The home economy.

Britton, Andrew ; Gregg, Paul ; Joyce, Michael 等

CHAPTER 1. THE HOME ECONOMY

PART 1 RECENT DEVELOPMENTS AND 'POSSIBLE FUTURES'

The Treasury and Civil Service Committee of the House of Commons said in their Report on the 1989 Budget that 'the Chancellor was on a tightrope', meaning that the economy faced dangers of recession on the one hand and of accelerating inflation on the other. In our central forecasts we show a long period of slow growth rather than a recession, and the rate of inflation measured by the retail price index slows down very gradually from its current rate of just under 8 per cent. On this occasion, however, we give special prominence to two variants on the central forecast, one with domestic demand continuing to grow quite fast despite the high level of interest rates and the other with an abrupt fall in demand in response to the tightening of policy last year.

The range of possible forecasts is always very wide, considerably wider than the range of most forecasts actually published. For example the rate of growth of output from 1989 to 1990 is expected at the present time by all eight leading City forecasters to be in the range 1.7 to 3.2 per cent; the rate of inflation at the end of 1990 is expected by every one of that group to lie in the range 4.0 to 5.7 per cent. Yet plausible variants on our main forecasts produce outcomes above or below those ranges. It is useful, on occasion, to draw attention to these 'possible futures' and the risks or opportunities they imply.

Recent Developments

The 1989 Budget was described as 'cautious', because it left the public sector in large surplus. As expected taxes and National Insurance contributions were cut: the direct effects on revenue of the Budget measures were under 2 pound sterling billion in 1989/90 and nearly 3-1/2 pound sterling billion in 1990/91. Compared with a 'no-change' Budget this adds about 1/2 per cent to GDP and about [pound sterling]1 billion to the balance of payments deficit in 1990. Compared with the assumptions about the Budget in our February Review the changes are trivial. The decision not to uprate specific duties on drink and tobacco reduces the retail price index by just under 1/2 per cent.

Although world interest rates risen the UK authorities have been able tohold rates in this country unchanged, the banks' base rate remaining at 13 per cent since November. The effective exchange rate has slipped back a little, reversing the rise in the first quarter.

The trade balance in the first quarter was significantly worse than we expected. Import volumes (of capital goods in particular) continued to rise, suggesting that the slowing down of domestic demand has not been abrupt. The first estimate of consumer spending was also a little higher than we had been expecting. The unusually mild weather may mean that building investment was also above expectation. Total GDP however will be held back by the low level of oil output following a series of accidents in the North Sea.

Several indicators point to a high rate of inflation. Wage settlements in particular are clearly rising, being now about 7-1/2 per cent. This may be associated with an increase in labour disputes and strike action. Wholesale prices in the first quarter were a full percentage point higher than we had expected.

There has been a further round of data revisions to the national accounts reducing some of the discrepancies. The effect has been to raise the growth of the expenditure estimate of GDP but to reduce the growth of the output estimate. Personal sector income has been revised up substantially and the savings ratio has been revised up a little.

The main data change however is to employment, following the publication of the 1988 Labour Force Survey. This has reduced considerably the growth of productivity in recent years. For 1988 the figure for the whole economy has been reduced from 3.3 per cent to only 1.4 per cent, putting the recent behaviour of the labour market in a very different light. The new data also do much to solve the 'mystery' of the fall in unemployment, as Paul Gregg explains in the note at the end of the chapter.

The Central Forecast

Apart from the 1989 Budget we have retained the sam fiscal policy assumptions as we used in February: taxes are cut by pound sterling 2 billiokn in each Budget throughout the forecast period.

The exchange-rate forecast is again determined by the principle of 'consistent' expectations. In other words the profile for the current account of the balance of payments in our forecasts up to the end of the model solution period (1997) together with the assumed path of interest rates is deemed to be consistent with the level of the exchange rate actually obtaining in the current quarter. If the market's expectations are the same as ours, the anticipated yield from holding sterling assets and holding foreign currency assets will be similar apart from a small premium in favour of sterling necessary to induce net capital inflows to offset the current account deficit. Interest rates are now higher in the UK than elsewhere mainly because the markets anticipate, as we do also, a tendency for sterling to fall in the medium term.

In Februarry we were assuming that interest rates would be held at 13 per cent for several years. Now, for the main case, in view of the more rapid rate of inflation in prospect, we assume that rates will rise to about 14 per cent next year, and will not fall back for two years after that. Our forecast for GDP growth this year is virtually unchanged at just under 2-1/2 per cent, which is also broadly in line with the Treasury's forecast at Budget time. Consumer spending is expected to slow down substantially compared with 1988 and by the latter half of the year fixed investment may be falling. Our forecast for growth next year has been raised significantly as compared with the forecasts in the February Review. One reason for the change is a higher forecast of oil output consistent with the new Department of Energy Brown book. Another reason is a higher forecast of the growth rate of earnings which maintains consumer real incomes, but which also adds to inflation. The outlook beyond 1990 is still for several years of slow growth, followed by a return to about 2-1/2 per cent growth in the more distant future.

The new employment data have led us to reassess our forecasts of productivity growth. We now expect employment to go on rising throughout this year and most of next. Corresponding to this is a rather more optimistic view of the prospect for unemployment. This now levels off at about 1.7 million early next year and only rises slowly even in the medium term.

The prospect for the balance of payments remains one of very slow improvement. The deficit on the current account is now expected to be larger in 1989 than in 1988, this change mainly reflecting the size of the deficit in the first quarter and the interruptions to the flow of oil production. It remains our view that the current account must be brought back close to balance, but this may not be achieved until the mid-1990s. The rate of progress must depend mainly on the growth of domestic demand. The sustainable rate of growth is raised a little by rather more optimistic views of the level of oil production and of the growth rate of world trade (both of which improve our balance of payments at any given level of domestic activity).

The prospects for inflation have deteriorated significantly in the last few months. We have raised our forecast of the growth in the retail price index at the end of this year to about 6-1/2 per cent, compared with the Treasury forecast of 5-1/2 per cent. This difference may be explained mainly by a difference of view on the exchange rate and the interest rate. Looking ahead to 1990 the contrast is greater. We now foresee inflation staying in 1990 at about 6-1/2 per cent, whilst the Treasury expect a fall below 5 per cent. As well as interest rates and the exchange rate the difference probably turns on judgements about the relationship between earnings and productivity growth. At the present time we observe the combination of earnings speeding up and productivity slowing down. This is not unusual in the later stages of a cyclical boom. Experience suggests that it will be followed by a year or more of quite rapid inflation before the downturn in economic activity has its effect on the labour market. In our main forecast the slowdown in earnings growth comes in 1991 with a year or two of low price inflation after that as a consequence.

The Nature of the Variant Forecasts

This time last year most forecasters, ourselves included, were expecting the growth of domestic demand to slow down; in fact it continued unabated. We could be making a similar mistake again, or alternatively we could be over-reacting and under-estimating the extent to which demand is now being cut back. In the two variant forecasts we make symmetrical changes up and down to the level of consumer spending by adding or subtracting a 'residual adjustment' throughout the forecast period. The scale of that adjustment has no precise significance, but it is thought to be about the limits to which forecasters' 'judgement' might reasonably be pushed in the case of that particular equation. Initially the adjustment adds or subtracts only 1/2 per cent to the level of consumption, but this builds up to about 5 per cent thanks to the dynamic structure of the equation.

In the main forecast the current level of the exchange rate is deemed to be consistent with the projections of interest rates and the current balance. In the variants the exchange rate is assumed to respond immediately to the different outlook, making an appropriate step change up or down. In fact it would presumably take some time for the markets to change their vie of the underlying situation, so the 'jump' might in practice be more like a 'slide' or a 'surge', but this need not make much difference to the main features of the model solutions.

The scale of such exchange-rate changes would undoubtedly be moderated by the response of the authorities. No account is taken of exchange market intervention, but it is assumed that interest rates are adjusted up or down simultaneously with the change in sentiment in the markets. Fiscal policy is assumed to be the same in all cases, as is implied by recent policy statements.

The Higher Demand Variant

The effect of higher demand is to delay the transition to slower growth, not to remove the necessity for it. Output goes on rising above trend until 1991, with employment still increasing and unemployment falling even in 1990. The effect is a temporary one, however, and by 1994 the level of output is the same as it is in the main forecast.

Meanwhile the fall in the exchange rate, and the higher pressure of demand in the domestic economy, imply a rate of inflation just into double figures. The exact timing of this surge in inflation, and the extent of the exchange-rate fall, is of course highly uncertain. Nevertheless the possibility of inflation of this magnitude certainly cannot be ruled out.

The policy response raises interest rates to 16 per cent. No doubt a really vigorous response might result in yet higher rates, or the same level held for a longer period. In either case the fall in the exchange rate would be smaller and the inflationary consequences less serious. On the other hand there are both political and economic costs to raising interest rates, of which the authorities will be well aware. (As mentioned above, fiscal policy is assumed unchanged in both variants.)

The exchange-rate and interest-rate response are such as to bring the current account back on to its path towards equilibrium by about 1991. This is a feature of all the variants. Whatever the level of domestic demand and of interest rates the external constraint must eventually be a binding one.

The scale of the exchange-rate movement, with an initial drop of 10 per cent, is not exceptional or inherently implausible. The resulting improvement in relative price competitiveness is barely enough to reverse the loss experienced in 1988. In terms of the 'real' exchange rate, calculated using relative wholesale prices, the level after the initial drop is similar to the average of 1987.

The Lower Demand Variant

The second variant is simply a symmetrical change to consumer demand in the opposite direction. The results confirm the rough linearity of our macro-economic model. In reality, however, it is possible that there may be asymmetries both in the behaviour of the economy and in the responses of policymakers.

Assuming symmetry, the path of output shows an earlier slowdown, especially in 1990, followed by an earlier recovery around 1993. This can just be called a recession, since GDP is lower at the end of next year than it is in the third quarter. Manufacturing output in particular falls quite sharply.

The implication would be a turnaround in unemployment about the end of this year, followed by a significant rise, reaching 2 million again some time in 1991 or 1992. The exchange rate is assumed to 'jump' up by about 10 per cent as the prospects for inflation and the current balance improve. The authorities take the opportunity to cut interest rates to 11 per cent, but as the exchange rate resumes its fall and the balance of payments remains in large deficit they are brought back up again to 13 per cent by 1992. This is a very high level in real terms, implying that even in this variant monetary policy remains tight. Although the current balance in this variant improves steadily, it is still necessary to make sterling assets attractive to investors and induce capital inflows on a substantial scale.

The most striking feature of this variant is the low rate of inflation, amounting almost to price stability by 1992. It is here that questions of asymmetry most obviously arise. Is it possible for example that the growth of average earnings could fall as low as 1-1/2 per cent in 1992? A more plausible view might be that an inflation rate of about 2 or 3 per cent could be sustained for a long period in the early 1990s.

Policy responses might also be asymmetrical, reducing interest rates more and sooner. It is perhaps unlikely that the authorities would really press for price stability if the cost was interest rates still at 13 per cent in 1992 and unemployment rising. That would tend to weaken the exchange rate (even initially, assuming 'consistent' expectation). That apart, the lower demand variant also seems to offer a plausible view of the future, although it is again in several respects outside the range of published forecasts.

PART 2 THE MAIN CASE IN DETAIL

Forecasts of Expenditure and Output (table 1)

The growth of output last year is now estimated at under 4-1/2 per cent, lower than earlier estimates had suggested--a reminder that revisions by the Central Statistical Office are not always in the same direction. The apparent pause in the first quarter of this year is accounted for by interruptions to oil production; the underlying growth of output has not yet lost all its momentum.

Even part from oil, the trade balance in the first quarter was dissappointing, and difficult to reconcile with a continuing good rise in manufacturing output. The 'residual' category of expenditure needed to reconcile the national accounts data is likely to remain large. We have settled on a level of 2 pound sterling billion a quarter as a reasonable level at which to extrapolate this residual. The implication is that, year-on-year, our forecast of GDP grows about 1/2 percentage point faster than the growth rate of its identified expenditure components. For 1990 the growth rates are virtually the same.

The profile of the main expenditure categories reflects the tightening of monetary policy in the course of last year. Consumer spending which has surged ahead for the past three years is forecast to slow down wuite abruptly. Fixed investment, having risen substantially last year, is now expected to show little further growth in the course of this year and to fall in 1990. The profile of stockbuilding is similar, positive this year, but becoming negative by the end of next year. This would be a profile of company spending not unlike that at the corresponding stage of previous cycles.

Subtracting exports from imports of goods and services at 1985 prices gives the net switch of resources from overseas, or the excess of domestic demand over domestic output. This rose by as much as 15 pound sterling billion (4.3 per cent of GDP) last year. This year we expect a more modest further increase, much of it accounted for by the fall in oil production. By 1990 resources will be switched back, but still on a modest scale, to improve the balance of payments.

Personal Income and Expenditure (table 2)

The growth of average earnings continues to run ahead of that predicted by our equation. Perhaps the best explanation would be in terms of regional mismatch and shortages of skilled labour. We described the disparity in regional unemployment rates in the February Review. The CBI Survey suggests that about a quarter of firms are experiencing skill shortages, and a few firms have difficulty in recruiting any kind of labour at all. The current rate of increase of average earnings is about 9-1/2 per cent; we foresee little change from that rate this year. By 1990 there should be a slight deceleration due to a reduction in average hours worked.

With employment only rising a little, and other personal incomes growing more slowly, total personal disposable income at current prices would be slowing Down over the forecast period, even if the growth of average earnings were not. As inflation accelerates, personal disposable income in real terms will almost come to half. Through 1988 RPDI rose by about 5 per cent, an unusually rapid pace by any standards. Over the next year, growth may be under 1 per cent a year, which is unusually slow; in 1990 we may see a fall.

The forecasts of consumer spending have been the area where the most serious errors have been made in recent years. This is also one area-where different forecasters now take substantially different views. Our model is based on the relationship of consumer spending with wealth, credit and liquidity as well as real incomes. (We are not convinced that demography or house prices play an important independent role.)

The data for personal incomes in the past have been revised up, so that the savings ratio last year is now rather larger, at about 4 per cent. Keeping a small positive residual on our consumption function, we still expect the savings ratio to fall this year, as it almost invariably does when real income growth slows down. Despite the past increases in interest rates we expect consumer credit in go on rising, if at a much slower pace.

In consequence we foresee total consumer spending continuing to rise over the short-term forecast period, although almost static during 1990. Within that total, spending on durable goods, which is more interest-rate-sensitive, should peak in the first half of this year and fall back during 1990.

Fixed Investment and Stockbuilding (tables 3 and 4)

The recorded levels of fixed investment in the latter half of last year have been revised up by some 8 per cent. The main change is to public sector investment and 'other' private investment (not shown in table 3). Carrying forward this higher level into 1989 and beyond raises total investment, although our view on the growth of both manufacturing and service investment is little changed.

In both sectors our forecast remains consistent with the DTI surveys, although both may now be a little out of date. These point to continuing growth at a brisk pace in the first half of this year (helped perhaps in the case of building by the mild first quarter). By the end of the year however we would expect the investment boom to be over, and to be followed by an actual fall year-on-year in 1990.

The housing market seems to hae responded already to the tightening of monetary policy last year. House prices are now static in the south of the country and few properties are changing hands. Starts and completions show no clear trend, but actual investment was falling in volume in the latter half of last year. We interpret this as the beginning of a downturn, due to continue through this year and next.

The figures for public investment are difficult to read, as they are affected by privatisation. Our forecasts make allowance for the transfer of electricity and water to private ownership, and assume that this makes their investment plans more subject to market forces. (In the context of the present forecast that means that their investment plans may be cut.)

STockbuilding by all sectors was high at the end of last year. This is a familiar experience at the top of the output cycle. As demand falls back this year, stockbuilding will at first serve to smooth the path of output. Later on, however, in the course of next year perhaps, stockbuilding itself will turn negative as firms seek to restore a normal relationship of stock levels with output or sales. These cyclical developments take place against the background of a continuing slow downward trend in stock to output ratios.

Exports and Imports of Goods and Services (table 5)

Our latest estimates of world trade have raised the growth rate for both 1987 and 1988 by a percentage point, to about 7 per cent a year. UK exports of manufacturers grew in volume a little faster than that in 1987, a little slower in 1988. On resent evidence the UK has been roughly holding its share of total trade despite a loss of relative price competitiveness (see also table 8). Our current model equation, which reflects average behavior over a longer period, has been under-predicting the growth of exports. We take account of this in our projection, as we did in February.

World trade growth in 1989 and 1990 is now forecast at 7 per cent and 5-3/4 per cent respectively, faster than previously expected. Our exchange-rate forecast implies that relative price competitiveness will be broadly maintained from now on, so the prospects for export volume growth would seem to be good. Against this, the CBI survey points to growing pessimism amongst exporters about the difficulties of competing internatinally. Despite these doubts we are forecasting an acceleration in the growth of manufactured exports this year, with some slowing down in 1990.

The forecast of total exports is complicated by the effects fo interruptions to oil production The assumed path for oil output is shown in table 6 below, with a sharp drop in the first half of the year, and a recovery to a more normal level by the fourth quarter. Exports of oil move roughly in proportion to output, and oil consumption is little affected; imports of oil can then be calculated as a residual item. Also subsumed within total exports are the UK's 'other' exports of goods, mainly food and basic materials. These have been depressed in 1987 and 1988, but we are hoping to see a recovery, encouraged by the high figure for food exports for the first quarter.

Imports of manufactures (especially capital goods) continued to rise rapidly in the first quarter, when we had been expecting to see a pause. Our equation cannot track the erratic quarter-to-quarter movements but has proved a good guide to last year as a whole. The forecast is made using the equation with 'zero residuals'. This indicates growth well in excess of final spending during 1989 (fourth quarter on fourth quarter) but a much more gradual profile through 1990, when capacity utilisation is lower.

As with exports, imports of food and basic materials do not seem to share in the rapid rise of trade in manufactures. In this case we have to project our equation with 'negative residuals'. Taking account also of trade in oil, total imports of goods will record another very large rise year-on-year in 1989, followed by a much smaller figure in 1990.

The upshot is that the visible balance, already over 20 billion pounds sterling in deficit last year, will be very slow to turn round--indeed there can be no guarantee that it will not get substantially worse before it shows a sustained improvement. Neither is there reason to expect the invisible balance to change greatly over the next two years.

Overall our view of the current balance remains much the same as it was in February. There has been a progressive move towards deeper deficit, exacerbated by the boom of last year. Even if domestic demand now grows very slowly, the path back to a sustainable external position will be difficult and protracted.

Output and the Labour Market (table 9)

In a forecast which builds up output by summing the categories of expenditure, the composition of output by industry can only be deduced indirectly. During last year manufacturing output rose about 7 per cent, although the sum of expenditure categories weighted by their manufacturing content rose by just 1/2 of one per cent. This is another of the discrepancies within the national accounts and it makes forecasting of the manufacturing sector especially difficult. We assume that the discrepancy does not widen any further. That implies, together with the forecasts of expenditure components, that manufacturing output will peak in the first half of this year.

The new employment data based on the 1988 Labour Force Survey are described in the note at the end of this chapter. In the light of the lower level of productivity now estimated we have looked again at the evidence for an acceleration in the underlying trend of productivity in the manufacturing sector. The results are described in the box below.

The results of this exercise confirm our view that the growth rate of productivity will slow down substantially as the growth of output slows down. In the manufacturing sector the sustainable average growth rate of output per head is probably about 3 per cent a year, compared with 6.7 per cent in 1987 and 5.6 per cent in 1988.

For the economy as a whole the corresponding figure may be about 2 or 2-1/2 per cent. Over the next year the figures for total output per head are distorted by the erratic path of oil production. Excluding North Sea oil, total productivity rose about 2 per cent in 1988 and is forecast to rise by 1 to 1-1/2 per cent in 1989, but by under 1 per cent in 1990.

Unemployment has again fallen rapidly in the first quarter (possibly helped by the weather). It is now much easier to understand the extent of the fall since 1986, although we would still attribute a significant role to Restart and similar measures designed to deter ineligible claimants. It still seems inevitable that the fall in unemployment will come to an end, given the profile of output in the central forecast. We are now more confident, however, that the fall can continue for the rest of this year. The number at which the claimant count turns up again now seems likely to be about 1.7 millions. (this compares with a 1979 trough of 1.05 millions on a consistent basis with the current count.)

Price Inflation (table 7)

Up to now the lack of any significant rise in import prices has held back inflation in this country. The deflator for imports of all goods and services actually fell year-on-year in 1988. Prices on world markets were still relatively depressed, and the exchange rate rose by as much as 6 per cent. Looking ahead to 1989 and beyond, this moderating influence will be removed.

From now on we expect the exchange rate to drift down (see table 8) whilst world inflation is accelerating (see Chapter 2). Oil prices in particular are sharply up in the first half of the year. Year on year the rise in import prices will not be great, since the level in the first quarter is still only just up on a year earlier, but by next year we would expect import prices to be rising at least as fast as those of domestic production.

Some acceleration also seems inevitable in the growth of unit labour costs; indeed it not seems that a speeding up was already under way last year. Average earnings in 1988 rose by 8.7 per cent, whole economy productivity (excluding the North Sea) by just 2 per cent. For 1989 the situation looks more serious again, wih average earnings growth over 9 per cent and non-oil output per head about 1-1/2 per cent.

The wholesale price index for manufactured output in the first quarter was 5-1/2 per cent up on a year earlier and 2 per cent up on the quarter. This is not significantly out of line with our equation, which takes account of the effect of capacity use on profit margins, although it is in fact rather higher than we were forecasting in February. Even though capacity use will now be falling wholesale price inflation should accelerate in response to until labour costs and to the prices of imports. By the end of the year the annual rate of change could be nearly 8 per cent. The consumer price index (CPIe is subject to similar influences and it too is likely to rise more rapidly as the year progresses, although in this case the Budget decision not to uprate specific duties has a moderating effect. The latest estimate, for the fourth quarter of last year, was already about 5-1/2 per cent; we expect about 6 to 6-1/2 per cent by the end of this year.

The path of the retail price index (RPI) would be similar but for the inclusion in it of mortgage interest rates. This accounts for the figure close to 8 per cent recorded in the first quarter of the year. By the end of 1989 the RPI should be more in line with the CPI at around 6-1/2 per cent.

In 1987 and again in 1988 the share of non-oil profits in gross domestic product rose substantially. This is not unusual at a time of rapid growth in output, when productivity is rising strongly and there is also scope for widening wholesale and retail margins. The share is now about 15 per cent, and in the circumstances of our main forecasts it is unlikely to rise much further.

The prospects for inflation beyond 1989 are obviously very uncertain, but the process of deceleration could well be a slow and difficult one. The key judgements on earnings, productivity and the exchange rate are described elsewhere in this chapter. Given those judgements, inflation as measured by wholesale prices will slow down a little during 1990, but the CPI could actually accelerate (because the moderating influence of the 1989 Budget will be over). The RPI could also accelerate if it is again necessary to raise interest rates next year.

Public Sector Finance (table 9)

The Chancellor in his Budget Speech this year reiterated his view that a balanced budget, or a zero PSBR, was the best guide to fiscal policy in the medium term. The projections published by the Treasury at the time suggested that the surplus now in prospect could be virtually eliminated by 1992-93 if taxes were cut by a total of 6 pond sterling billion in the interventing years. The basis for these calculations has not been published in sufficient detail for informed criticism, but they seem to understate the likely buoyancy of revenue.

For 1989/90 our forecasts of the public sector debt repayment are above the figures in the Budget Statement. For 1990/91, assuming tax cuts moreover of 2 pond sterling billion in next year's Budget, the surplus seems more likely to rise than to fall.

Part of the explanation of the difference in forecasts may lie in the assumptions about inflation. Higher inflation raises revenue at least in proportion; to the extent that cash control is effective, the result for spending is much less than proportionate. Hence a lower assumption about inflation would naturally produce a forecast of a smaller PSDR.

If we are right in predicting that the PSDR remains large, there will be pressure to raise the volume of public spending this year and subsequently. We assume that such pressure will be resisted, since the projections at current prices may already be over-run as a result of general price inflation.