The world economy.

Barrell, R.J. ; Gurney, Andrew

CHAPTER II. THE WORLD ECONOMY

The overall outlook: longer-term prospects

The second half of the 1980s has seen relatively strong, and rising growth in most of the major seven economies. This performance has been accompanied by a slow increase in the average rate of inflation. This combination was aided by the fall in world oil prices in 1986 and weak commodity prices. Only toward the end of the decade did strong growth begin to put pressure on commodity prices, and then only temporarily because of stock shortages of metals and minerals. The world economy was also operating with some degree of spare capacity in 1986 and 1987, allowing a period of sustained non-inflationary growth.

The performance of the advanced economies in the last few years is even more remarkable when account is taken of the serious worries that have been expressed both about large scale current account imbalances and the size of the US Federal deficit. Many commentators saw the current account imbalances as unsustainable and hence ultimately destablising. This position appears to have been rather pessimistic. In a world with minimal capital controls, exchange rates are likely to be determined by the demand for asset stocks rather than by flows across the exchanges. Simple portfolio models can be constructed that suggest that even a small preference on the part of non-residents for US assets is sufficient to produce large scale capital flows into the US.(1) Fiscal deficits also appear to be financable, at least in the short run.

We have tried to take account of these lessons when constructing our forecast for the next decade. Table 1 outlines our forecast. (We have, for the first time produced a forecast base up to the year 1999.) Although we expect some slowdown in world activity through to 1992, it is nowhere as severe as that experienced in the early 1980s. We also anticipate that growth will thereafter strengthen. Model based forecasts do have a tendency to smooth the future. We may expect cycles, but if they are produced as a response to shocks then they are difficult to forecast. However, we do feel that the average performance of the world economy in the 1990s will be significantly-different from the 1980s. The inflationary pressures at the beginning of the decade are unlikely to be so severe, and the monetary response is unlikely to be so savage as in the early 1980s. Also, cycles are often produced by sharp fiscal policy changes, and we do not see any need for major fiscal policy changes in any of the major economies. Fiscal consolidation has proceeded successfully in Germany, Japan, and France, and there may even be room for increased government expenditure growth in these countries. Fiscal policy in the US is tightening, but the political problems involved in reconciling the views of a Democrat Congress with a Republican President inevitably mean that the process of deficit reduction will proceed slowly.

Developments in Europe during the 1990s are likely to be much influenced by the moves to closer economic integration that are resulting from the single market programme. The gradual reduction of barriers to trade that will take place is in many ways a continuation of the process that has been underway in Europe since the Treaty of Rome in 1958 and in the world at least since the Kennedy round GATT agreements in the early 1960s. It may be true to say, as the Commission does, that the level of output in Europe will eventually be 5 per cent higher with the single market programme than it would otherwise have been. It is however wrong to necessarily conclude from this that growth in Europe will in future be higher than it has been in the past. The Community has already reaped major output gains from freer trade, and although the 1992 programme may be broader than those seen earlier, it has had a rather small effect on our projections of the average growth rate in Europe in the 1990s. Nevertheless, we do expect higher growth in Europe in the next ten years than in the last ten. This is partly because of the 1992 programme, but it is also the consequence of our presumption that macroeconomic policy will be both more stable and better coordinated than it has been in the 1980s.

It is to be hoped that further integration in Europe does not lead to increased external barriers to trade. Prospects for world trade and growth in general, and for the less developed countries in particular, depend upon an absence of increasing protectionism. The outlook for the third world is rather diverse. The prospects for Far East Asia look good. Increasing coordination of the Korean and Taiwanese economies with Japan seems likely. Large scale direct investment flows have been changing the structure of production in the area, and we expect these flows and the consequent output growth to continue. Our model of the Far East economies embeds in its export equation the rapid gain in world trade share that has taken place over the last two decades, and we are projecting that this will continue. Prospects for Latin America also look bright, partly because we foresee no major recession in North America. However, political developments there will remain crucial to the prospects for growth. In contrast to these two groups, prospects for the sub-Saharan African economies remain bleak. Their continued dependence on primary products facing slow demand growth suggests to us that their exports will, as in the past, grow only slowly. Only large scale capital inflows can augment their ability to import.

The forecast is, of course, not without its risks. We have not attempted to forecast the implication of the major changes that are taking place in the political and economic structure of Eastern Europe. We have judgementally increased the Centrally Planned Economies imports, and allowed them a small current account surplus. However, over a ten-year period we could easily see the emergence of the Baltic states as the new Hong Kong, Poland and Czechoslovakia could regain their pre-war economic dynamism, or Russia could again become a major participant in world manufacturing trade. Perhaps the most significant possibility would be the reunificiation of Germany. This would transform both the country groupings on our model and the prospects for economic growth in Europe. As reunification is highly probable, especially when considering a ten-year horizon, our forecast may well have to be recast.

The overall outlook: shorter-term prospects

Despite our relatively sanguine position about developments in the 1990s, there are risks in the short term. As chart 1 shows, interest rates have continued to rise in the last few months. The process of exchange-rate adjustment has possibly been set back by the recent strength of the dollar, but we do not currently foresee any prospects for a major realignment of exchange rates.

Consumer price inflation in the major seven has risen from 3.3 per cent in 1988 to 4.1 per cent in 1989. However the recent rise in interest rates should, along with more restrictive fiscal policy partially reverses this rise. Chart 2 gives both recent and prospective developments in the rate of inflation in the major economies. The outlook for inflation will be aided by recent commodity price developments. Oil prices have been relatively buoyant recently, but have only just managed to stay around the OPEC marker price of $18 per barrel. Strong world growth could put upward pressure on oil prices, but we are assuming that they will be basically stable in 1990 as the resumption of full Iraqi production after the end of the Gulf war increases available supply. Table 2 gives our commodity price forecast, and chart 3 plots the recent developments upon which it is based. The free market dollar prices of developed country food have declined by around 10 per cent since the end of 1988. The grain and vegetable crops in North America and in parts of northern Europe have been good, and prices have dropped from their drought induced peaks of last year. The free market prices of food-stuffs from developing countries have fallen even more sharply, as coffee and cocoa prices have fallen. We are expecting that these prices will now stabilise, and that they will remain around their current levels for the next two years. Metals prices have also fallen by around 10 per cent from the peaks reached at the end of 1988, and although there are still some supply difficulties in the copper market, sufficient new capacity has been brought on stream to cope with the moderating demand growth that we are projecting.

World trade growth is also likely to slow down, partly because of a decline in activity in world oil markets, and also because lower commodity prices will reduce export revenues in some of the less developed countries, leading to a reduction in their import volumes. Also Chinese imports, which had been rising strongly, are liable to be greatly reduced after the political upheavals of this summer. Trade growth, however measured, will remain relatively buoyant in the early 1990s. This is partly the consequence of increasing imports in the oil producing countries, and partly because the newly industrialised countries of South East Asia increase their imports. The former is in response to relatively buoyant oil revenues, and the latter is the consequence of major declines in competitiveness over the last eighteen months in both Korea and Taiwan.

Exchange rates and interest rates

Our forecast has once again been constructed on the premise that operators in the foreign exchange markets are rational and forward looking. We project that

exchange rates will change in line with current and anticipated interest-rate differentials, and that the future path for interest rates can be read from a combination of current long rates and government policy pronouncement. Historically long rates have exceeded short rates, perhaps reflecting the presence of a liquidity premium on assets at the short end of the market. Currently long rates, which can be seen as the average of future expected short rates, are below short rates in all of the major seven economies except Japan, implying that short rates are expected to decline. Table 3 sets out our forecast for interest rates.

We are projecting that current interest-rate differentials will be basically unchanged, but that short rates will generally decline. There are signs that the US Federal Reserve is loosening its policy stance, and that the Bank of Japan is tightening theirs. We are projecting that US rates fall by almost half a percentage point over the next year whilst Japanese rates stay firm and then decline slightly. These differences reflect the different worries about emerging inflationary pressures. The US has been tightening its monetary policy for longer than the Japanese, and there are already signs of the American economy slowing down. German interest rates have also risen sharply recently in response to inflationary pressures, although not all European countries have been emulating them.

In the last few weeks the DM and French franc and the lira have all strengthened against the dollar, reducing inflationary pressures in those countries. Chart 4 plots recent and projected exchange-rate changes. Our exchange-rate forecasts are given in table 4. We have presumed that French interest rates will remain two percentage points above those in Germany over the forecast period, and rates in Italy are projected to be some 5 per cent higher than in Germany. These differentials partly reflect risk premia (of 1/2 per cent and 2 per cent respectively) but the removal of capital controls in Europe has reduced the ability of the French, Danish and Italians to maintain artificially high interest rates (see box 1). The exchange-rate forecast that follows from our interest-rate projections will require a number of re-alignments of the EMS over the next decade, the first probably being triggered by the lira in mid 1991 (unless of course they reduce their tolerance bound from its current 5 per cent to 2 1/2 per cent, when the realignments would have to occur earlier). We have not assumed that a monetary union will be formed in Europe during our forecast period.

Our interest-rate projections imply that the recent strengthening of the dollar will unwind only slowly, at least after the correction that appears to be taking place at the end of 1989. Table 4 contains our exchange-rate forecasts and also gives our projections for real exchange rates. We are projecting that the yen continues to appreciate in real terms, whilst the franc and the lira depreciate. The DM appreciates only slowly, reflecting the constraints placed on it by the operation of the EMS.

The combination of interest rates and exchange rates in our forecast along with projected fiscal policies, produces a process of slow adjustment in world current account surpluses. Chart 5 plots the current account to GDP ratios for the major economies. Capital outflows from Germany and Japan are maintained, and the US maintains a capital inflow. The scale of these flows slowly reduces, but we are projecting the US current account deficit will only have fallen from 3.4 per cent of GDP in 1987 to 1.1 per cent in 1999. Over that period we expect US real interest rates to fall from their current level of 4.7 per cent to around 3 1/2 per cent. Over the same period Japanese real interest rates are also projected to fall to around 1 1/2 per cent, and we expect German real rates to be around 2 1/2 to 3 per cent for most of the decade. These real interest-rate differentials will continue to be a major force sustaining capital flows behind the current account imbalances.

There is always a risk that the exchange-rate paths we have projected will prove to be unsustainable. In particular we are curently assuming that there will only be a slow decline in the dollar over the next ten years. This leaves the US with a sequence of large but we think financable current account deficits. However, sentiment may change, and the markets may decide that a path that heads towards a current account deficit for the US of 1 per cent of GNP in 1999 is not sustainable. The dollar would then have to drop. We think that this is a serious risk. Indeed in our last forecast we considered that this was such a serious possibility that we adopted it as our main case. Since that forecast there has been sufficient new information, such as that detailed in box 2, that we felt that the balance of probabilities has shifted just sufficiently for us to change the basis for our main case forecast. The forecast contained in the August 1989 Review could be seen as a variant on the current chapter, based on the presumption that the dollar has to fall by 10 per cent in the next two years.

United States

The US economy appears to be slowing down, but at a relatively moderate pace. Industrial production fell 0.8 per cent at an annual rate in September, and it has risen by only 1.4 per cent since the beginning of the year, compared to 4.9 per cent growth during 1988. Consumer prices rose by 1.9 per cent at an annual rate in September, below both the 4.4 per cent rate since December, and the same rate in the previous year. However, retail sales rose moderately in both August and September. Auto sales were very buoyant largely as a response to heavy price discounting. Consumer confidence, as measured by questionnaire surveys, is also fairly high. The signs of a moderate slowdown of growth in activity are reinforced by declining capacity utilisation in manufacturing, which stood at 83.7 per cent in September compared to 84.4 per cent in the second quarter. Civilian unemployment rose slightly in September to 5.3 per cent, but this is still below the 5.5 per cent seen in 1988, and well below its recent peak in 1982.

Further indicators of a moderate slowdown in activity are also available in the preliminary estimates of GNP for the third quarter. Overall growth in real GNP in the quarter was 2.5 per cent at an annual rate, much more moderate than growth rates observed in 1988 and early 1989. However, consumer spending rose by 5.8 per cent at an annual rate in the quarter, buoyed up by incentives for car purchase. Imports of goods and services rose 15.1 per cent at an annual rate, whilst exports were flat. This change in the external position has to be interpreted with some care, as merchandise trade has not deteriorated quite so rapidly. The strength of the dollar in the third quarter would have reduced the real value of property income from abroad, reducing GNP immediately. Although this effect is statistically correct, the declines in GNP growth in the second and third quarter that it has produced are unlikely to have much behavioural impact. They may also be rapidly reversed if the dollar depreciates.

We are forecasting that the slowdown in activity will continue into 1990, and even 1991, but that output growth will only be reduced to 1 3/4 per cent per annum. This is however, considerably below both the output growth of 3.3 per cent projected for 1989 and the 4.6 per cent experienced in 1988. The growth in domestic demand is forecast to slow down much less rapidly than output, falling from 2.8 per cent in 1988 to 2 per cent in 1990 and 1 3/4 per cent in 1991. The strength of the dollar during the middle of 1989 has eroded some of the competitiveness gains achieved in 1986 and 1987. This will reduce the growth of exports of goods from over 20 per cent in 1988 to around 5 per cent in 1990. The appreciation will also reduce the contribution of invisibles to real GNP, both through the effects of competitiveness on trade in non-factor services, and also through a decline in the real value of property income receipts. The interpretation of trends in the contribution to GNP of services receipts and payments in the balance of payments has been clouded somewhat by recent methodological changes. This issue is discussed further in box 2.

The Federal Reserve has been easing monetary policy, and in October the federal funds rate was 1/4 per cent below its summer average of 9 per cent. This was partly in response to instability in world financial markets. A weaker dollar in October, along with signs of a slowing economy also gave space for some easing of the policy stance. We are projecting that US interest rates will continue to fall slowly during 1990 and 1991, but that some positive differential will be maintained against the rest of the major seven economies. This should allow the dollar to continue falling at about 1 per cent per annum from its current level. The gradual easing of the monetary stance should prevent any precipitate decline in the rate of inflation but a slowing economy should raise unemployment to around 6 per cent in 1990, and average earnings growth should slow down to around 4 per cent in the early 1990s. The rate of increase in consumer prices is forecast to decline from 5 per cent in 1989 to 4 1/2 per cent in 1990, and to continue to fall by 1/4 per cent a year thereafter. The inflation prospects are much aided by projected slow growth in both producer prices and in manufacturing earnings. The former is the result of weak demand and falling or weak commodity prices, whilst the latter is a continuation of the recent trend of declining real earnings in US manufacturing.

The prospects for any major improvements in the US current account are slight, and we are anticipating a larger deficit in 1990 than in 1989. Table 6 gives the details of our current account forecast. The visible balance in 1989 has benefitted both from the strong growth in exports, and from the improvements in the terms of trade consequent on the rise in the dollar during the year. However, the visible trade balance will deteriorate in 1990, as the subsequent loss of competitiveness is likely to reduce export growth next year, and some adjustment in trade prices in response to the stronger dollar is likely. Slower domestic demand growth in 1990 and 1991 will reduce the growth of imports, and the visible balance is not forecast to deteriorate further. The non-factor services balance is expected to continue to improve, and in combination with a projected reduction in defence related non-requitted transfers, should be sufficient to offset the continuing decline in net property income from abroad. Large US deficits have reduced the net overseas asset position, and continuing deficits are likely to make the US a net debtor in 1993. (Measured asset stocks, which have shown a deficit for some time, do not reflect the underlying value of all assets, and it is probably better to assess the debt position from income flows.)

There has been much discussion of the US government's fiscal position as a result of the temporary sequester order for $16 billion issued on 16th October.(2) However, the prospects for gradual reductions in the government deficit remain good. Table 7 gives our forecast. We have adjusted our basic fiscal assumptions for the short term for this forecast, producing a slighter greater negative impact on GNP. We are assuming that a $30 billion tax package will be introduced in the period up to 1991Q1. The current budget reconciliation bill introduces some increases in excise and income taxes, and we expect the 1991 budget (to be passed in 1990) to contain further increases of $10 billion in both excise and income taxes. Even if the administration persuades Congress to introduce changes to capital gains tax, they would be unlikely to have the same impact on the economy as an income tax increase. The proposal to reduce capital gains tax for two years would beyond doubt raise considerable revenue, as the realisation of gains was brought forward, but this increase in revenue would not affect activity, and hence we would have to ensure that it had no effect on our output forecasts in the short term. This tax reform would, however, reduce revenue in the longer term, and if it were introduced we would anticipate higher income taxes after 1992. We have revised down our projections for government spending in the medium term, as real defence spending is likely to continue to decline, and non-defence purchases will rise only slowly. In the longer term a neutral fiscal stance requires that the ratio of government debt to GDP remain constant, and to ensure that this can happen we are forecasting that the public sector deficit will settle at around 0.6 per cent of GNP. The Federal deficit is likely to remain somewhat higher than this because it excludes the surplus on the social security fund, and includes grants in aid to state and local governments.

The slowdown in the US economy along with the gradual elimination of the current account surplus and of the government deficit all point to the natural solution of the `twin deficits problem'.

Japan

The Bank of Japan's monetary policy has tightened somewhat over the last three months. This is partly in response to the strength of the dollar and the fall in the yen to below 140 per dollar. The depreciation of the yen is a cause for concern because it adds to increasing inflationary pressures. These pressures appear to be the authorities' main concern. The underlying rate of increase in consumer prices is now around 3 per cent, with the Tokyo CPI for September giving signs that inflation may be increasing. Domestic wholesale prices have also been rising although they have been moderated by weaker metals prices. Wholesale prices are probably rising at 3 per cent per annum, and in August were 2.8 per cent above the same month in 1988.

The labour market has tightened sharply in the last quarter. Average earnings growth has accelerated, and the projected level of bonuses for 1989 is 7 1/2 per cent higher than that in 1988. As bonuses are such a large proportion of Japanese compensation we should expect to see rises of 6 1/2 per cent for the year as a whole. The unemployment rate has continued to fall, and has dropped to 2.2 per cent of the workforce. Overtime hours worked in manufacturing rose in both June and July, and survey evidence suggests that labour shortages are continuing to spread among enterprises. The job offers to applicants ratio has increased sharply in recent months, and settled at 1.34 during the summer months.

The second quarter's GNP figures pose some problems. Output declined by 0.8 per cent after a sharp rise of 2.3 per cent in the first quarter. Private consumption fell 1.3 per cent, mainly in response to the introduction of the first comprehensive indirect tax Japan has seen. Housing investment also fell, and housing starts are also down. These falls reflect bottlenecks in construction, as unskilled labour is drawn into the manufacturing workforce, and as business construction absorbs resources. The combination of rising interest rates and tight labour markets is likely to produce further falls in residential construction into 1990. Business investment has been remarkably buoyant, and we are projecting that it will grow by more than 16 per cent in 1989. Survey evidence from enterprises published by the Bank of Japan suggest that investment plans have been revised upwards during the summer. Orders received for machinery, a good leading indicator of investment one year ahead, are however, growing more slowly than of late, suggesting that investment growth will slow down into next year. This may reflect capacity constraints in the capital goods industry, where output has been growing at over 12 per cent at an annual rate for the last year. Unit labour costs in manufacturing have also begun to rise after a period of falls. This comes both from rising earnings growth and falling productivity growth. Output per worker had been rising at 4 1/2 per cent a year, but the growth rate has now fallen to 3 per cent.

Japanese output growth is expected to continue to slow down through 1989 into 1990 and 1991. Table 8 gives our forecast for Japan. As part of a process of fiscal consolidation, the Japanese government are keeping expenditure growth low, and this is reducing output growth in the immediate future. Consumption is expected to grow much more strongly in 1990 than in 1989, when it has been affected by the introduction of the consumption tax, and also by a drop in consumption in the second quarter in deference to the late Emperor Hirohito. The savings ratio has as a consequence, risen to a high level during 1989. Housing investment will probably decline further in 1990. All these factors, when taken together, are likely to reduce domestic demand growth from 5.6 per cent in 1989 to 4.9 per cent in 1990. There is likely to be some increase in domestic demand growth thereafter as housing investment recovers, and as government expenditure growth rises after the completion of a successful budget consolidation process.

The weakness of the yen during 1989 has boosted competitiveness, and exports of goods and services have been growing strongly. Table 9 contains our forecast of Japanese trade and payments. After several years of sluggish export growth and declining market shares, exports of goods have grown strongly in 1989, and this is likely to continue into 1990. As a result, we expect that GNP growth will exceed domestic demand growth in 1990, but with the resumption of yen appreciation export growth will decline, and import growth will stabilise. As a result GNP growth after 1990 will in general be 1/2 to 1/4 per cent less than demand growth.

The deterioration of the current balance in 1989 is largely due to the worsening in the terms of trade produced by the weaker yen. However, we expect this J curve effect to work its way out during 1990 and 1991, and we expect the current balance to improve from $65 billion in 1989 to $78 billion in 1991. Strong demand growth along with an appreciating yen will in the longer run reduce the current account surplus to under 1 per cent of GDP by the end of the 1990s. Much of the turnaround will come from growth in imports of services such as tourism. These imports have in the last few years been rising at more than 30 per cent a year in dollar terms. We do not expect this to be sustained, but the increasingly affluent Japanese population and the increasingly open society are expected to sustain considerable service import growth. Net property income will continue to rise as Japan becomes the largest net creditor in the world economy. However, continued appreciations will raise the dollar value of investment income debits, reducing the contribution of IPD to the balance of payments.

The more moderate level of domestic demand growth, along with stable real commodity prices, should, in combination with the appreciation of the yen, help to reduce inflationary pressures. We are expecting consumer price inflation as measured by the consumers expenditure deflator rather than the more volatile Japanese consumer price index to peak at 2 3/4 per cent in 1990 or 1991 as the effects of the recent devaluation feed through. Thereafter inflation should gradually decline.

Germany

The latest indicators suggest that the German economy is continuing to grow very steadily. GNP growth in the year to the second quarter was nearly 5 per cent. Investment and exports have grown even more strongly than in 1988, whilst consumption growth has slowed following the increase in indirect taxes enacted at the start of the year. It now seems likely that growth for the year as a whole will be 4 1/4 per cent, the highest since 1976. This performance has exceeded earlier expectations and prompted the German authorities to tighten their monetary policy in an attempt to prevent strong growth fuelling inflation.

Business survey evidence indicates that confidence remained high through the summer. The IFO index of business climate fell slightly in August but remained above the levels for January-May. The index for new orders in manufacturing showed a 6.7 per cent increase in the year to July. Domestic orders were up 5 per cent and orders from abroad up 9.7 per cent. The new orders growth was strongest in investment goods. These figures confirm the impression that growth in Germany has been stimulated by external demand, and in particular by demand for investment goods.

A background of continued high growth inevitably leads to concern over whether the economy is liable to `overheat'. In common with other major economies Germany has experienced two years of particularly high growth following five years that were moderate to good. Historically such a sequence has often been accompanied by rising inflationary pressures as the economy runs into capacity constraints. The German economy is currently displaying some evidence of these symptoms. Business survey evidence indicates that the rate of capacity utilisation in manufacturing has risen throughout 1988 and 1989. Unemployment fell from 8.9 per cent to 7.9 per cent between the second quarters of 1988 and 1989, whilst unfilled vacancies stood at 253,000 in August 1989, against 193,000 a year earlier.

Whilst such evidence suggests that the German economy is running closer to capacity constraints, there have been minimal effects on inflation. Consumer prices have increased by around 3 per cent this year, but part of this was due to the increase in indirect taxes at the start of the year. Wholesale price inflation reached 6.5 per cent in April and May, but declined to 4.5 per cent in August. It would therefore appear that capacity constraints have not yet been reached.

The Bundesbank's response to this situation has been to progressively tighten monetary policy. At the beginning of October the official discount rate was raised by 1 percentage point, following increases of 0.5 percentage points in January, April and June. These increases were partly a reaction to the increases in market rates that had followed from tightened credit conditions. This policy should dampen domestic demand and hence reduce the chances that the economy will overheat.

The latest rise in interest rates has also contributed to the recent appreciation of the DM. Between January and September the German effective exchange rate averaged 112.6, down 2.4 per cent on its average value in 1987. By the end of October it had regained its 1987 value, appreciating 9.3 per cent against the US dollar, 6.6 per cent against sterling and 4.7 per cent against the yen in a period of just six weeks. This appreciation will be welcome to the authorities, firstly because it should dampen external demand next year; secondly because imports should be cheaper, and hence inflation lower; and thirdly because it will contribute to the reduction of the German current balance surplus, which remains stubbornly high at around 5 per cent of GNP.

Our forecast for Germany is presented in tables 10 and 11. We anticipate that the German economy will in future face less favourable external conditions than it has recently experienced. The growth in export markets will slow to 5 per cent in 1990, compared to 7.6 per cent this year, and competitiveness will deteriorate as a consequence of the recent DM appreciation. Export growth is therefore expected to be much lower.

Domestic demand in 1990 should again grow at around 3 per cent, as increased consumption growth compensates for a reduction in investment expenditure. The increase in consumption is mainly attributable to the reduction in German income taxes which will be enacted in January 1990. The decision to cut taxes then was taken in 1987, and as 1990 is an election year, the cuts are only likely to be rescinded if the economy is experiencing considerable overheating. The decline in investment reflects the less vigorous growth in overall demand, and also the effects of this year's increases in interest rates.

Our forecast of slower growth in demand means that we anticipate that the risk of overheating will recede. Inflation should come down next year as this year's increases in indirect taxes cease to contribute to the annual rate. The currency appreciation and reduced aggregate demand will also help to alleviate inflationary pressure. These developments should enable the Bundesbank to reduce interest rates gradually, which should further encourage the growth of domestic demand in the second half of the year and in 1991.

We expect that the pattern of growth established in 1990 will be repeated in 1991. Domestic demand is forecast to grow at around 3 per cent, with some recovery in investment growth as monetary policy is relaxed. Net exports will however reduce the increase in GNP to around 2.5 per cent, as export market growth declines to 4.6 per cent, and as the DM continues to appreciate in line with the interest-rate parity condition. This appreciation together with demand growing in line with productive potential should allow inflation to remain at around 2 per cent.

In 1992 and beyond we are forecasting that GNP in Germany continues to grow at around 2.5 per cent per annum, which we think reflects the medium-term growth potential. Inflation should therefore stabilise at 1.5-2 per cent per annum. Domestic demand growth is expected to exceed GNP growth, and this will enable the current account surplus to decline gradually as a proportion of GNP.

The economic outlook for Germany in the 1990s will be considerably affected by the further development of economic cooperation within the European Community, and by political developments in Eastern Europe, especially East Germany. The effects of 1992 are discussed in box 3. We expect that the move towards a single European market will provide a significant stimulus to European growth, spread over a number of years.

The recent changes in Eastern Europe could also be beneficial to Western European growth, and especially to West Germany. This year so far the Federal Republic has accepted over 170,000 immigrants from Eastern Germany, with the prospect of more to come. There has also been around 360,000 ethnic German immigrants from the rest of Eastern Europe. This influx may cause short-term difficulties in terms of accommodation, but also represents an addition to the labour force. In the longer term this should be a supply-side stimulus particularly as many of the newcomers are young and reasonably well educated. However, in the short-term, such large inflows of migrants are likely to raise the level of aggregate demand, as they tend to be heavy consumers of government provided services.

Aside from the effects of immigration, Western Europe should also benefit in the 1990s from the proposed restructing of Eastern European economies. This will lead to increased demand for western technology as the east attempts to modernise its industry. This development should create opportunities for the western economies as a whole, but is liable to be particularly beneficial to West Germany because of its geographical proximity to the eastern bloc.

France

The French economy continued to grow at over 3 per cent per annum in the first half of 1989. There has however been some shift in the composition of expenditure. Investment grew by only 5.4 per cent in the first half of 1989, compared to 7.5 per cent in 1988, and government consumption was also lower. GDP growth was however maintained by a better net exports performance, particularly in the first quarter.

Business survey evidence suggests that confidence remains high, although slightly less buoyant than in the early half of the year. The rate of capacity utilisation in manufacturing continued to rise in the second and third quarters, but much less strongly than in 1988Q4 and 1989Q1. Consumer price inflation remained almost constant in the first eight months of the year, declining from a peak of 3.7 per cent in May to 3.4 in August. The unemployment rate in August was 9.5 per cent, virtually unchanged from February, whilst unfilled vacancies stood at 80,000 up 7,000 over the same period. All these indicators suggest that the growth of demand in the economy has been close to the growth-rate of productive potential.

French interest rates were raised in line with German rates at the beginning of October, in order to maintain the relative rates of return on short-term capital between the two countries. The finance minister, M. Beregovoy, has also stated that the government would look to maintain the current franc-DM parity in any future realignment of the EMS. In our forecast we have assumed that French interest rates remain 2 points higher than German rates, but there is a possibility that this differential will narrow further as markets believe the exchange-rate commitment has greater credibility or if it is expected that a monetary union is to be formed. At present France retains some capital controls, which are due to be phased out by July 1990, but it is thought that these remaining controls are of limited significance. (See box 1 on the EMS and Capital Controls.)

At the start of October the government presented its budget for 1990. The budget deficit is projected to fall from FF--100.5 billion to FF--90 billion. Spending is projected to increase by 5.3 per cent and revenues by 6.8 per cent. As the government is expecting nominal GDP to increase by 5.5 per cent, the effect of this package on aggregate demand is likely to be broadly neutral.

Table 12 presents our forecast. Domestic demand now looks likely to grow more slowly in 1989 than in 1988 as consumption and investment respond to the monetary tightening earlier this year, and to the anticipated slowing of economic activity. Net exports are however expected to add 0.5 points to GDP growth with exports benefitting both from the strength of external demand and the decline in the effective exchange rate in late 1988 and early 1989. That depreciation also accounts for the slight increase in inflation this year, as pressure-of-demand effects remain small.

We expect that GDP growth will decline to 2.6 per cent in 1990. Domestic demand growth will continue to slow down in response to the tighter monetary stance, whilst net exports will cease to add to GDP growth, principally because both weaker external demand and the recent appreciation of EMS currencies will contribute to a lower growth of exports. Appreciation will however enable inflation to decline again.

The forecast for 1991 and beyond indicates that the French economy can be expected to grow at around 2.5 per cent per annum, whilst maintaining an inflation rate of 2-2.5 per cent, slightly higher than we are forecasting for Germany. As with Germany, the move towards a single European market, the prospective move towards full monetary union, and the possible opening up of East European markets should all support the maintenance of a high rate of growth into the 1990s.

Italy

The most recent information on the Italian economy reveals that GDP grew by 3.9 per cent in 1988, slightly higher than we had thought, but indicates that growth in the first half of 1989 averaged only 3 per cent, mainly because of a worse net exports performance. Industrial production appears to have declined in both the first and second quarters of this year, although recovering somewhat in June and July.

Business survey evidence is a little more ambiguous, although also suggesting that economic activity may have been less buoyant in the early months of this year. Confidence in overall economic prospects fell markedly between November 1988 and March 1989, but had recovered somewhat by August, and a similar, though less dramatic profile is discernable from survey replies on the future tendency of production. Capacity utilisation in manufacturing fell in the first quarter of 1989, but rose in the second quarter to its highest level in the current cycle.

Inflation has been higher this year, with consumer price inflation registering 6.5 per cent in July and wholesale prices 6.4 per cent. Wholesale price inflation was however down from 7.1 per cent in March. In the latter half of the year inflation should benefit from the strengthening of the lira within the EMS which occurred over the summer, but which was reversed in October when the Bank of Italy did not raise Italian interest rates in line with the increases in other European countries. Labour market evidence also suggests that inflationary pressure remains muted. Hourly wage rates in industry had increased by 6.1 per cent in the year to August, but employment in industry decreased by 200,000 between the third quarter of 1988 and the second quarter of this year, and unemployment remains little changed at around 12 per cent of the total labour force.

The Italian budget for 1990 aims to reduce the public sector deficit to 10.4 per cent of GDP, by raising L11,500 billion extra revenues and reducing spending by L8,500 billion. This may be an ambitious target, but it is vital that it is pursued as vigorously as possible as Italian gross public debt continues to grow rather faster than GDP.

At the time of the budget it was also hinted that Italy may opt to put the lira within the narrow band of the ERM, possibly following the elimination of its remaining capital controls. The current exchange-rate policy appears to have gained credibility with the markets judging by the lira's strength over the summer. This allowed the authorities to forego an interest-rate rise in October, which would have had unwelcome consequences for the financing of the budget deficit. For our forecast we have assumed that the Italian-German interest-rate differential remains at around 5 percentage points, of which 2 points are a risk premium attributable to continued uncertainty about the sustainability of the exchange rate, and to concerns about the higher level of Italian public sector debt. Further increases in the credibility of exchange rate policy would allow Italian interest rates to decline more rapidly than we have assumed.

Our forecast for Italy is presented in table 14. As noted above, GDP growth seems to have decelerated in the early months of this year, and our projections suggest growth of only 2.7 per cent this year and 2.3 per cent next. Domestic demand growth slows in response to the projected fiscal tightening, and to the slight decline in real earnings this year. Net exports are expected to add to GDP next year, as the effective exchange rate declines and as domestic demand grows more slowly than demand abroad. Inflation should turn down again in response to lower growth in import prices and in trend unit labour costs.

In 1991 and beyond we expect that domestic demand will begin to recover as monetary policy is relaxed gradually, in line with other ERM participants. This should also allow some recovery in the growth rate of GDP, to around 3 per cent per annum. A continued commitment to exchange-rate stability should enable inflation to fall below 5 per cent. Inflation is however likely to remain above other European countries, leading to a gradual worsening of the current account, which may require further fiscal tightening in order to maintain exchange-rate credibility.

Canada

Canadian output rose by 5.6 per cent in 1988, driven largely by domestic demand growth of 6 per cent. After the current government was successfully reelected in November 1988, it introduced an austerity package, reducing government expenditure plans. The Bank of Canada had already been tightening its monetary stance. Short-term interest rates rose to 10.7 per cent in the last quarter of 1988, and firmed further in 1989. As a consequence we are anticipating that output growth will have slowed considerably in 1989. We are forecasting that GNP will be no higher in the last quarter of 1989 than it was in the first quarter of the year. The slowdown has been marked in consumption and housing investment, but as table 15 shows, business investment growth has also slowed from its precipitate rate of increase in 1988. We are expecting a further slowdown of domestic demand into 1990, with investment in housing actually falling, and some destocking taking place. Output growth should exceed demand growth as a result of the recovery of exports from their current level.

The tighter policy stance, along with the reduction in demand growth, are beginning to show signs of reducing the rate of inflation. The consumer price index rose only 0.1 per cent in both August and September. However, sharp falls in fresh food prices helped reduce aggregate price increases and the underlying rate of inflation does not yet seem to have fallen below 5 per cent. Labour market conditions remain very tight, with unemployment falling to only 7.3 per cent in September, 1/2 per cent below its level a year ago. Unit labour costs in manufacturing are also growing by around 2 1/2 per cent, and average earnings growth in manufacturing remains above 5 per cent.

Despite the lack of a large number of promising signs with respect to inflation, we do expect the output slowdown to have some effect, with consumer price inflation falling to 4 per cent in 1990. In the longer term, however, we do not expect inflation to fall much below this level. Canadian interest rates exceed those in the US, and as a consequence of our use of the open arbitrage condition in setting our exchange rate projection, we expect the Canadian dollar to resume its decline against the US currency. This is a reversal of the strong appreciation experienced in 1987 and 1988. The appreciation helped reduce inflationary pressures, whilst the depreciation is likely to increase them. We expect Canadian inflation to stay 1 to 2 per cent above that in the US, reflecting the exchange-rate path and maintaining competitiveness at around current levels. The strong appreciation worsened the Canadian current account balance, but appears to have been more than matched by long-term capital inflows. Although we are projecting that the balance of payments will improve somewhat, we are forecasting that the current account will stay in deficit.

World trade and balance of payments

World trade growth appears to have slowed down somewhat since 1988. Trade in manufactured goods, as measured by the manufactured exports of the 13 largest OECD members, rose at a rate of 10.1 per cent in 1988. This was well above average rate of 3 1/2 per cent experienced in the previous decade. We expect manufacturing trade growth to slow down into 1989 and 1990, and in the early 1990s slower output growth than recently experienced in both the US and Europe should mean even slower growth in world trade. However, the projected general buoyancy of the world economy should mean that trade growth accelerates in the second half of the 1990s. Our wider measure of world trade is based on the UN series for exports of all goods by all trading economies. We are forecasting that world trade on this basis will grow more rapidly, reflecting the growing share of world trade held by the Far East in particular. Table 16 gives our forecast of world trade as well as trade shares by the countries and groups on our model. This suggests that total world trade grew by over 10 per cent in 1988, and has slowed down to 8 per cent in 1989. The slight downturn in the cycle in world output that we are forecasting for the early 1990s reduces this measure of world trade growth to just over 5 per cent per annum, before recovering to 6 3/4 per cent in the second half of the decade.

Our forecast is predicated on the belief that there will be no major rise in world protectionism as a result of the persistence of current account imbalances. Table 17 gives our projections for world imbalances. The combination of strong growth in Japan and Europe and expanding services trade means that we are predicting that the pattern of imbalances will improve, albeit only slowly. The importance of services trade has been emphasised by GATT in its post Uruguay publications, and we are assuming the recent process of trade liberalisation will continue in this area.

NOTES

(1) See Bank for International Settlements Economic Paper no. 25, September 1985, `The US external deficit and associated shifts in international portfolios', by Michael Dealtry and Jozef Van't Dack for further details. (2) The Gramm-Rudman target in 1990 is $100 billion. However a leeway of $10 billion is allowed before any automatic cuts are invoked. Therefore faced with the official administration deficit forecast of $116 billion, Congress only needed to find $6 billion to avoid cuts. However, it is only in the absence of any agreement that the original $100 billion target comes into play and automatic cuts of $16 billion must be made.

ANNEX I. CHANGES TO THE WORLD MODEL(1)

The Institute's Global Econometric Model (GEM) contains over 300 estimated relationships and the processes of data revision and of structural change in the world economy necessitate continual revision to the model. (Many new equations are reported in the World Economy chapter of the Institute Review, but many others are also introduced without comment.) The model used for the current forecast contains 41 new equations, and these have changed the properties of the model in significant ways.

(a) New Equations (1) There are new investment equations for the US, Japan, France, Germany and Canada. They in general have a slower business investment response to rises in interest rates, both directly and indirectly than had the equations included on the last version of our model. (2) New equations have been introduced to allow for the effects of exports and imports of non-factor services on GDP and also of factor services on GNP. The services and IPD equations have been included on the model for over a year, and as we have judged their behaviour to be satisfactory, we have decided to integrate them fully into the model. (3) The trend unit labour cost equations have been replaced. These equations are central to the wage price sector of the model, as they affect wholesale prices and hence all other prices and costs. The new equations are of the form: ULT = [(ER/IP).sup.*][100.sup.*][e.sup.bt] where ULT is trend unit labour costs, ER is manufacturing earnings, IP is industrial production and b is an estimate of underlying productivity growth in the economy. These estimates are derived from work undertaken at the Institute by Julia Darby and Simon Wren-Lewis and are summarised in table AI below.(2) There are also new equations for consumer prices in the US, Japan, Italy and Canada and new wholesale price equations in Japan, Germany, Italy and the UK. These, along with unit labour cost equations have changed the price responsiveness of the model, accelerating the feed through of inflationary shocks. The characteristics of all these new equations are given in table AII. There are also additional changes to the Italian and UK sectors of the model, which have in the past displayed some instabilities in simulations over a ten year period. These have been reduced with the introduction of new employment equations for both countries, new earnings or compensation equations, and a new personal sector transfers equation for the UK. These additional equations are reported in table AIII.

(b) A current balance adjustment system None of the previous versions of our model ensured that current balance improvements in one country would be matched by deteriorations elsewhere. After some experimentation we have introduced a new system to ensure that this happens. (1) Trade volumes. The export equations on our model in general have exports depending upon imports elsewhere in the world. The presence of both time trends and non-unit elasticities on world trade has meant that earlier versions of the model would produce export growth some 1 per cent per annum greater than import growth. This tendency had to be corrected by residual adjustment during the production of forecasts. The model has now been adjusted so that export and import volumes grow at the same pace. (2) Trade prices. There were previously no inbuilt conditions to ensure that average export and import prices grew at the same rate. They now do so. This has changed model properties in a major way. In particular the response to commodity price shocks has altered. (3) Invisibles. Earlier versions of the model did not ensure that the invisible surplus of one country would be matched by invisible deficits elsewhere. This is now the case. Box A1 compares an oil price shock with and without the various parts of the current balance adjustment system. These balance of payments changes now ensure that the model is internally coherent. They do not, however, ensure that the world current account balances. Indeed, the world has been in deficit with itself for most of the last two decades. Table AIV gives our measure of the world's current balance discrepancy since 1984. The IMF in the October 1989 World Economic Outlook explain this discrepancy as the result of poor (and inconsistent) recording of invisibles, and especially of investment income.(3) They also project that the absolute value of the current account discrepancy will continue to grow, but that it is likely to decline as a proportion of the flows involved as the accuracy of the underlying statistics is continually improved. Our current forecast embodies this judgement, but it is possible for model users to adopt an alternative position. [Tabular Data 1 to 17 Omitted] [Charts 1 to 5 Omitted] [Tabular Data AI to AIV Omitted]

NOTES

(1) The work included in this version of the model has been undertaken by research staff at NIESR and the London Business School, and has benefited from the comments of economists at HM Treasury, the Bank of England and the World Bank. (2) Worktaken from J. Darby and S. Wren-Lewis, `Changing trends in international manufacturing productivity', NIESR, mimeo. (3) World Economic Outlook, October 1989, published by the International Monetary Fund, Washington D.C. See especially pp. 59-63, `Discrepancies in the World Current Account'.

BOX 1. THE EUROPEAN MONETARY SYSTEM AND CAPITAL CONTROLS

The European Monetary System

The European Monetary System (EMS) has been in existence for over a decade. It was designed to reduce exchange-rate uncertainty and improve the coordination of macroeconomic policies within the Community. Exchange rates do seem to have become less variable since the formation of the EMS(1), and the frequency of realignments (as reported in table 1) has declined, with only 4 of the 11 realignments taking place in the last five years. There has recently been much discussion of the role and significance of capital controls in the EMS. In particular there has been some concern that the progressive removal of capital controls will cause a major realignment of central rates, and this has been seen as a barrier to early entry by the UK into the exchange-rate mechanism. French and Italian controls are due to be removed by July 1990, and the UK, the Netherlands and the Federal Republic of Germany currently have no controls at all. The smaller nations, such as Spain, Greece, and Portugal have until 1992 (or even beyond if necessary) to remove their capital controls. However, the stability of the EMS depends much more on the lira and franc exchange rates than upon those for the peseta, drachma and escudo. This is not only because France and Italy are large wealthy nations, but also because they have well developed capital markets that may well react to changes in regulations. To analyze the likely effect of the removal of controls during 1990 we need to analyse both the factors which determine exchange-rate movements and also judge the scale and impact of any remaining controls.

Interest rates and exchange rates

In a foreign exchange market where there are forward-looking operators who wish to make profits we would anticipate that the bilateral exchange rate expected for a period ahead would differ from the current exchange rate by approximately the difference in the interest rates on the two currencies involved. If expectations are on average correct, then we would observe that in the past interest-rate differentials have been reflected in exchange-rate changes. There are a number of reasons why the `open arbitrage' condition for exchange rates may not have held in the past. In particular, there may be a systematic `wedge' between exchange-rate changes and interest-rate differentials. Table II gives the interest rate differentials between the major members of the European Exchange Rate Mechanism over the last decade and the changes in exchange rates for that period. The uncovered interest differential between France and Germany has averaged -0.3 since 1978, whilst that between Italy and Germany has averaged -3.2. The differentials have been larger in the second half of the 1980s. Systematic uncovered interest differentials of this magnitude can arise for at least three reasons: (1) There may be `risk premia' on some assets. Asset holders may have a preference for one currency rather than another, and may be willing to forego some return in order to be able to hold their preferred portfolios. As long as there is no operator who is both indifferent between the two currencies involved and who is also extremely dominant, then the market will produce a rate of depreciation for the `risky' currency that is less than the observed interest differential. (2) There may have been systematic expectational errors. If, in the case of Italy, the market had expected a higher rate of inflation (and hence a greater depreciation) than that observed ex post, then there might well be ex post uncovered differentials on this size. Although there is clearly a case for saying that Italian anti inflationary policy in particular was more successful in the mid 1980s than had been anticipated, the size and persistence of the uncovered differential cannot easily be explained in this way. (3) Capital outflow controls can also help raise domestic interest rates for a given rate of exchange-rate depreciation. The EMS has existed for some time with exchange controls, and they may even have been strengthened in the early 1980s. Capital controls are now being removed, and it is unlikely that they explain the differentials observed over the last few years.

Capital controls in Europe

The UK, Germany and the Netherlands have had no direct controls on capital movements for some time, and controls are minimal in Belgium and Luxembourg. (However, prudential regulations on the currency composition of institutional portfolios in these countries may have similar effects.) Denmark has already removed most of its controls, and the remainder are due to be removed by 1992. Until recently French companies faced restrictions on their foreign exchange dealings, and individuals still face restrictions on holding bank accounts abroad or in foreign currency (except ECUs). All controls will be removed by June 1990. Even when controls existed, it was always possible for companies to find channels through which funds could flow. The laxity of French controls may be reflected in the smallsize of the average uncovered interest differential. Italian controls have been more stringent. For the 32 years up to September 1988 there were considerable restrictions on all foreign currency operations of both banks and companies. All transactions had to receive permission from the Bank of Italy. Controls were considerably relaxed in 1988. It is now only required that all foreign transactions go through a recognised bank. It is also still forbidden to hold lira deposits abroad, to open credit lines to external residents, or to purchase short-term foreign securities. These restrictions will change portfolio composition but are unlikely to prevent large scale capital flows. The persistence of a large uncovered differential on Italian interest rates in 1988/9 in the absence of signficant capital controls by July 1990. However, the removal of capital controls, which has effectively already taken place, at least among the larger economies, has not destabilised the ERM and has not even necessitated a realignment of central rates. [Tabular Data I to II Omitted]

NOTES

(1)See R. Barrell and F. Eastwood National Institute Economic Review, May 1988 no. 124 and for a more detailed study see Artis, M. and Taylor, M., `Exchange rates and the EMS: Assessing the track record', CEPR Discussion Paper no. 250.

BOX 2. REVISIONS TO US GNP

The July 1989 issue of the Survey of Current Business contained significant revisions to the estimated contribution to GNP of the levels of exports and imports of both non-factor and factor services. This revision followed from the substantially revised current account estimates produced in the June 1989 issue of the SCB. The largest revisions were made in the other services exports category, which were revised upwards for 1988 by $28 billion after $18 billion in 1986 and $21 billion in 1987. These revisions followed on in part from improvements in the survey information on foreign visitors to the US, as well as in the method of estimating expenditure of foreign students' whilst they are resident in the US. There were improvements in both coverage and reporting of business service exports. There were also substantial upward revisions to estimates of profits from US investment abroad, reflecting the incorporation of information in the latest survey of direct investment. These revisions have a substantial impact on both GNP estimates and on the current account. Their interpretation is central to the analysis of recent developments in the US, and crucial to the construction of the forecast. Our model contains equations for exprots and imports of services in the balance of payments, as well as for foreign investment receipts. Estimation was undertaken on the unrevised data in 1988. The revisions have two implications. Firstly, if our income and relative price elasticities are correct, then the larger base volume of services exports will produce a larger improvement in the current account for any given improvement in competitiveness. Secondly, the revised estimates of services exports may lead to revised competitiveness elasticities. Unfortunately, the revisions to data on services exports has not been fully incorporated in the historical series. The Bureau of Economic Analysis suggest that the methodological revisions will not be fully incorporated into the historical National Income and Product Accounts until late 1990, and unless separate balance of payments series are produced with revisions back before 1986, 1984 or 1981 (depending upon the series involved) we will be unable to reestimate our equations until after that date. The pattern of revisions suggests that there may be a greater response to competitiveness improvements, especially in travel, than we have previously estimated, and we have adjusted our residuals accordingly. Further details of the revisions can be found in the June and July issues of the Survey of Current Business for 1989.

BOX 3. 1992 AND ALL THAT: THE INTERNAL MARKET AS A STIMULUS TO ECONOMIC ACTIVITY

In 1985 the European Commison published a White Paper, entitled `Completing the Internal Market', which aimed to remove may of the barriers to trade which still existed within the community. The White paper contained 279 proposals, and the Commission set 1992 as its target date for completion of the internal market. We are currently a little over half way through the programme for completion of the process in Brussels and at present the timetable is being met in terms of the number of measures that are passing through the system. However, many of the most important areas are still to be agreed. Progres has been fastest on the legislation for technical harmonisation and standardisation of industrial products and that on the abolition of capital controls is well advanced. But in the areas such as Commission has expressed its concern. However, progress in Brussels is a misleading indicator. The European legislation has in the main to be incorporated in the national legislation of the member states before actual implemnetation can begin and this process is well behind schedule with only a handful of the measures having been approved throughout the Community. The Commission's greatest achievement has perhaps been to instill an awareness that the 1992 programme provides increasing opportunites of a whole range of industries. It is less easy to quantify the scale of the benefits that might emerge, or indeed may already have emerged as companies seek to prepare themselves for the new economic environment. The Cecchini (1988) report suggests that the elimination of all types of obstacles to free trade between member states should led to substantial economic benefits, derived from the integration of fragmented markets, and increased competition stemming from easier access across nation state boundairies. These benefits cannot be questioned, but there is some doubt as to their likely scale. Table I contrasts the benefits calculated by Bakhoven (1989) and Emerson et al (1988). Dell and Mayes (1989) suggest that the benefits will be around half those claimed in the Cecchine report. Some commentators believe that the 1992 programme has already contributed to economic growth. Growth in Europe in 1988 and 1989 has been led by a surge in investment. The European Commission claims that the "investment-led growth process [has been] sustained by the positive expectations redited by the completion of the internal market' (European Commission, 1989), but they, of course, are not disinterested observers. The difficulty here is that investment has increased worldwide, in response to cyclical increase in world economic activity, which we suspect is largely independent of the 1992 programme. The investment equations on our model are able to explain the recent upturn in investment adequately, which accords with our judgement that the 1992 programme has had only limited effects on economic activity to date. We anticipate that the 1992 programme will lead to increased exports in the years to come, as firms in sensitive sectors increasingly adapt their industrial strategies to meet the new opportunities and challenges that the programme will promote. The European Commision has identified some of the main areas where the 1992 programme may be expected to have a particular impact. These include industries where there is a large element of public procurement, which has often tended to favour domestic industries. Examples here include railway rolling stock, energy production plant, information technology and telecommunications equipment. A second group is where there are national regulations of industrial standrards. Examples here include pharmaceuticals, electrical appliances and toys. A study in European Commission (1989) claims that some of these benefits have already begun to be realised but we feel the case has been exaggerated, since it appears that firms have only recently begun to consider the likely impact of the internal market and the strategic respnoses that it will entail. We therefore expect that the restructuring of European industry will take place mainly in the early and md 1990s, as the impact of the programme becomes more immediate and more clear. A research programme into this topic is currently being undertaken at the National Institute by David Mayes.