The UK economy.

Kneller, Richard ; Riley, Rebecca ; Young, Garry 等

Section I. Recent developments and summary of the forecast

The economic outlook is sometimes strongly influenced by what people expect to happen. In the late summer of last year, prospects for the British economy worsened partly because people had become more pessimistic about the world economic outlook. Among other things, it was feared that this would lead to postponement of investment projects and extra precautionary saving which would dampen demand, weaken output and so justify the greater pessimism.

At the time, we felt that the lack of confidence seen in business surveys pointed to a gloomier outlook than was warranted. To a large extent, we were relatively more optimistic because we felt that monetary policy could be eased substantially to ensure that demand did not collapse. In fact, the cuts in interest rates over the past six months have been much sharper than we had expected. Base rates in the UK and the rest of the world are now over I per cent lower than we had expected them to be. The swift easing of monetary policy has played a key role in preventing the loss of confidence causing a severe economic downturn.

Now that confidence is returning, it is likely that the UK will narrowly avoid an actual recession in the current downswing of the cycle. Latest estimates suggest that output growth remained positive around the turn of the year when the risks of a decline were greatest. While nothing is guaranteed, growth is likely to increase throughout the rest of this year and into 2000.

It is perhaps a sign of renewed confidence that commentators are now returning to their earlier theme of the 'two-speed' economy. While the overall economy appears to have avoided a recession, this is not true of the manufacturing sector. Manufacturing output in the first quarter appears to have fallen for the third quarter in succession. The most obvious cause of this weakness is the strong pound. This has reduced the profitability of domestic firms who compete with those overseas on the basis of price and reduced the attractiveness of those who attempt to maintain their profitability. Accepting an estimate of around ??1.30 as a fair value of sterling, then the exchange rate has been overvalued since the end of 1996.

It is possible to argue that the strength of sterling over the past two years or more has been caused partly by unnecessarily high interest rates in 1997 and the first part of 1998. But for the most part the slowdown in activity over this period has been close to that which was needed to meet the government's inflation target. In general, the Monetary Policy Committee (MPC) has not been sensitive to the needs of particular sectors or regions in setting interest rates and rightly has taken account only of the needs of the overall economy. In this case, it has been an imbalance in macroeconomic policy, with taxes not high enough to allow lower interest rates, that has caused problems for manufacturing industry, rather than the overall policy stance.

The weakness of the euro since its launch at the beginning of the year has meant that the value of sterling is putting further pressure on the traded sector of the economy. Towards the end of April, sterling is trading at around ??1.50, above its old central rate in the ERM. It is likely that further reductions in UK interest rates would weaken the pound and so relieve some of this pressure. But having been reduced so sharply, it is now possible that further cuts in interest rates would threaten the inflation target. This depends importantly on how further cuts in interest rates would be interpreted by the foreign exchange markets and how this would affect the value of the pound. Put simply, if the pound were to stay at its current high levels, then there is ample scope for further cuts in interest rates. But if the pound were to fall to a level at which the traded sector could more easily compete internationally, then further interest rate reductions would pose a threat to the inflation target by the end of next year.

Expectations about the possible future course of the exchange rate are one of the key factors in determining the outlook for the economy more generally.

In producing this forecast, we have considered three different possible scenarios. In each case, the sterling exchange rate against the euro is fixed from the first quarter of 2003 when the UK is assumed to become a full member of the monetary union. UK interest rates are then set equal to those in the rest of the Euro Area. We consider two different entry rates for sterling. The first is a rate of ??1.33, equivalent to DM2.60. This is a rate broadly compatible with the competitiveness of the traded sector, only marginally above what we consider fair value. The second is a rate of ??1.42, equivalent to DM2.775. This is the rate implied by four-year interest rates on sterling and euro-denominated risk-free bonds. It is approximately equal to the rate in January when the traded sector was clearly struggling to compete internationally, indicating that sterling would be overvalued at that rate.

These assumptions about what might happen in 2003 are likely to affect the way in which the economy behaves in advance of this date. For example, if firms in the traded sector knew that the UK was going to join EMU at an uncompetitive rate, then they would take some action in advance to bring their costs under control. This in turn would affect the actions of wage bargainers who would find it harder to push up costs. This restraint would thereby make, the uncompetitive rate more manageable. The resulting downward pressure on inflation would allow interest rates to be lower than would otherwise be the case.

The first case we consider is where everyone expects the UK to join EMU in 2003 at the more competitive exchange rate and depreciate gradually to ??1.33. We assume that interest rates remain at 5 1/4 per cent until the middle of next year, after which they are changed according to the gap between actual inflation and its target rate.

In this case, inflationary pressure in the very short term is largely absent, consistent with recent UK evidence. The target measure of inflation (RPIX) falls through the remainder of this year to 1.8 per cent in the fourth quarter. But as growth picks up in 2000 there is an increase in inflation such that the inflation target is missed at the end of the year. Interest rates a re then raised in an effort to offset the upward pressure on prices.

Beyond the end of next year interest rates would have to be raised further in response to inflation persistently higher than target, finally reaching 7 1/2 per cent just prior to joining EMU.

A difficulty with this particular scenario is that it embodies a different view of the future from that currently implicit in financial market expectations. This is not a crucial objection as markets are often wrong, failing to notice when asset prices are out of line with their fundamentals. But it is clearly preferable in forecasting to take a view that is consistent with the expectations of those whose actions are likely to be important in determining outcomes.

The second case we consider is where everyone expects the UK to join EMU in 2003 at the less competitive exchange rate of ??1.42, and that interest rates broadly follow the path implicit in the current term structure of interest rates. At present, expectations in the short-term money markets indicate little change in interest rates over the coming twelve months, but the bond market suggests possible convergence to euro levels in 2003. Thus we assume that interest rates remain at 5 1/4 per cent until the first quarter of 2000 and then fall by 0.15 of a per cent every quarter until reaching 3.45 per cent, equal to the expected short-term euro rate, in the first quarter of 2003.

In this case, inflationary pressure remains subdued at least until 2003, despite interest rates being lower than in the first case we considered. There are two key reasons for this. First, the sterling exchange rate remains stronger throughout, putting downward pressure on prices as the competitiveness of the traded sector is squeezed. Second, the expectation that the exchange rate will be stronger reduces expectations of inflation over the next four years and so reduces inflation itself. Chart 1 shows the inflation path and the accompanying path of UK base rates under both scenarios.

The difference between these two cases emphasises the importance of expectations in affecting macroeconomic outcomes. In particular, inflation is easier to control when inflation expectations remain low and this can be achieved with lower interest rates than when inflation is expected to be higher. It is primarily for this reason that the Bank of England was so keen to establish its anti-inflation credibility when it first became independent and why the European Central Bank may be reluctant to cut interest rates further until it has built its reputation.

Thus actual inflation is partly determined by inflation expectations and these are in turn influenced by what people expect policymakers to do. This picture is complicated by policymakers also needing to make an assessment of what other people expect them to do.

It is in understanding the possible interaction between interest rates, the exchange rate and expectations wherein lies one of the main challenges in monetary management. The MPC's approach in constructing its inflation forecast is similar, but not identical to the second case discussed above. The February Inflation Report states that "in judging the path for sterling, conditional on unchanged UK nominal interest rates, the MPC has taken into account both interest rate differentials and risk considerations".

There are two critical points to make about this approach. First, there is an inconsistency in producing a forecast on the assumption of unchanged UK interest rates while simultaneously projecting the exchange rate from market interest differentials based on the expectation of falling interest rates. In current circumstances this procedure would lead to monetary policy being too loose. An assessment of inflation pressures based on higher interest rates but the same exchange rate path as the market assumes would imply tighter monetary conditions in future than are likely to occur in practice and hence an over-optimistic inflation forecast. Second, the exchange rate path implied by the expectations of the financial markets may be completely incorrect and, more importantly, different from that which underlies the expectations of those involved in determining prices in the real economy.

The importance of this second point is brought out in the two cases we considered above. Forecasting on the basis of the second case, similar to the MPC's approach, would lead to an excessively loose monetary policy if expectations are truly as set out in the first case. By contrast, forecasting on the basis of the first case when market expectations build in low inflation expectations would lead to an excessively tight monetary policy.

Clearly this is an area where judgement is required, as purely mechanical forecasting rules might be misleading. While this may appear to be a fairly esoteric point, it is not possible to ignore the importance of inflation expectations in determining actual inflation or avoid making an assumption about how they are formed.

The general view is probably that the sterling exchange rate is strongly overvalued at present and that if the UK is to join EMU in 2003, as the government seems to want, the pound will need to depreciate faster than the markets are currently expecting. We have built this as an assumption into our short-term forecast where expectations are determined consistently with the UK joining EMU at the beginning of 2003 at an exchange rate of ??1.33. But as this date becomes nearer, we assume that people begin to realise that the entry rate will in fact be higher. So, from the beginning of 2001, we have assumed that expectations of the entry rate are gradually revised upwards from ??1.33 to ??1.42.

None of this affects the short-term forecast, but it does affect the behaviour of the economy in the early part of the next decade. To see this, Chart 2 shows the changing pattern of interest rates and exchange rates as people learn that sterling will join EMU at a high rate.

This shows a trade-off between higher interest rates and a higher exchange rate which prevents the overvalued [TABULAR DATA FOR TABLE 1 OMITTED] exchange rate doing too much damage to the real economy. It does reduce output relative to what it would otherwise have been, particularly so in the manufacturing sector, as shown in Chart 3.

The difference in output in the two cases gives some indication of the cost of joining EMU at an overvalued exchange rate. The cost would be much higher if interest rates were not adjusted to offset the deflationary implications of the high exchange rate. It would also be higher if price setters failed to recognise the likelihood of entry at the high rate. This perhaps illustrates the importance of not suddenly joining EMU at an overvalued rate, because that would not allow the possibility of any interest rate adjustment.

Monetary conditions (table 1)

In the short term, we have assumed that interest rates remain at 5 1/4 per cent until the end of next year. Our earlier discussion emphasises the difficulties in judging the outlook for inflationary pressure and how the MPC will react to the available information. But with evidence of generally improving confidence and a robust labour market, it is now unlikely that interest rates will be reduced much further over the coming year.

The main factor that could justify further cuts in interest rates is the strength of sterling, especially against the euro. We have assumed a slight decline in sterling against the euro from the second quarter of this year. This is consistent with little change in the sterling-dollar exchange rate. As noted above, the expectations underlying the short-term forecast are based on UK entry into EMU at ??1.33 at the beginning of 2003.

Share prices have recovered substantially from their weakness in the late summer of last year. While there are strong grounds for believing them to be overvalued, we have forecast that they continue to rise at the same rate as nominal income.

[TABULAR DATA FOR TABLE 1 OMITTED]

Fiscal policy (table 2)

The March Budget did little to change the thrust of fiscal policy in the short to medium term. Overall, the many measures in the Budget implied a small fiscal relaxation of up to [pounds]3 1/2 billion in 2001-2. But in cyclically adjusted terms, public sector net borrowing remains set to be approximately zero over the next five fiscal years.

The cyclically adjusted budget deficit is not necessarily the best way of measuring the stance of fiscal policy. In our Fiscal Reports, we have emphasised the contribution that the public sector budget makes to national consumption and saving. On this basis, the overall stance of fiscal policy is expansionary in 1999-2000 as public sector consumption is set to rise by over 3 per cent, whereas more of the increase in taxation is being paid for out of saving rather than private consumption.

The outlook for the public finances has been transformed in recent years. After recording public sector net borrowing of [pounds]6.6 billion in 1997-8, itself a massive improvement on earlier years, the accounts showed a surplus of [pounds]5.2 billion in 1998-9. The improvement was achieved by continued firm expenditure control and some increase in the tax share. Spending rose by 3 per cent while cash receipts rose by 6 per cent. These may be compared with an increase of about 4 1/2 per cent in money GDP over the same period.

Our forecasts for the public finances are shown in Table 2. The outlook for public spending is consistent with that shown by the Treasury in the Budget report. But our revenue projections are less buoyant for 1999-2000 than those of the Treasury. This mainly reflects a difference in view on the outlook for corporation tax receipts, which has been made more uncertain than usual by a range of changes in the system being introduced this month. These include the abolition of Advance Corporation Tax, a cut in the corporate tax rate and the introduction of quarterly payments of the tax. Coming at a time when profits have been weak, but business investment and hence capital allowances strong, there is potential for receipts being lower than the Treasury expects. Our forecast for net borrowing of just over [pounds]9 billion is less optimistic than the Treasury's own forecast of borrowing of [pounds]4 1/2 billion. In 2000-1 and the years ahead, our forecasts are virtually identical to those of the Treasury.

These forecasts suggests that the government is well on course to meet its own fiscal targets of avoiding a deficit on the current balance and keeping its net indebtedness below 40 per cent of GDP.

Summary of the forecast

It now appears likely that the UK will narrowly avoid a recession this year. The rate of growth in the first quarter is estimated at 0.1 per cent, the same as in the fourth quarter of last year. With confidence returning to both the household and corporate sectors, and with government spending set to rise at its fastest rate since 1991, growth is now likely to pick up. We are forecasting that output will be close to 2 per cent higher by the end of the year than it was at the end of 1998. This is consistent with an annual growth rate of about 1 1/4 per cent in the year as a whole.

Domestic spending in 1999 is expected to grow at an annual rate of around 2 per cent, driven by household consumption growth of 2 per cent and growth in fixed investment and government consumption of around 3 per cent. An anticipated fall in inventory accumulation is set to reduce the growth rate by 1/4 per cent. Growth will be further reduced by a worsening trade position. A strong exchange rate and difficult international conditions are expected to lead to growth in exports of goods and services of a little over 1 per cent. By contrast, imports are expected to grow by close to 4 1/2 per cent. This worsening trade position reduces the growth rate by around 1 percentage point.

The manufacturing sector will not avoid a recession. Its greater reliance on external demand means that it will do much worse in 1999 than the economy as a whole. We now expect it to contract by 1 3/4 per cent in 1999. Output in the public sector is expected to rise by around 3 per cent in response to faster growth in public spending. Growth in the private services sector is likely to remain positive, but at a much lower rate than in recent years.

With only a mild slowdown, employment is expected to continue to rise in 1999, albeit more slowly than in the past four years. Productivity growth is likely to be slow, as is usual when growth is weak. We do not expect to see any substantial increase in unemployment.

Part of the explanation for the rise in employment over recent years has been the lack of any significant wage pressure. The introduction of the National Minimum Wage this month will add about 1 per cent to the level of wages, but with the economy slowing and overtime falling, the growth of average earnings is expected to moderate slightly to about 4 1/4 per cent. Unit labour costs are expected to grow by about 4 per cent over the course of the year. With demand weak, profit margins can again be expected to contract. As a consequence inflation is expected to remain on target. We expect RPIX inflation to decline to below 2 per cent by the end of this year.

Looking ahead to next year, the recovery is expected to become more firmly established. Indeed there are clear risks of renewed inflationary pressure. Household consumption is forecast to grow by over 2 1/2 per cent, supported by further firm income growth and a good background of low nominal interest rates, increasing wealth and receding worries about the jobs market. Fixed investment is also likely to benefit from a low cost of capital and improving confidence. Unless the exchange rate rises further, the outlook for the traded sector should also improve. We expect overall growth of about 2 1/2 per cent, although manufacturing is likely to continue its recent weakness.

The improving economic situation could pose a threat to the inflation outlook. While unit labour costs are expected to grow more slowly than in 1999 at 3 1/2 per cent, this rate is above that which is consistent with the inflation target in the long run. Furthermore, import prices and indirect taxes are set to increase. With profit margins reduced from their 1996 peak, companies will attempt to pass these cost increases on to prices as demand strengthens. We expect RPIX inflation of 2.7 per cent at the end of 2000.

A deteriorating trade position has had a negative influence on growth since 1996. Yet over this period the current account of the balance of payments has improved slightly as a deterioration in the goods balance has been offset by improvements in the services and income balances. We expect the current account to remain in balance in 1999, but to show a deficit of [pounds]10 billion in 2000.

There are quantifiable risks to the forecast. We estimate the probability of a recession, with output no higher at the end of 1999 than in the first quarter, to be around 17 per cent. The probability of a steeper slowdown with average output growth below zero in the year as a whole is estimated to be around 7 per cent. On the upside, we estimate the probability that growth exceeds 2 per cent to be about 16 per cent.

With regard to inflation, we see the chance of inflation being below 2 1/2 per cent at the end of the year to be around 75 per cent, with a 40 per cent chance that it is below 1 1/2 per cent. Looking almost two years ahead, we estimate a 42 per cent chance that inflation is below 2 1/2 per cent at the end of 2000.

Section II. The forecast in detail

The components of expenditure (table 3)

The background for renewed growth in demand is favourable, with all of the main domestic sectors well placed to increase spending. Households are in a strong financial position and are likely to experience good growth in income. Additionally, the low cost of capital should encourage further growth in capital spending by [TABULAR DATA FOR TABLE 3 OMITTED] companies. Government consumption is also expected to rise strongly in line with announced plans. The main restraints on growth are likely to come from the effects on confidence of some lingering fears of recession, possibly excessive stocks and the continued depressing effects of an overvalued exchange rate and slow world trade growth.

It does now appear that the threat of recession this year is receding. The main concern had been that output would fall around the beginning of the year. We estimate that in the first quarter of 1999, output grew by about 0.1 per cent. Within this total, exports of goods are forecast to have been static, in line with the trend in recent trade statistics, while imports are likely to have risen slightly. Fixed investment is forecast to have fallen back slightly from the high levels seen in the fourth quarter. We have forecast a rise of a little under 1/2 per cent in household consumption over the fourth quarter and expect a further period of inventory accumulation in the first quarter, at a similar rate to that seen in 1998. On this basis, domestic demand rose at a quarterly rate of about 0.2 per cent, but its effect on output was offset by a further worsening in the trade position.

In the year as a whole we expect to see a similar pattern of relatively modest growth in the domestic demand components partly offset by a negative contribution from net trade. Household consumption is forecast to grow at about 2 per cent, the slowest rate since 1995. Fixed investment is expected to rise by about 3 1/2 per cent as strong growth in the non-manufacturing and public sectors offsets a sharp fall in manufacturing investment. Government consumption is expected to be a major source of growth, rising by over 3 per cent in 1999. We expect inventory accumulation to make a negative contribution to growth of about 1/4 per cent, as companies attempt to prevent their stock levels building up excessively. Exports of goods and services are expected to remain weak, growing by just over 1 per cent over the year as a whole, the lowest rate since 1991. But a decline in imports growth as the economy slows will reduce the negative contribution to GDP growth of net trade.

Growth is then expected to pick up steadily to a rate of about 2 1/4 per cent in 2000 and 2001. This is accounted for by an improving contribution from net trade as export competitiveness improves and by continued growth in government consumption. The improved economic environment is expected to lead to a resumption of faster consumption growth of about 2 1/2 per cent per annum. The growth in fixed investment is expected to continue at about 3 1/2 per cent per annum.

Household sector (table 4)

Recent figures relating to the household sector probably disguise the potential for strong spending growth over the next two years or so. Figures on income, s have been distorted by the timing of tax changes and dividend receipts, whereas slower spending growth looks likely to be just a temporary pause. In our view, robust growth in underlying incomes, together with very buoyant asset prices and low interest rates, are likely to encourage a big improvement in consumer confidence and spending.

The rate of growth of household consumption has fallen over the past year, largely as a consequence of the unwinding of the effects of the windfalls received in 1997. Household consumption in the fourth quarter of 1998 was only 1.7 per cent higher than a year earlier, the lowest annual growth rate since 1995. Within this total, there has been a noticeable change in the composition of household spending. Over the same period, spending on durable goods is down by 0.7 per cent, spending on non-durable goods is down by 0.8 per cent, whereas spending on services is up by 4.6 per cent. It is likely that the slowdown in consumption growth observed recently is a temporary pause reflecting the earlier build-up of durable goods in 1996 and 1997, financed partly out of windfall receipts. There is now some evidence that spending on durables is beginning to pick up again, although the timing of spending on vehicles, the main component, has been affected by the introduction of the new biannual registration system.

The changed composition of spending has affected the relationship between household consumption and retail sales, which is worth about 40 per cent of the total. While the relationship is likely to be distorted in the future by Internet shopping and other methods of spending, there is little evidence to suggest that this has happened yet. In the three months to February, retail sales were 0.25 per cent higher than in the previous three months. This is consistent with our forecast of 0.4 [TABULAR DATA FOR TABLE 4 OMITTED] per cent growth in overall household consumption once some allowance is made for faster growth in spending on services.

The recent development of household income needs to be interpreted with some care. The main measure of household income, real household disposable income (RHDI) was stagnant in 1998. By contrast, real pre-tax labour incomes (compensation deflated by the consumer price index) rose by over 5 per cent. The reason for the difference is that household tax payments rose by more than 16 per cent, while net property income fell by over 8 per cent.

Households paid [pounds]15 billion more in income tax in 1998 than they had a year earlier. This partly reflects the introduction of self-assessment, which the government believes has led to a permanent increase in the tax base. This also had an effect on the timing of tax payments with more coming at the end of the tax year in 1997-8, thus distorting the comparison between calendar years. However, it also reflects the effect of earlier measures to raise taxes and efforts to secure the tax base. These included the ending of tax relief for profit-related pay and restricting to a lower rate the married couples allowance and tax relief on mortgage interest payments.

The main impact of this appears to have been a step-change in the average tax rate. While this would have a permanent impact on the level of household income, its effect on the growth rate will not now be repeated. In fact, the tax measures in the 1999 Budget prevent the average rate rising further. The introduction of a 10p starting rate of income tax is estimated to reduce income taxes by [pounds]1.5 billion this year, while for 2000-1 the effects of the abolition of mortgage interest relief and the married couples allowance (yielding [pounds]3 billion) are partially offset by the cut in the basic rate of tax (costing [pounds]2.25 billion).

The fall in net property income in 1998 was partly due to a sharp decline in dividend payments. Dividend payments by private non-financial corporations fell by 8 per cent in 1998 as a whole and, seasonally adjusted, were more than 40 per cent lower in the fourth quarter than in the first. While company profits have been weak, the recent pattern of dividend payments almost certainly reflects changes in the corporate tax system. The ending of dividend tax credits may have had a permanent effect on dividend payments while the abolition of Advance Corporation Tax (ACT) is likely to have encouraged companies to wait until after April to pay dividends. The other factor reducing household net property income was higher interest payments arising from the increases in mortgage rates seen until the middle of 1998. This is now set to be reversed following the recent cuts in rates.

In summary the factors which account for the stagnation in household incomes were either step-changes, such as the increase in the average tax rate and any reduction in company dividend payments due to the ending of tax credits, or are likely to be reversed, such as the change in the timing of dividends due to the abolition of ACT and higher mortgage rates. Without such depressing effects in 1999, real household disposable income growth is likely to be more closely related to the growth in real earnings. We are expecting that in 1999 average earnings will grow by around 4 1/2 per cent in nominal terms and by 2 1/2 per cent in real terms. With employment continuing to grow slightly this will raise the real disposable income of households by about 3 1/4 per cent. Excluding property incomes, real incomes are expected to grow by over 4 per cent.

Looking further ahead, the prospects for household income growth remain good. We expect earnings growth to rise to around 5 per cent in 2000 as the labour market begins to firm and productivity picks up as the economy grows more quickly. Real household disposable income is expected to grow at about the same rate as this year.

Alongside an improving outlook for household incomes, wealth is also providing strong support for consumer spending. At the end of 1998, households had net financial wealth of about [pounds]2000 billion (enough to finance current levels of consumption for about four years) and housing wealth of [pounds]1500 billion. While the stock market appears significantly overvalued, there are no strong grounds for forecasting a correction in the short term. Housing wealth is likely to increase further as a consequence of rising house prices. There have been reports of strong housing demand in certain parts of the country and there is a clear possibility that prices could rise rapidly in response to low mortgage rates. Against this, the outlook for house prices has been adversely affected by the ending of tax relief for mortgage interest payments. But this is now worth only about [pounds]1.5 billion per annum, which when capitalised is equivalent to about 1 1/2 per cent of the value of the housing stock. This will have only a modest impact on house prices. The increase in stamp duty for properties worth more than [pounds]250,000 will have a larger impact on prices at the top end of the market, but this is unlikely to affect significantly the overall value of housing. Our forecast for end-year house price inflation of 3 1/2 per cent is in line with that of the Halifax bank. This could prove to be conservative, but we would expect interest rates to be raised if there was a significant risk of house prices [TABULAR DATA FOR TABLE 5 OMITTED] rising very strongly. This would prevent a house price bubble of the sort seen in the late 1980s.

Against this background, household spending has considerable room for further expansion. Overall, we expect consumption growth of around 2 per cent this year. This is likely to pick up through the year and into 2000 when we expect to see growth of a little under 3 per cent. The household saving ratio is expected to rise back to around 8 1/2 per cent in 1999 and 2000.

With relatively subdued investment by the household sector, the rising saving ratio is expected to lead to renewed acquisition of financial assets. This is likely to be associated with relatively modest growth in mortgage and consumer credit, since households in aggregate are able to meet their spending plans out of their own resources.

Fixed investment and stockbuilding (tables 5 and 6)

Last year was another year of strong growth in fixed investment. Fixed investment grew by 8.6 per cent over the year, rising by 2.8 per cent in the last quarter alone. Growth in business investment was driven by growth of over 15 per cent in the non-manufacturing sector. Investment in manufacturing dropped by a little under 3/4 per cent.

We expect more moderate growth in investment in 1999. This mainly reflects the slowdown in the economy and continued uncertainty about the outlook for demand. Furthermore, despite relatively strong balance sheets, the financial position of companies has weakened slightly following strong capital spending. Nonfinancial companies moved into financial deficit in 1997 after four years of paying back debt. Their financial [TABULAR DATA FOR TABLE 6 OMITTED] deficit reached [pounds]8.5 billion in 1997 and almost [pounds]21 billion in 1998. This is expected to decline to about [pounds]7 billion in 1999 as interest rates fall and investment is more subdued. We do not however expect investment to stagnate. Recent business surveys suggest renewed business optimism. The British Chambers of Commerce survey shows an increase in the number of firms who have revised their investment plans upwards in the first quarter of this year, both in manufacturing and services. Also, the January CBI Industrial Trends survey reports that investment intentions are less negative than in October last year. The low cost of capital is another factor supporting investment now, although expectations of further reductions in the cost of capital over the next few years may mean that firms are still willing to postpone investment in light of slowing demand. Overall we forecast growth in fixed investment of a little less than 3 1/2 per cent this year, remaining relatively similar in 2000 and 2001.

Growth in business investment is expected to be very unequal between the traded and non-traded sector. This reflects the continued overvaluation of the pound and slow growth in world trade. We expect non-manufacturing business investment to rise by just under 8 per cent in 1999, declining to roughly 4 per cent in 2000 and 2001. Manufacturing investment is expected to be much weaker than this, with a fall of almost 10 per cent expected for 1999. Thereafter manufacturing investment is expected to remain roughly stagnant.

Private sector housing investment is expected to drop by roughly 2 1/4 per cent this year, in line with housing starts in 1998. However, low mortgage rates are likely to increase housing demand and we expect renewed growth in private sector housing investment of approximately 3 1/2 per cent in 2000. Public sector housing investment is expected to pick up quite sharply, rising by around 14 per cent in 1999, albeit from a low level. This is accounted for by the local authority capital receipts initiative. Other general government investment is expected to rise broadly in line with the government's plans. We expect this to rise by 13 1/2 per cent in 1999 and over 6 per cent in 2000 and 2001.

Inventory accumulation was a little over [pounds]3 1/2 billion in 1998. With the slowdown in demand in the last quarter of 1998 we expect some of this has been involuntary. We forecast stockbuilding of a little over [pounds]1 billion in 1999. With slower growth in the traded sector, we expect manufacturing inventories to be run down in 1999.

Balance of payments (tables 7 and 8)

The last couple of years have seen very difficult trading conditions for UK export firms. There is little prospect [TABULAR DATA FOR TABLE 7 OMITTED] of an improvement in the short term. Slowing growth in the world economy reduced the pace of world trade growth from over 10 per cent in 1997 to 7 per cent in 1998. By our current estimates world trade growth is likely to fall further to 5 per cent during 1999. This has contributed to an increase in competitive pressures on UK firms both in overseas markets and at home.

Sterling weakened on the foreign exchanges towards the end of 1998, helping to alleviate some of the pressures on UK exporters. Since then it has appreciated somewhat but remains well its highs at the start of last year. By our estimates sterling is currently overvalued. There is some evidence that its recent appreciation has not further reduced the optimism of UK exporters, probably on the expectation that sterling is likely to continue a downward trend this year. It is also worth noting that the measures of optimism started from a low base. The effective exchange rate is forecast to depreciate from around 102 in the second quarter of this year to about 100 at the end of 2000.

The lack of any substantial change in the nominal exchange rate will leave UK industry in an uncompetitive [TABULAR DATA FOR TABLE 8 OMITTED] position. Chart 8 shows unit labour costs in the UK relative to the Euro Area from 1984 to 2003. This provides some indication of the difficult conditions faced by industry both at present and in prospect.

Poor competitiveness and sluggish growth in world trade contribute to our view that the level of manufacturing exports will remain static this year, growing by only 1/2 per cent on the year earlier.

If competitiveness is unlikely to improve markedly in the short term, then the best hope for exporters is that world trade picks up. Our forecast of the global economy suggests that some increase in trade growth is likely in 2000 and 2001. Because of this, the growth of manufacturing exports is expected to rise next year, assisted by a slight sterling depreciation, to a rate of 5 per cent per annum.

The performance of the manufacturing export sector also compares relatively badly with the services export sector, which is worth about 40 per cent of goods exports. Exports of services grew strongly towards the end of last year despite the slowdown in world trade growth and is 0.5 to 1.0 per cent above that predicted within the model. This might suggest a favourable change in the import propensities of foreign countries for UK service exports. Whether this change is permanent, from a change in tastes or improved competitiveness from non-price factors such as after-service care, or temporary, the effect of 'Cool Britannia', remains to be seen.

Reductions in import prices resulting from the appreciating exchange rate have meant that the terms of trade have risen to levels last seen in 1992. Our forecast is that the terms of trade will rise remain at around this level throughout 1999 and 2000.

The strength of sterling combined with strong consumer demand within the UK has contributed to a rise in import growth over the last few years. Imports of both manufactured goods and services grew by over 10 per cent in 1996 and 1997 and by over 8 per cent in 1998. This is similar to the growth rates witnessed in the late 1980s. However, recent figures have suggested that import growth has now begun to decline. The figures for year-on-year growth for the fourth quarter of 1998 were 4.7 per cent in service imports and 7.8 per cent for manufacturing. As the UK economy slows further the growth of imports is expected to continue to fall through 1999 to just over 4 per cent (compared with 8.4 per cent in 1998) and 5 per cent in 2000.

Perhaps the most worrying feature of the balance of payments is the forecast deterioration of the goods balance. This reached a deficit of [pounds]6.3bn in the fourth quarter of last year and is forecast to increase to around [pounds]8 billion by the end of next year. This is a result of the continued strength of sterling and the recovery in UK demand next year. The goods deficit in 1998 is the highest since 1989 and is set to rise to [pounds]27 billion in 1999 and over [pounds]30 billion in 2000.

The balance on services, transfers and income is expected to have peaked in the first quarter of this year at [pounds]8 billion. This surplus is forecast to fall to [pounds]22 billion in 2000.

The turmoil in international financial markets had some benefits for the UK balance of payments in the third quarter of 1998. The losses by foreign owned banks and other financial institutions meant that these were marked as credits on the income balance such that the current account as a whole was some [pounds]2.5 billion in the black. In the fourth quarter the same effect was apparent but now augmented by the repatriation of losses of foreign owned oil companies based in the UK abroad.

[TABULAR DATA FOR TABLE 9 OMITTED]

The current account was in surplus by [pounds]1.9 billion for the whole of 1998. We forecast that the surplus will continue in the first part of the year before falling into deficit thereafter. We expect the current account to be in approximate balance in 1999, but a deficit of [pounds]10 billion is expected for 2000.

The current account balance is now a relatively minor contributor to the overall flows into and out of sterling assets. Table 8 indicates how the current account balance is financed. The so-called basic balance is composed of the current account balance plus net direct investment and portfolio investment. Recently this has been negative, reaching -[pounds]41.2 billion in 1998. The size [TABULAR DATA FOR TABLE 10 OMITTED] of the deficit is not due to the current account being in deficit, but is because direct and portfolio investment abroad has exceeded such investment into the UK. This balance has to be financed either by a repatriation of other investment back to the UK or by other foreign investment in the UK. In practice, this has been achieved by a net increase in foreign deposits in the UK banking system. Overall, the international investment position is such that the UK has [pounds]58 billion more foreign liabilities than it has assets. Even though we expect the current deficit to rise over the next two years, it is likely to remain a relatively small proportion of national income and will not have a substantial effect on the net foreign asset position.

Output and employment (tables 9 and 10)

The slowdown became apparent in the last quarter of 1998 when output grew by around 0.1 per cent, down from 0.4 per cent in the previous quarter and 0.7 per cent in the last quarter of 1997. Output in manufacturing fell sharply by 1.3 per cent, the largest drop since spring 1991. Output in financial intermediation also fell sharply by 1.5 per cent after frenetic activity in the third quarter, however growth in business services on the whole was still healthy. Growth was particularly strong in the electricity, gas & water supply industries and in the post & telecommunications industries at 1.5 per cent.

With confidence returning to both the household and corporate sectors and with the easing of fiscal policy, growth is likely to improve through the remainder of this year. Public spending is expected to contribute significantly to growth in 1999. Annual growth in public sector output is expected to rise from 2 per cent last year to just over 3 per cent this year. Private service industries should continue to grow in response to renewed expansion in demand, albeit at a slower rate than last year. In the distribution and business services sector we are expecting growth of around 1 1/4 and 3 1/4 per cent respectively in 1999, picking up slightly in 2000.

Compared with the last months of 1998, output in manufacturing has remained relatively stable in the first months of 1999, declining by approximately 0.3 per cent. With the less pessimistic outlook for growth this year overall, manufacturing is unlikely to suffer severe decline. However, manufacturing continues to be constrained by foreign demand. The pound remains overvalued and world trade is not expected to grow strongly. On average we are forecasting a decline in manufacturing output of 0.4 per cent per quarter for the first three quarters of the year. With expectations of a small depreciation in the exchange rate, we forecast positive growth in manufacturing output towards the end of the year as competitiveness improves.

Productivity growth in 1999 is expected to fall. Slowdowns are normally associated with reductions in productivity growth as capacity utilisation falls and firms hold on to labour to avoid firing costs and subsequent retraining costs. The likelihood that the slowdown will be short-lived accentuates the fall in productivity, since firms will be less willing to adjust labour to temporary fluctuations in demand. This is indeed reflected in the fact that the growth slowdown so far has not been associated with a reduction in employment.

The introduction of the National Minimum Wage (NMW) is likely to affect productivity in the opposite direction. Introduced at the beginning of April this year, the NMW guarantees employees a minimum hourly pay of [pounds]3.60, or [pounds]3.00 for employees between 18 and 21. This is initially expected to affect the pay of approximately 1.9 million workers, raising salary costs by roughly half a per cent. It is likely that firms will attempt to avoid a similar rise in unit labour costs by taking steps to increase productivity. As activity picks up next year, productivity growth is expected to rise again.

By way of contrast to the general picture, growth in manufacturing productivity might improve this year. Manufacturing productivity grew by 1 per cent in 1998. The latest figures suggest that growth in manufacturing productivity is picking up. Compared to a year earlier, manufacturing productivity increased by 1.7 per cent in the three months to February 1999. This is likely to reflect recent reductions in manufacturing employment and prolonged competitive pressure in export markets.

Employment continues to rise. According to the Labour Force Survey, in the quarter from December to February this year employment is at 27.34 million, up by 80,000 on the previous quarter. This also brings a rise in the employment rate. We are expecting employment to grow by just under 1 per cent in 1999, rising to roughly 27.5 million by the end of the year. This is less than in the past few years, reflecting the general slowdown. Unless the slowdown is more severe than we are predicting, we do not expect employment to fall. This is supported by recent surveys of employers and recruitment consultancies, indicating a marked easing in the rate of job losses throughout the economy and a rise in the number of permanent placements.

ILO unemployment rose by 32,000 from the three months to November to the three months to February this year, leaving the ILO rate at 6.3 per cent up from 6.2 per cent. The claimant count has remained roughly stable at 4.6 per cent. The rise in the ILO rate reflects an increase in economic activity rather than a decrease in employment. We expect the ILO rate to remain at 6.2 per cent throughout the rest of the year. The claimant count is expected to remain around 4.5 per cent. Our forecast of stable unemployment is contingent on relatively subdued growth in earnings. The latest figure for average earnings growth was up from 4.5 per cent in the year to January to 4.6 per cent in the year to February. However, this is unlikely to indicate increasing wage pressure. Growth in basic pay settlements has fallen from 3.1 per cent in the three months to January to 3.0 per cent in the three months to February. Also, in comparison with settlements growth of 3.5 per cent last year, this does not suggest increasing wage pressure.

The government's New Deal (ND) programme is also likely to help keep unemployment in check, particularly long-term unemployment. Initially the New Deal provides assistance to the long-term unemployed in job search and provides subsidised employment or training [TABULAR DATA FOR TABLE 11 OMITTED] to those who are unsuccessful in finding employment. In December last year 200,000 people were participating in either the ND for the long-term unemployed or the ND for young people. The already extensive programme was extended with a special ND for the over 50 year olds in the March Budget. By helping the long-term unemployed in job search and by providing people with basic skills, we expect the ND to increase the effective labour supply, alleviating wage pressure and increasing sustainable employment.

Our forecast for employment and unemployment is also affected by the introduction of the NMW. By increasing wage pressure, the introduction of the NMW is expected to have some adverse effect on employment. Although employment growth in sectors dominated by low pay has been quoted as evidence that the NMW will not adversely affect employment, this could equally imply that employment growth would have been stronger in the absence of the NMW. It is however unlikely that the NMW will affect employment by much in the short term. Survey evidence suggests that employers are comfortable with the level set for the NMW, suggesting that there should be little immediate job loss associated with it.

Earnings and prices (tables 4 and 11)

There is mixed evidence on trends in pay and their implications for inflationary pressure. The latest evidence on pay settlements suggests a reduction in the rate of earnings increase. The IRS survey recorded that the median rate of increase in settlements fell from 3.5 per cent in December to 3.1 per cent in January and 3.0 per cent in February. By contrast, the recently reinstated official average earnings index showed a rise in earnings growth to 4.6 per cent in the year to February, up from 4.5 per cent in January. Past experience (e.g. in 1997) shows that the index does sometimes diverge from the comprehensive national accounts data available up to a year later. This has to be borne in mind. In terms of their broad trends, both series indicate that pay pressures are not strong but need to be monitored carefully.

Our forecast suggests a slowing of earnings growth on the national accounts measure from 4.9 per cent at the end of 1998 to about 4 1/2 per cent at the end of this year. It is therefore compatible with the information on settlements. This is despite the allowance of a 1 per cent increase in the level of earnings to take account of the effect of the NMW. Real earnings growth is projected to fall when the Consumer Price Index is used as the deflator and to rise when the Retail Price Index is used. The former suggests continued growth in real pay of about 2 per cent per annum, consistent with long-term productivity trends. This indicates that the labour market is probably near equilibrium at present.

Low productivity growth is partly responsible for an acceleration in unit labour cost growth to 5.3 per cent for the fourth quarter of 1998. This is somewhat faster than the 4.8 and 4.6 per cent seen in the second and third quarters respectively. As productivity begins to increase through 2000 and 2001, growth of unit labour costs is expected to slow, despite the rise in average earnings growth mentioned above. Our forecast for growth of unit labour costs is 4.1 per cent this year and around 3.0 per cent in 2000 and 2001. This is very slightly faster than is consistent with the government's inflation target.

Elsewhere in the economy inflationary pressures remain reasonably weak. The GDP deflator, the most comprehensive measure of cost pressures, grew by 2.1 per cent in the fourth quarter of 1998, up slightly on the 1.7 per cent registered in the third quarter. Our forecast for 1999 shows inflation by this measure to remain steady at close to 2 per cent before rising slightly to 3 per cent through 2000 and 2001.

Producer input prices fell again in the fourth quarter of 1998, by close to 5 per cent on the previous year. This represents a further deceleration in cost pressure after year-on-year falls of 3.7 per cent in the second quarter and 4.2 per cent in the third. Producer input prices are expected to continue to fall through this year before rising through 2000 at around 0.5-1.0 per cent per annum.

A large part of this expected increase in input price inflation is accounted for by increases in the price of oil. The historically volatile oil price index currently stands at 62.0 (1995 = 100). We forecast the price of oil to rise by around 33 per cent this year. However, despite this increase, our forecast is for the index of producer output prices to remain relatively fiat. This is explained largely by continued competitive pressure from overseas so that growth in unit labour costs is offset by lower profit margins.

The imports deflator fell again in the fourth quarter of 1998 and the index now stands at 86.7 (1995 = 100). These price falls, which have been going on throughout 1997 and 1998, have been faster in the goods and oil sector than in the service sector. The index for the price of imports of manufactured goods stands at 84.4 in the fourth quarter of 1998 (the last available data point) while the index of service imports stands at 96.5 (both 1995 = 100). The difference between sectors reflects differences in the ability of firms in different sectors to widen their profit margins.

We forecast that the index for import prices will cease falling around the middle of this year. Our forecast is that import prices will fall by 1.9 per cent in 1999 but rise by around 1.2 per cent in 2000 and 4.7 per cent in 2001.

The strength of sterling has meant that UK exporters have been forced to reduce their prices in order to remain competitive. According to the latest available data the index of export prices stands at 91.6 in the fourth quarter of 1998 (1995 = 100) against 92.3 in the third quarter. Price cuts are greater than those predicted by normal fundamentals and have been achieved by squeezing profit margins. Reductions in export prices by this means can be regarded only as a short-term solution and help explain the relief of UK exporters to recent cuts in UK base rates. However weaker demand conditions here and abroad mean that firms are unlikely to be able to reverse these cuts in the near future. The falls in the index of export prices have been slower than those for import prices and, all other things being equal, also suggest a loss of competitiveness of UK exporters on world markets.

Again there is significant contrast between sectors of the economy. The majority of falls in the aggregate export price index have been achieved by falls in the price of UK goods rather than services. Prices of manufactured goods are around 10 per cent lower than the base year of 1995, while prices in the service sector are only 2 per cent lower. This difference can be explained in part by the strong growth in average earnings, which has been fastest in the service sector and accounts for a large part of costs. The index for manufacturing export prices is forecast to continue to fall through 1999, whereas the index for the price of services exports is forecast to grow over the year as a whole. The overall export price index is expected to continue to fall up until the middle of the year before rising thereafter. Our forecast is for export price inflation of -0.6 per cent this year and around 11.5 per cent in 2000 and 2001.

RPIX inflation rose in March to 2.7 per cent, marginally higher than the government's target of 2.5 per cent. This was mainly a result of indirect tax changes made in [TABULAR DATA FOR TABLE 12 OMITTED] the Budget. More importantly the fall in RPIY inflation, which excludes the effect of taxation, suggests that inflationary pressures in the economy remain weak. RPI inflation, the headline measure of retail price inflation, also fell over the period, mostly because of reductions in mortgage interest payments. The index for mortgage interest payments is forecast to fall continually through 1999 and into 2000 as recent cuts in base rates feed into lower mortgage payments which account for around 6 per cent of the RPI index.

We expect RPIX inflation to remain stable during 1999. Average earnings growth is forecast to quicken in 2000 and 2001 as the economy recovers and sterling weakens on the foreign exchange markets. However the effect on RPIX inflation is forecast to be small as some of the increase in average earnings growth is offset by increases in productivity and import price inflation remains low. We forecast that RPIX inflation remains below target through 1999 and only rises above target by the end of 2000.

RPIY inflation excludes from it both changes in mortgage interest payments and indirect taxation. Changes in indirect taxation have meant that the RPIY has remained below the RPIX measure over recent years and is expected to continue to remain below RPIX inflation through 1999 to 2001.

Inflation in the Harmonised Index of Consumer Prices (HICP) for March currently stands at 1.7 per cent, up slightly on the figures recorded through the second half of 1998. However this is higher than the average for the EU 15 of 1 per cent. We expect this measure of inflation to rise slightly in 1999, averaging 1.6 per cent for the year. The target range for the European Central Bank is 1.5 to 2 per cent.

National and sectoral saving (table 12) Table 12 shows our forecasts of national and sectoral saving. This puts together parts of the forecast that have already been discussed. This shows that in 1998 saving and investment in the UK were almost exactly balanced with domestic investment financed by domestic saving. A similar pattern is expected for 1999.

Over the following two years, national investment is expected to rise relative to saving. This is the equivalent of an increase in the deficit on the current account of the balance of payments, which is expected to worsen to 1 per cent of GDP by 2000. In this way the overseas sector provide the necessary savings to finance the domestic shortfall. A current account deficit of this size is not a problem, although it will mean that the net financial position of the UK will worsen further.

Within the domestic economy, the household sector is in its customary position of saving more than it invests and lending the surplus to other sectors. Investment by the company sector is expected to continue at about 12 per cent of GDP, larger than it is able to fund from retained profits alone. As a consequence the company sector is expected to draw in funds from other sectors to finance its investment plans. Its deficit is expected to reach over 2 per cent of GDP by 2001.

[TABULAR DATA FOR TABLE 13 OMITTED]

The government sector is in approximate balance at present. While saving is expected to remain positive (thus meeting the Golden Rule on one possible definition), it is not expected to be sufficient to finance rising investment (so that the government continues to run a deficit). Thus it too is likely to need to borrow from other sectors to finance its need for capital.

The economy in the medium term (table 13)

The way in which the economy behaves over the medium term is determined partly by a range of shocks that are inherently unpredictable. But there are other important influences on its development that can be foreseen. These include trends in the size and composition of the population, forthcoming changes in the policy framework as well as adjustments to existing disequilibria. It can be argued that the development of the British economy over the 1990s has been largely shaped by the need to adjust to the recession of the early part of the decade. The main problem for the economy over the past two years or more has been its grossly overvalued exchange rate. The adjustment to this will have a continuing effect on the economy over the coming years. But in other ways the economy is well balanced with inflation low and unemployment probably close to its sustainable rate.

Foremost among the policy influences is whether the UK decides to adopt the euro. We have assumed that this will take place at the beginning of 2003, but that sterling will be fixed at ??1.42 (equivalent to DM2.775), a higher rate than most people currently anticipate. UK interest rates are assumed to have converged on euro rates by the beginning of 2003. On the basis of current market rates, this would mean that UK interest rates fall to around 3 1/2 per cent in the early years of the next decade, before rising to 4 3/4 per cent by the end of the next decade.

Inflation is forecast to be very low over this period. This is based on a view that the European Central Bank (ECB) will be as successful in controlling inflation as the Bundesbank has been in recent years. With the sterling exchange rate fixed, there will not be room for large differences in inflation across the Euro Area. So long as this is clearly understood, expectations, which are crucially important in the inflationary process, should help anchor inflation itself.

It is possible that sterling will enter the Euro Area at an exchange rate which is overvalued by about 10 per cent, after allowing for UK inflation between now and 2003. This will also contribute some downward pressure on inflation. The effective exchange rate is expected to appreciate throughout most of the decade as the euro rises against the dollar.

The outlook for interest rates and inflation is consistent with real interest rates of around 2 to 3 per cent. This is much lower than has been normal in the UK over the past twenty years, but is consistent with real yields on UK government index linked debt.

Fiscal policy is assumed to be tight on average over the period in line with the current government's targets. But if the UK adopts the single currency, then fiscal policy will need to be used for stabilisation purposes. To have sufficient flexibility for this purpose, it will be necessary for the budget deficit to be small on average. In keeping with this, government consumption is expected to grow by less than GDP. This would allow some increased spending on transfer payments as the proportion of pensioners in the population rises.

Unemployment, on the ILO definition, is forecast to settle at about 6 1/2 per cent of the workforce. This is slightly higher than is the case now, and is a reasonable estimate of the sustainable rate of unemployment. The rate of real wage growth is sensitive to which deflator is used. Using the GDP deflator at basic prices, the real wage is forecast to grow at about 1 3/4-2 per cent per annum, the same as the rate of growth of productivity. Other price indices are expected to grow a little more quickly than this, reflecting small increases in indirect taxes, a slight worsening of the terms of trade and, in the case of the RPI, an increase in mortgage costs once interest rates begin to rise.

The government has identified slow productivity growth as one of the important problems of the UK economy and it is possible that policy action over the coming years will be successful in raising it. However, we have made no special allowance for this. Our forecast of productivity growth is consistent with long-term trends in the economy.

On this basis we would expect to see economic growth of between 2 to 2 1/2 per cent per annum over the next ten years, with lower growth beyond that as the population of working age rises more slowly. Among the expenditure components, household and government consumption are expected to grow by about 2 per cent per annum with slightly faster growth in fixed investment.

The current account deficit is set to average about 1 1/2 per cent of GDP throughout the period.

Forecast errors and probability distribution (tables 14 and 15)

Table 14 provides a set of summary information as regards the accuracy of forecasts that have been published in the April Review. The latest National Accounts information available when these forecasts are constructed is for the fourth quarter of the preceding year.

A rule of thumb is that a 70 per cent confidence interval for a variable of interest can be obtained by adding a range of one absolute average error around our central forecast. Thus we can be 70 per cent sure that GDP growth in 1999 will be between 0.4 and 2 per cent. The size of the average errors indicates that some variables are easier to forecast than others. For example, the errors in forecasting consumers' expenditure are smaller than those in forecasting fixed investment. It is also the case that the error in forecasting GDP growth is smaller than that made in forecasting its components. This arises because of offsetting movements among the components. The probability distributions around the growth and inflation forecasts have been calculated assuming that the distributions are normal and are shown in table 15. The standard errors have been calculated from the historical forecast errors underlying table 14. These estimates of the probability distribution around our central forecasts provide a quantitative assessment of the limits of our forecasts.

[TABULAR DATA FOR TABLE 14 OMITTED]

[TABULAR DATA FOR TABLE 15 OMITTED]

The forecast was compiled using the latest version of the Natinal Institute Domestic Econometric Model. We are grateful to Ray Barrell, Nigel Pain and Martin Weale for comment and discussion.