interest rates fixed the real rate of interest is initially reduced, giving a modest impetus to fixed investment.

As the EMS is not a monetary union it is of course possible that the exchange rate and interest rates would in fact change, but the results of the first simulation are interesting nevertheless as a limiting case. It is possible, moreover, either in the EMS or in a monetary union, that a fiscal policy change would be made in response to wage pressure; indeed some response might be unavoidable. But, as we shall see below even the direction of such a response is not certain.

The effects of wage pressure in the fixed exchange-rate case can be seen by comparing the figures in table A and B(i). The effects on the levels of nominal earnings and prices peak in 1993; thereafter the effect on inflation is trivial. This means that there is a permanent loss of competitiveness which reduces the volume of exports and increases the volume of imports. Nevertheless, after 1994 the effect of wage pressure is to improve the balance of payments.

This rather surprising result follows from the lower level of output and economic activity. Looked at from the supply side of the economy, wage pressure reduces the sustainable level of capacity use both in terms of labour and of capital. The `non-accelerating inflation rate of unemployment' (NAIRU) is raised. In fact unemployment by the mid-1990s is higher by about 100 thousand. Capacity utilisation in manufacturing is also down by about 1 1/2 per cent.

On the demand side, real consumer spending is permanently higher thanks to the higher level of real wages, but investment is permanently lower (despite the initial fall in real interest rates), because profits are lower. As a result the growth rate of capacity is slowed down, by about 1/4 per cent a year after 1992.

To sum up, in this variant wage pressure is `successful' in that real wages are permanently raised and the UK terms of trade `improve' without any lasting adverse effect on the current balance. The cost is a burst of rapid inflation lasting for a few years, a loss of output and employment in perpetuity, and also a small reduction in the rate of growth of productive potential.

Comparing table B(ii) with table A shows the effects of the variant with the exchange rate free; interest rates are again held fixed. The result of extra wage pressure is exchange rate appreciation. Far from easing the problem of price competitiveness by depreciating, the exchange rate actually rises relative to the main forecast by about 1/2 to 1 per cent a year. This moderates the initial burst of inflation, which is now followed by a long period when inflation is rather lower than in the base case. The price level is the same in the variant as in the main case towards the end of the decade.

The consequences for the real variables, beyond the first year or two, are very similar whether the exchange rate is fixed or free. More wage pressure in both cases results in lower output, higher unemployment and a slower growth of capacity for the whole of the forecast period.

The consequences for the current balance in the free exchange-rate case are similar initially to those of the fixed exchange-rate case: the adverse effect in the first few years arises from the loss of competitiveness, and this is again followed by a period when lower output improves the current balance. But as the exchange rate continues to appreciate, the effect in the long run is slightly adverse in table B (ii), although it was slightly favourable in table B(i).

The precise movement of the exchange rate, the size of the initial jump, and the scale of the subsequent appreciation, are not of any great significance. A more robust conclusion is that the behavior of the exchange rate is unlikely to alter very greatly the consequences of wage pressure for real variables. The exchange rate, it seems, does not respond to cost or price inflation directly, but rather to the prospects for the supply and demand for sterling in the medium-term or long-term future. The effect on those prospects of a loss of competitiveness from higher wages may be at least offset by a lower level of output. The implied movement of the exchange rate could in theory go either way.

A similar point applies to the appropriate response of fiscal policy in the fixed exchange-rate case discussed above. Inside a monetary union, fiscal policy may take the place of interest rates and exchange rates as a means of correcting balance of payments disequilibrium. In the variant described here the implication would be that fiscal policy could be marginally less tight on average over the next decade if wage pressure were more intense. This somewhat counter-intuitive result follows from the entirely plausible finding that wage pressure improves the current account in the long run. But even if fiscal policy were more expansionary the effect of wage pressure on unemployment would not be nullified. That is a supply-side effect, which would still be there in the long run, even if (ex ante) demand were to be higher. [Tabular Data Omitted]

PART THREE. THE MAIN FORECAST IN DETAIL

Forecasts of expenditure and output (table 1) Much of the data for late 1989 is still provisional but it seems that last year GDP grew by around 2 1/2 per cent which, after taking into account the fall in oil production, is above the long-term trend growth rate of output. However, this masks a fairly abrupt deceleration in output growth during the latter part of 1989 which we expect to continue through 1990. Output growth this year will probably be in the region of 1 1/2 per cent if the expected recovery in oil production actually occurs, adding to our forecast of under 1 per cent growth in non-oil GDP.

Our assumption of UK ERM membership towards the end of this year allows interest rates to fall substantially, boosting growth to 2.3 per cent in 1991. Although these figures are similar to our November forecast the behaviour of the demand components is somewhat different. We are now a little less optimistic about investment this year as we expect it to fall slightly in real terms. However, we are not too gloomy about the supply-side of the economy; our forecast now sees improvements in both exports and imports and our interest-rate path encourages healthy investment growth next year. We now have consumers' expenditure a little higher this year and next and de-stocking occurs on a smaller scale in 1990. [Tabular Data Omitted]

Wage and price inflation (tables 2 and 8) The underlying increase in whole economy average earnings was around 9 1/4 per cent at the end of 1989 (see appendix B on UK Wage Developments at end of chapter). Combined with slowing output growth and the corresponding decline in productivity growth this has resulted in a substantial rise in the rate of increase of unit wage costs. Over a similar period the increase in producer output prices has not shown any acceleration with profit margins presumably dropping in order to absorb costs increases. We are forecasting a year-on-year growth in whole economy average earnings of 9.3 per cent for this year compared with over 10 per cent last year. We give a large weight to manufacturing productivity in our earnings equation as this is a leading wage-setting sector. This, because of our forecast of a manufacturing output fall in 1990, helps restrain earnings this year. Next year should see a further slight slowing of earnings growth as the increase in unemployment, which will begin some time this year, reduces labour demand pressures offsetting the small recovery in productivity resulting from increased output growth.

We have decided on this occasion to treat the poll tax as if it were an expenditure tax similar to the domestic rates which it replaces. We have made a small increase to the c.p.i. in the belief that the changeover will not be neutral. In fact the national accounts will treat the poll tax as a deduction from disposable income and we shall fall in line with this convention in due course.

Manufacturing wholesale prices, which feed into consumer price inflation, will probably increase faster this year as slower output growth and rising earnings increase costs. The assumed small fall in sterling also increases import prices with manufacturing imports increasing in price more than the total imports deflator shown in table 8. We expect consumer price inflation (as we define it) to exceed 6 per cent this year and to remain fairly high in 1991 as demand pressures return while import prices and earnings remain buoyant.

The annual growth of the r.p.i. was unchanged at 7.7 per cent in December of last year but the increase in mortgage interest rates this month will result in rather higher figures for the early months of the year From April onwards the poll tax will increase the r.p.i. by about 1/2 per cent, but falling interest, and hence mortgage, rates will bring retail price inflation down to below 7 per cent by the end of 1990. Further mortgage rate falls in 1991 should bring the r.p.i. down to around 4 1/2 per cent through next year. [Tabular Data Omitted]

Personal income and expenditure (table 2) Data for December of last year actually showed a decline in outstanding consumer credit. In addition, the volume of retail sales fell in January. These figures suggest that consumers are responding to high interest rates although consumers' expenditure data shows less of a slow-down in purchases of non-durables.

The impact of interest-rate changes upon consumption is an important feature of our forecast as we assume interest rates fall substantially as the UK joins the ERM. In addition to the effects of mortgage rates upon real personal disposable income the expected falls in the interest rate will also influence real consumers' expenditure via credit and wealth effects. Our economic model explains the growth of consumer credit in terms of consumers slowly adjusting towards some target interest payments to income ratio. Therefore, as a sharp rise in interest rates would increase this ratio above that targeted by consumers we could then interpret the recent fall in outstanding credit as an adjustment by consumers back towards their desired interest payments to income ratio. Changes in interest rates also result in wealth revaluation effects (primarily gilt and share prices in our model) which feed into consumption. However, it is difficult to judge either the magnitude or timing of these credit and wealth effects upon consumption. In our equations the credit effects are quite slow-moving and the impact upon consumption of a change in interest rates is still felt after 2 years.

Our consumption and wealth equations are backward-looking but it is probable that consumers are influenced by expectations. If this is the case it seems logical for consumers to smooth their consumption rather than immediately cut it or increase it in response to short-term revaluation effects etc. It may well be the case that some of the spending associated with future falls in interest rates has already happened (due to the expectation of lower interest rates in the not too distant future) and may explain why recent strong spending has persisted in the presence of high interest rates.

This forward-looking reasoning may also apply to wealth effects. We have therefore slightly dampened our model's forecast for the growth of real consumers' expenditure and real wealth (especially real share prices). An additional reason for a more muted response from consumption as interest-rate reductions revalue wealth is that the latter includes pension fund wealth which probably has no effect on spending.

Fixed investment and stockbuilding (tables 3 and 4) A fall in investment and/or destocking may explain why import volumes decreased in the final quarter of 1989. As capital goods imports volumes experienced one of the strongest declines it seems likely that a slow-down in investment may be well under way. The latest DTI company liquidity survey shows that the liquidity of non-manufacturing companies has continued to fall whereas the liquidity of manufacturers has increased. Combined with the evidence above this may suggest that manufacturing investment in particular is being curtailed as the company sector begins its adjustment back towards financial balance. The January CBI Industrial Trends Survey confirms that manufacturing investment intentions have weakened with the largest balance of companies since October 1982 expecting to authorise less capital expenditure this year than last year. Our manufacturing investment equation is forward-looking and also indicates that investment in this sector may already be responding to the prospect of lower output and capacity utilisation this year. We therefore expect a small cyclical fall in manufacturing investment this year with a limited recovery in 1991 as output improves and our assumed interest-rate declines reduce the cost of capital. All of the other investment categories, with the exception of a more buoyant oil sector, follow a broadly similar pattern probably resulting in a total investment fall of around 2 3/4 per cent this year rebounding to about 4 percent growth next year. [Tabular Data Ommited]

As stockbuilding is quite import-intensive it may be that the improved trade figures are also indicative of de-stocking. However, the stockbuilding data is unclear as the official statistics now attribute part of the `statistical discrepancy' (the difference between expenditure and output measures of GDP) to stockbuilding. Official data actually indicate a recent build-up of stocks - unintended stock accumulation usually occurs soon after output reaches a peak - which, if correct, one would now expect to be reversed. The CBI surveys agree that stocks are high in both the distributive trades and manufacturing industries and expect significant de-stocking during this year. Although we still believe substantial de-stocking will occur this year, we now expect it on a smaller scale compared to our November forecast as we now assume interest rates fall further. In our stockbuilding equations this works via a cost of stock-holding term (that is, the opportunity cost) which is roughly equivalent to the real interest rate. As these equations are forward-looking, expected interest-rate falls and a future output recovery restrain de-stocking from becoming recessionary. However, stockbuilding in the distribution sector seems more sensitive to company liquidity constraints and we expect this sector to experience the largest de-stocking as companies attempt to restore more sustainable liquidity ratios.

Balance of payments (table 5) Provisional figures show that the current balance deficit for 1989 was just over 20 billion [pounds]. Although this is a further 5 1/2 billion [pounds] deterioration compared to the 1988 deficit most of the decline occured in the oil and invisibles balances. The non-oil visible balance showed very little deterioration over the year and the better than expected trade figures for the final quarter of 1989 were largely due to an actual improvement in the UK's visible trade balance. Our poor performance in oil trade was mostly due to temporary production problems and a recovery is expected this year. The main adverse influence upon invisible earnings seems to be IPD which has declined partly because the relative rate of return - for both profit and interest related assets - has shifted in favour of overseas owners of UK assets.

Exports of manufactures have continued to rise rapidly, especially finished manufactures such as motor cars and capital goods. This good performance may be related to a relaxation in domestic demand pressure allowing manufacturers to switch to the export market. It seems that UK exporters of manufactured goods have adapted to the changing composition of world manufacturing demand, and, helped by foreign direct investment, have improved our underlying non-price competitiveness. Although we include a term for the non-price effect, our equation for manufacturing exports - which includes a unit elasticity for world trade - is underpredicting by a growing amount. Simultaneously, oil - the largest component of non-manufacturing exports - seems to be recovering.

The last quarter for 1989 also showed a turnaround in imports with virtually all import volumes falling. One implication is that investment may have slowed down more than expected as capital goods imports declined markedly. Our imports of manufactures equation is overpredicting, which provides some more evidence that we may have too high a figure for last year's fourth quarter investment and/or stock-building.

This year we expect a substantial reduction in the current balance deficit. The encouraging import and export trends of the final quarter of last year should continue with the combination of low domestic demand and world trade growth of around 5.5 per cent. Our forecast of a fall in investment and substantial de-stocking has a considerable restraining influence upon imports as these categories are extremely import-intensive. In addition, a recovery in oil output should improve the trade balance by around 2 billion [pounds] this year. However, we see little further improvement in the current balance in 1991 when fairly strong import growth returns as a consequence of increased GDP growth. The IPD surplus begins to grow again as a result of the returns to previous large portfolio outflows and of lower interest payments on UK liabilities.

This forecast starts from a higher exchange rate and incorporates a smaller sterling depreciation compared to our November Review. This partly explains our more optimistic view, since the November Review forecast for the 1990 current account included a J-curve effect from exchange-rate depreciation. Conversely, this forecast actually includes a beneficial terms of trade effect for this year partly due to the slow growth of some basic materials world prices. [Tabular Data Omitted]

Output and the labour market (tables 6 and 7) Manufacturing output seems to have been flat since the beginning of last year with output of the consumer goods industries experiencing a more rapid deceleration than the investment and intermediate goods sectors. However, there will be year-on-year growth for 1989 in manufacturing due to the acceleration in growth through 1988. The energy sector is still registering low levels of output and its recovery may have been slowed by the relatively warm winter weather Employment in manufacturing has been falling for a while in response to stagnating output in this sector, but the latest data show that towards the end of last year service sector employment was still growing. This suggests that the tightening liquidity position of manufacturing mentioned earlier may also be affecting employment in this sector.

We are forecasting a small decline in manufacturing output this year accompanied by a deceleration in output growth of the construction and distribution sectors. Our expectation of declines in investment and stockbuilding obviously play a part in this output slow-down. Conversely, the oil sector should experience an output recovery, and business services output may remain robust as exports of services continue to grow. Next year output should grow in all sectors - especially business services and distribution with manufacturing still a little subdued - with total non-oil GDP increasing by more than 2 per cent.

Data to the end of last year show that total unemployment is still falling but at a much slower pace than previously. However, this official unemployment count is based upon the `claimant status' of those not working. This makes forecasting unemployment difficult as attention has to be paid to the eligibility criterion for claiming unemployment benefit as well as sectoral employment flows. For example, as non-working females are less likely to be claiming benefit than non-working males a given increase in employment in a sector such as distribution, which employs a large number of females, would reduce unemployment to a lesser extent than an equivalent increase in manufacturing employment. Under this system married women not eligible to claim benefit who take part-time employment actually increase unemployment in the long run, as they increase the numbers eligible for benefit. Anomalies such as these between those officially unemployed and those seeking work help explain why we find that unemployment is a poor indicator of excess supply in the labour market in terms of explaining earnings.

As our forecast sees the female employment ratio increase as business service employment continues to grow and manufacturing declines, ceteris paribus this results in more unemployment compared to an alternative sectoral output mix. Conversely, our measure of excess supply in the labour market - the proportion of the population of working age that is not employed - is not influenced by changes in the composition of output, and earnings are unaffected. Given the slow-down in output and its assumed industrial composition we expect unemployment to begin rising in the middle of this year as employers begin shedding labour. By the end of 1991 unemployment could be back to around 1.75 million. [Tabular Data Ommited]

Public sector finance (table 10) The public sector debt repayment seems to be heading for a lower than expected surplus for the 1989/90 financial year. Unless corporation tax receipts are remarkably high in January it seems that the surplus will fall substantially short of the Autumn statement forecast of 12.5 billion [pounds]. Some higher spending - indicated by increased local authority borrowing perhaps connected with the costs of implementing the poll tax - helped depress the surplus at the end of the last calendar year. Higher inflation should have raised tax revenues although it should also increase public sector pay and costs. Revenues may soon begin a decelaration as employment responds to slower output growth but the path of earnings will also have an important influence. We have now revised downwards our forecast of the public sector surplus for the years 89/90 and 90/91 to 7 1/2 billion [pounds] and 12 billion [pounds] respectively.

Our forecast of a substantial decline in interest rates associated with UK ERM membership implies that a tight fiscal policy is necessary to control demand pressures. Fiscal restraint is especially important given our expected continuation of a current account deficit combined with the possibility of further falls in the personal sector savings ratio. This implies no tax cuts in the near future and the continuation of public sector surpluses. [Tabular Data Omitted]

THE MEDIUM-TERM FORECAST AND THE NAIRU

The non-accelerating inflation rate of unemployment (NAIRU) is the level of demand consistent with a stable rate of inflation. Short-run deviations of demand away from the NAIRU will result in either falling or rising inflation. Government policy can change both the NAIRU and the path or speed of adjustment towards the NAIRU. But only at the demand be in equilibrium.

The National Institute's economic model has a NAIRU or, strictly, a sustainable level of activity. We can express the model's wage and price equations in terms of a `target real wage' and a `feasible real wage' (see Joyce and Wren-Lewis (1989). (1) The former represents the desired real wage of employees and the latter represents the inverse of the desired mark-up of firms. Stable inflation takes place at the NAIRU when the level of demand is such that the `target real wage' and the `feasible real wage' are equalised. If these differ then inflation will decelerate/accelerate via the price-wage nexus until the level of demand is brought into line with the NAIRU.

If we impose the model's `target real wage' to be equal to the `feasible real wage' we can arrive at the following expression for the model's NAIRU:

.3938 In CU - .7524 PNWR = ......

The left hand side is the level of activity consistent with stable inflation. As the price and wage equations use different measures of activity (capacity utilisation and proportion of the population not working respectively) we have a weighted combination of both goods market and labour market utilisation terms. Therefore, we should make it clear that the following discussion centres around a non-accelerating rate of utilisation which is different to the normal NAIRU which refers only to unemployment. We will refer to our measure of weighted utilisation as the sustainable level of utilisation (SLU). The RHS variables are the determinants of the model's SLU and consist of factors which drive a wedge between wage and price inflation. In our model these factors are the real exchange rate, the cost of stockholding, taxes and an industrial mismatch term.

According to our model a rise in the UK's real exchange rate results in an increase in the sustainable level of utilisation. This occurs because a higher real exchange rate reduces the import costs of the firm. This implies an increase in the `feasible real wage' which allows a lower rate of unemployment and/or higher rate of capacity utilisation consistent with a stable inflation rate. The path of the SLU in the forecast is largely dependent upon the behaviour of the real exchange rate.

The sustainable level of utilisation (SLU) and the actual level of weighted utilisation (ALU) for the forecast are shown below.

At the end of 1989 actual utilisation is above the SLU and the UK economy is experiencing accelerating inflation. In response to this inflationary situation actual utilisation declines through 1990 and then subsequently grows steadily throughout the rest of the forecast period. Sustainable utilisation actually increases at first until the beginning of 1992 and then falls until recovering in the late 1990s. Our forecast for consumer price inflation is therefore characterised by decelerating inflation in the early 1990s (as actual utilisation is below the SLU) followed by accelerating inflation in the late 1990s.

As our forecast assumes no tax cuts the future SLU is largely determined by the real exchange rate and the cost of stockholding. The rise in the SLU through this year is partly caused by an unsustainable increase in the real exchange rate. Later on the depreciation of sterling - in line with our `Italian-type' membership of the EMS - brings falls in the real exchange rate and the SLU begins steadily to decline.

Real wealth and liquidity effects explain the slow-down and subsequent recovery in actual utilisation over this period. The `mix' of utilisation between the labour and goods market is disappointing for the former category; unemployment continues to rise until 1993 even though actual utilisation recovers two years previously. This is due to the movements of the relative price of capital and labour favouring an increase in the capital to labour ratio.

After reaching fairly low levels of consumer price inflation (below 3 per cent) around the mid-1990s our SLU calculations indicate that accelerating inflation begins to occur. This is a result of actual utilisation reaching a higher level than the SLU. However, the gap between these does not widen as the NAIRU itself grows in the late 1990s. This seems to be partly the result of our forecast growth for world trade which increases over the same period. This would boost UK exports and an increase in the real exchange rate (thus increasing the SLU) would be consistent with current balance equilibrium at the end of the forecast period.

APPENDIX A. MAJOR INTERVENTIONS IN THE FORECAST

Forecasts produced with a behavioural model reflect a combination of the properties of the model itself, judgemental intervention, and `exogenous' variable projections. In the case of the Institute's domestic forecast this last category comprises almost exclusivelyworld variables using our world model, GEM, and policy judgements. In the past, a cynical view has been that forecasts owed rather more to the forecasters than the model, with judgemental intervention being extensive. In the last few years this has not been true of the Institute's forecasts, which has been heavily model based. This appendix outlines the interventions (residuals) in this forecast sector by sector. A printout showing all the residuals in all equations is available on request.

Consumption

Small negative residuals have been added to total consumption and the consumer credit equations. This reflects both the latest information suggesting a slowdown in consumption and credit and also a view that consumers are already responding to expectations of a fall in interest rates later this year and through next year and hence there will be slower growth in consumption and credit than the model would predict over this period.

Investment

There is a small negative residual in 1990 for manufacturing reflecting the pessimistic CBI Survey evidence. Distribution and Business services investment contain growing negative residuals, which offset positive time trends in the estimated equations. A small positive residual on `other' investment reflects projected investment in the water and electricity industries.

Stockbuilding

The CSOs national accounts adjustment for stockbuilding is allocated across the three sectors in line with total stocks held in these sectors in 1988. This addition, plus possible unintended stockbuilding last year has resulted in positive residuals in our distribution stockbuilding equations. As a result we have phased out a positive residual in distributionwhich continues with growing negative residuals after 1991 to prevent stock output ratios rising in the medium term as companies financial position improves. A similar pattern exists for the `other' sector but starting from zero. For manufacturing a small flat negative residual is imposed as thecompany liquidity system does not feed in here.

Trade volumes

The average residual for the volume of manufactured exports for the past yearis about 5 per cent and this is projected flat for the whole forecast period and a smaller negative residual is placed on manufactured imports. Small negative residuals on the price of both imports and exports of manufactures are phased out by 1993. There are small negative flat residuals on the volumes of exports of food and basics and services and imports of other non-manufacturesand services. There are also interventions on the prices of imports of food and basics reducing the impact of the recent metals price fall from our world index as they are a small proportion of UK imports and exports.

Output, employment and unemployment

There is a flat positive residual on manufacturing output which adds 3 1/2 per cent to its level and a growing negative residual on `other' (predominantly public) services. This raises growth in the `rest' category. A small negative residual in construction reflecting the lownumber of housing starts currently is phased out by 1991. For employment there is a growing negative residual for `other services' which induces some productivity in this sector and a compensating positive residual in the `rest' category. This has a net effect of reducing mismatch and hence wage growth in the long run. A decreasing negative residual on unemployment in 1990/91 reflects changes in eligibility criteria for the claimant unemployed.

Exchange rate

This is kept exogenous in the forecast but at levels designed to reflect the theory behind the model equation. Movements in the exchange rate prior to entry into the EMS inlate 1990 reflect the `open arbitrage' condition with a risk premium. After entry into the EMS sterling behaves like the Italian lira with a 2 per cent depreciation against the German mark. Our world model, GEM, then calculates the UK effective index and dollar rate compatible with the world forecast. The current account by the end of the model run (1999) is near balance.

Prices and wages

Consumer prices have a small positive residual reflecting pricing formulae for water companies after privatisation and the `fattening up' of electricity prior to privatisation. This increases the underlying NAIRU in the model marginally as it represents increased monopolistic power in the economy. The other key price equation, wholesale prices, has a zero residual.There is intervention on wages in 1990 reflecting information on the current wages round. This gives positive residuals which are phased out over 1991/92.

There are other interventions in the price of investment goods (a flat small negative residual) and a smoothing of the price of new houses reflecting again anticipation of interest-rate cuts later in 1990 causing people to postpone house sales. This reduces the house price fall in 1990 and reduces the price increases in 1991 and 1992. Share prices have been exogenised with an imposed path reflecting the long-run returns on gifts.

APPENDIX B. WAGE DEVELOPMENTS IN THE PRIVATE AND PUBLIC SECTOR

Media and governmental attention is once more focused on pay settlements bothin the private and public sectors with the Ford settlement of 10.2 per cent seen as influential in the private sector and the ambulance workers holding out for a settlement of 8.5-9 per cent. Yet as settlements have steadily risen over the last eighteen months the Department of Employment's calculated average earnings index has shown no clear trend. Two arguments have been offered for this current settlements, so their index will lag by around 6 months; secondly wage drift above settlements will have been eroded as productivity and output growth have slowed down.

Table B1 shows that while the second argument may have some validity, the first has little. Settlements, averaged over the last year to match the Department of Employment index in terms of the time period covered, were steadily rising through 1989, while the Department of Employment index has only shown some upward pressure in the last three months. The productivity slowdown is shown in column 3, with whole economy productivity slowing from around 3 per cent in 1987 to just 1.1 per cent in 1989Q3 when allowance is made for the disruption of oil production. As much of the `wage drift' on top of settlements is related to productivity, drift is likely to have turned down. The other major source of wage drift is overtime working which, as table B1 shows, has declined a little over the last year but is still above the level of early 1988.

So has the 1 1/2 per cent rise in settlements in 1990 been mostly offset bydeclines in wage drift? The answer is `only partially' if CSO data on income are to be believed. For wages and salaries per head calculated from the national accounts show average earnings as rising steadily through 1988 and 1989, with `wage drift' peaking in late 1988 and early 1989 at around 3 1/2 per cent on top of settlements and as declining to 3 per cent in 1989Q3, and probably a little further since then. If this measure of average earnings is to be believed rather than the Department of Employment index then a far more consistent picture emerges, with rising settlements and lower productivity causing a steadily increasing growth rate of unit labour costs through 1989. Indeed the picture is of a classic surge in inflationary pressure at the end of a period of rapid expansion. The labour market is still tight when the product market is stagnant, and so although real wages growth actually slows, firms' costs rise as the mix of wages shifts from productivity related `wage drift' to larger settlements.

The prospects in the near future are no brighter, as non-oil output is likely to be flat in the next six months further reducing `wage drift'; and as settlements are still going to be higher than their equivalents the year before, even if they peak at the current levels of 8.5 to 9 per cent. So while the growth of wages and salaries per head is likely to peak at 10 - 10.5 per cent, unit labour costs are set to continue to accelerate till the summer, when settlements ease as firms shed jobs, and output begins to grow again. Hence in the forecast the growth of wages and salaries per head is assumed to fall slightly in the first six months of 1990 but unit labour costs continue to rise a little. In the latter half of this year settlements will ease a little, and productivity begins to accelerate; hence average earnings will grow at a constant rate fora while at the end of this year as `wage drift' responds to the improving product market and settlements ease. We expect that this rate will be around 8 1/2 per cent.

Direct government influence over wages is limited to the public sector. Even in the public sector, influence over the nationalised industries is at most indirect and the Police and Fire services have an indexing formula which determines pay automatically. Foster, Henry and Trinder (1984) demonstrated that public sector pay in the long term kept broadly in line with the private sector but would fall behind for a sustained period and then havea period of catch-up as in the Clegg awards in 1980.

Trinder and Biswas (1985) and Trinder (1988) showed how, since the Clegg awards, public sector pay had fallen behind the private sector and showed no signs of catching up. In the last two years it has been argued that the public sector is a cause of the current inflationary pressure, because there is little or no productivity growth to pay for public sector wage settlements Table B2 updates the earlier work done by Trinder to include the last two years. The overall picture is much as before with public sector real wage growth averaging 2 per cent since Clegg, but the speed-up of inflation has meant that real wage growth has been below average in the last two years, although the private sector, has done considerably better. Nevertheless, over the last two years some groups within the public sector have been able to begin the process of catching up with the private sector, or at least to prevent further relative deterioration. The police have consistently done well with their indexation formula and the large nursing settlement in 1988 has restored their 1981 position relative to private sector pay levels. In the last year local authority employees, manual and 'non-manual, and civil servants have kept pace with the private sector while teachers, the armed forces and NHS auxiliaries have lost further ground.

The erosion of relative public sector pay is causing problems for the government as the recruitment and retention of staff is becoming increasingly difficult for several key groups. One way of easing the problems at a time when control of wages and inflation as well as public expenditure is particularly important is for phased settlements under which awards are implemented in steps through the year. This reduces expenditure in the first year but does offer a higher base for the next settlement and the prospect of improving wages through the year. This formula was used for civil servants in the 1989/90 financial year and is now proposed for 1.3 million public sector workers, one-third of the total, in teaching, the health service and the armed forces. Most will receive 7 per cent in April and an additional stage bringing the average up to 10 per cent in the following January. With private sector earnings growth likely to peak in the later half of the year this could mean that 1990 will, like 1980, be a year of public sector wage `catch up'. [Tabular Data Omitted]

REFERENCES

Foster, N., Henry, S.G.B. and Trinder, C. (1984), 'Public and private sector pay: a partly disaggregated study', National Institute Economic Review, no. 107. Trinder, C. and Biswas, R. (1985), 'Public services pay in the 1980s', National Institute Economic Review, no. 112 Trinder, C. (1988), Appendix to Home Economy Chapter, National Institute Economic Review, no. 124.

APPENDIX C. THE CAPITAL ACCOUNT OF THE BALANCE OF PAYMENTS

This Appendix contains a discussion of a number of factors underlying our short and medium term forecasts for the capital account of the balance of payments. Developments over the 1980s are describedelsewhere in the Review. As shown in the summary table our chosen measure of imbalance, the 'basic balance', is projected to remain high by historical standards in 1990 and 1991 at levels equal to 6.2 and 3.9 per cent of GDP. While these deficits are much reduced from the recent level, suggesting that exchange market pressures on sterling may have peaked last year, the possibility of further financing problems arising in 1990 cannot be entirely excluded. However this is not reflected in our central forecast due to the projected entry into the EMS and the consequent fall in the risk premium on sterling. In the medium term the basic balance continues to show a deficit although the forecast levels of net banking sector inflows are only small proportion of world wealth and should therefore require only a small interest rate premium.

The remainder of this appendix concentrates upon the prospects for portfolio and direct investment. Outflows of non-bank portfolio investment amounted to 21.6 billion [pounds] in the first three-quarters of 1989 and we are forecasting an outflow of approximately 29 billion [pounds] over the year as a whole, implying a net outflow of some 24.7 billion [pounds]. This is forecast to contine on into 1990, although net outflows are projected to decline somewhat thereafter, returning to the levels experienced in the 1980s.

Subsequent to the last forecast, figures for investment by the major non-bank financial institution in the third quarter of 1989 have been released. These confirm trends apparent in the first half of the year, with a series of cash finance takeover bids amounting to some 4 billion [pounds] leaving these institutions as net sellers of domestic equities for the first time since 1974. With the continued Public Sector Debt Repayment reducing the stock of outstanding government bonds and the expectation of currency gains from potential exchange-rate movements, much of the third quarter's net investment of 7.7 billion [pounds] went into overseas securities.

Such portfolio outflows may be expected to persist for some time in the absence of an expansion in the UK bond market. Changing demographic patterns mean that pension funds in particular are becoming more mature with a marked increased in current liabilities. This implies an enhanced demand for gifts and other assets whose terminal capital value is certain in order to match income with known short-term financial payments. With the public sector forecast to remain in surplus and the domestic corporate bond market expanding slowly, international bonds may well remain an attractive investment, despite the associated exchange rate risk. In any case this risk can be hedged for at least a year or two ahead.

Revaluations from changes in the prices of assets held by institutions, along with a continued inflow of contractually-related savings mean that we are forecasting that total institutional assets will grow at an average annual compound rate of 11 per cent up to the end of 1995. Thus there is plenty of scope for the financing of additional overseas investment.

Net outflows of portfolio investment are however forecast to be progressively reduced with the assumed entry into the EMS reducing the risk premium on sterling and the relative strength of domestic equity prices attracting overseas investors. In the medium term further privatisations, notably electricity, and new rights issues encouraged by the strength of the equity market might be expected to increase the number of suitable domestic investments. (One important caveat concerns developments in Eastern Europe. As made clear in the world economy forecast, we have not attempted to take account of potential changes in our forecast of European equity prices).

A further source of uncertainty arises from the continued existence of a large sectoral balancing item. This has been projected to fall at a faster rate than in our previous forecast. In part this reflects an assumed improvement in the quality of the accounts following the recent introduction of a new reporting system for securities dealers. These changes have already resulted in an upward revision to past foreign investment in UK corporate bonds.

Our short-term projections for direct investment flows reflect both recent trends in cross-border activity and the outlook for corporate profits. The latest data for cross-border acquisitions indicates a marked increase in both the number and value of EC takeovers within the UK. For the first three quarters of 1989 there were 47 such acquisitions worth nearly 2 [pounds] billion, with a number of reported deals, including the bid for Morgan Grenfell by the Deutsche Bank, yet to show in the figures. Our equation for inward investment does not capture any '1992' type effects as it runs off current activity and current costs relative to those in foreign markets. Thus we have made a residual adjustment to the equation as we assume that EC-based deals will persist in the run up to 1992, although they may taper off somewhat as companies increasingly focus upon Eastern Europe.

This adjustment helps to bring the direct investment balance back to zero in 1990, although outflows are forecast to fall somewhat due to the lack of growth in corporate profitability. In the medium term there are once again net outflows of direct investment, with a projected decline in real world unit labour costs relative to those of the UK, and investment finance provided by the recovery in profitability. [Tabular Data Omitted] (1) Joyce, M. and Wren-Lewis, S., (1989) `Does the NAIRU matter?' NIESR discussion paper no. 154.