The world economy.

Barrell, Ray ; Gurney, Andrew ; Veld, Jan Willem In't 等

CHAPTER II. THE WORLD ECONOMY

Prospects for the short term The start of hostilities in the Gulf in January appears to have removed some of the uncertainties surrounding the oil market, and oil prices have dropped to around 20 dollars per barrel. This development should help sustain growth and reduce inflation over the next two years. Box A sets out some calculations of the effects of the change in our oil price assumptions on our forecast. The appreciation of the D-Mark bloc and the emergence of a recession in the US driven by a wave of bank failures has persuaded us to be less optimistic then we were in our last forecast. Table 1 summarises the outlook. We are forecasting a slowdown in the rate of growth in the major economies in 1991, with some recovery in 1992 and thereafter. The slowdown has already taken place in the US, the UK and Canada, whereas in 1990 Japanese and German growth was at historically high levels. Chart 1 plots levels of capacity utilisation in the major economies. Only in the US has output clearly fallen below capacity, but record levels of utilisation in Japan, Germany and France inevitably imply some slowdown in growth from recent levels. We are projecting that Japanese growth will slow in 1991 and 1992 to around capacity growth, but the appreciation of the ERM currencies is projected to reduce growth in Europe. We are projecting French growth at only 16 per cent in 1991, and German growth is expected to be only around 2.5 per cent. The appreciation of the D-Mark block has caused the US to experience a major gain in competitiveness, and gains in export share along with a decline in import growth contribute positively to our forecast of US GNP, helping the economy to avoid a domestically generated recession.

Table 1 also gives our forecasts of world trade and industrial production. In the short term the failure of the GATT round is unlikely to have a depressing effect on trade, but we have stressed in previous issues of the Review that the longer-term prospects will be clouded if the dispute between the US and Europe over agriculture subsidies is not settled. We are projecting a rise in total world trade growth in 1991. This is driven by a combination of oil producers spending the revenue they have received from higher oil prices in the second half of 1990 and, also as unification in Germany raises imports. The slowdown in world activity in 1991 is expected to reduce the growth of industrial production, but from 1992 onward we expect it to resume growing at around 3 to 3 1/4per cent. In the longer term we expect world trade growth, which covers all goods and countries, to be around 7 per cent a year, whilst the manufactured exports of the major 13 industrial countries are expected to grow at around 6 1/2 percent a year. The differential between these two growth rates represents a combination of the long-term loss of trade share by the advanced world to newly industrialising countries and a slowly changing composition of world trade that will result from the projected relative decline in advanced world agricultural production. (a) Developments in Europe The authorities in Germany have continued their debate over the inflationary consequences of unification with the East, and interest rates have risen by in the last few months by almost a point in Germany whilst they have fallen by a similar amount in the US. As a result the D-Mark has appreciated by around 4 per cent since November. Since June of 1990 there has been a 15 per cent appreciation the D-Mark against the US dollar. The decision by the authorities in Italy and France to avoid an ERM re-alignment, along with the decision by the UK to join in October 1990 has meant that the European currency block as a whole has been appreciating against the dollar.

The unification of Germany has given a boost to growth both in the Federal Republic and elsewhere in Europe. The cost of unification has been much greater than was initially anticipated, and we are projecting German public sector deficits of 70 billion D-Mark in 1990 and 157 billion D-Mark in 1991. Table 12 below gives details of our German public sector forecast. The increase in demand has raised German imports, and we are forecasting a much reduced German current account surplus in 1991 and thereafter. Chart 2 plots the twin deficits in Germany both over the past and for the immediate future.

The emergence of the large public sector deficit in Germany has led the Bundesbank to raise interest rates, and in the short-term real interest rates in Europe, which are plotted on chart 3, have risen. In the longer term we would expect that the real rate of return on capital in Germany will rise because of unification. A large pool of unemployed but trainable workers will lower real wages in relative terms, although not absolutely, and profit rates will hence be higher. If as a consequence German real interest rates rise then, given the integrated nature of the European economies, real interest rates in the whole of Europe have to rise because funds will flow out of the rest of Europe into the new Germany. This would inevitably have a contractionary effect on output throughout Europe, but this could be offset by the greater level of demand generated by the new Germany. Box E below reports some simulations on our model that indicate growth in Europe outside of Germany is likely to rise by 0.1 to 0.2 per cent a year as a result of higher demand. However, the debate over policy formation and the fiscal stance in Germany has led the Bundesbank to keep nominal interest rates very high, and this in combination with the rise in the D-Mark block currencies has reduced growth by much more. Box D below attempts to calibrate these effects. We calculate that the combination of higher German imports, high European interest rates and the appreciation of the ERM exchange rates will lower French and Italian growth by 1/2 a percentage point in 1991, and by a similar amount in 1992. By 1993 Italian output is likely to be 1 1/2 percent lower than it would otherwise have been whilst French output is likely to be 1 per cent lower. The combination of high interest rates and a high exchange rate will however put downward pressure on inflation, and the authorities commitment to the ERM is likely to reduce inflation by between 1/4 to 1/2 a percentage point a year. If we treat the interest rate and exchange rate as beyond the control of the authorities in France and Italy, as indeed they would be in a monetary union, then the combination of impulses from Germany and the effects of the US recession are sufficient to explain all of the slowdown in activity in France and two thirds of that in Italy. We believe that the German fiscal authorities will tighten their stance, and that as a consequence interest rates in Germany will fall slightly. If this does not happen the outlook for Europe would be even more bleak than we currently foresee. In the short term German inflation is likely to exceed that in France, although it will be below the Italian level. This will produce a part of the real appreciation of the D-Mark that is needed to deal with the effects of unification on profits and the trade surplus. However, Europe faces the options of generating a German real appreciation by either allowing Germany to inflate more or by forcing the rest of Europe to deflate. The Bundesbank is opposed to an upward realignment. and is also opposed to higher German inflation. The real realignment that German unification necessitates appears to therefore be being produced by a recession in the rest of Europe that is generated by high interest rates and a commitment to the ERM. This may not be the least costly way to proceed. Tighter fiscal policies in Germany would allow interest rates to fall throughout Europe, and hence would avoid the effects of appreciation and high interest rates that are causing Europe to teeter on the brink of recession. (b) The slowdown in the US and the effects of the Gulf war The preliminary estimates of the outturn for GNP growth in 1990 in the US have been slightly worse than we had anticipated in November. The crisis in the financial sector appears to be having a strong impact on investment in business and on consumer spending. Institutions may have also become much more cautious about their lending habits. As a result both business investment, consumption and housing investment have been lower than they would otherwise have been. We have attempted in box B to analyse the effects of the slowdown in the US both in terms of its impact on that country and also in terms of its effects on the rest of the world. The combination of the depreciation of the dollar and the rapid slowdown in domestic demand has completely changed the prospect for the US current account over the next decade. We calculate that the recession will improve the US current account by over 1/2 a per cent of GNP in each of the next three years. A 10 per cent depreciation of the dollar will, we calculate, improve the current account of the US by a further 1/2 per cent of GNP after three years, although the initial effect is negative (at around 0.2 per cent of GNP in the first year). Box B shows the effects of a devaluation in the US current account. These two factors account for most of the 1/2 percent of GNP (in 78 billion dollars) improvement in the US current balance that we are forecasting will take place between 1990 and 1992. Much of the rest of the improvement in the US current account comes from the delayed effects of the payments that have been promised to the US for its role in the Gulf war. These amount to over $50 billion, but those paid in the US will count as balance of payments transfers and hence will have no effect on GNP. We do not believe that the Gulf war will provide much of a stimulus for growth in the west. in the short term we have raised our projection of US government spending by around $10 billion a year in 1991 and by $15 billion in 1992. These are small amounts relative to GNP (1/8 to a 1/4 per cent). Only extra current expenditure affects GDP. The use of military stockpiles has no national accounts counterpart as they are treated as current expenditure either in the year of purchase or over the period of construction. National income will only be affected if the stocks used up are replaced in future. We are assuming that only part of the military destocking will be replaced, and hence that the effect on GNP will be both small and spread out. These assumptions are based on a belief (and hope) that the war will be brief, successfully concluded and will have no adverse effects on the prospects for peace in the world.

Such a large turn around in the US current account, when combined with the autonomous rise in German imports means that the pattern of current account imbalances that we saw in the 1980s will basically disappear. However, as that pattern of imbalances involved some flows that are still sustainable we would expect it to re-emerge eventually. albeit in a more muted form. The German need for a current account surplus may well have disappeared, but the demographic and regulatory factors that drove the Japanese surplus in the 1980s are unaltered. (c) Oil and commodity market developments We have been arguing since last summer that the fundamentals underlying the world oil market should be putting downward pressure on prices. The onset of the Gulf war seems to have supported this position.

The fall in oil prices at the start of the Gulf war reflected the effects of the removal of uncertainty from the markets. The loss of iraqi and Kuwaiti production in the last five months of the year will have reduced world output by 4 per cent in 1990, and would reduce it by 7 per cent in a full year. Despite this output loss world output, at close to 65 million barrels a day, was close to 1979's record volumes. OPEC production rose by 5 per cent in 1990, with Saudi Arabia, Venezuela and Libya increasing their production markedly. These output increases took place partly to take advantage of higher prices, but also as a result of a political decision on the part of the Saudis to increase output to replace production from Kuwait and Iraq. Table 2 sets out our current oil price forecast along with our projections for other commodity prices.

The downturn in the growth of activity in the English speaking world is already having a depressing effect on commodity prices, and the slowdown in activity growth in continental Europe and in Japan will have a further depressing effect. Chart 4 plots recent commodity price developments. Food prices have been very weak in recent months because good harvests, especially in the USSR, have reduced the free market demand for wheat and other grains for animal feeds. Metals prices, except for aluminium, have also been very weak as the increased output that followed from the sharp rise in prices in 1988 has been met by declining demand. Aluminium prices were influenced in the short term by the closure of the Kuwaiti smelting capacity, and the transformation of bauxite ore into the metal is very energy intensive, and hence prices are sensitive to oil costs and energy. Chart 5 plots longer term commodity price developments along with our forecasts. All commodity prices have fallen in real terms in the 1980s, but we are expecting this process to be less severe after the slowdown in world activity in 1991 is reversed. Free market food prices depend upon surpluses generated in the advanced world, and the successful completion of the GATT round should reduce dumping of surplus products and hence lead to higher free market prices. Metals prices are however expected to continue to decline in real terms as improvements in extraction technology reduce the cost of production. Exchange rates and interest rates Exchange rates have been quite volatile over the last eighteen months. The yen depreciated by 10 per cent up until early May 1990 , and has appreciated by almost 20 per cent since then. The D-Mark effective exchange rate has appreciated more than 10 per cent over the last year and the Franc has risen with it, whilst the Lire has risen by less. Chart 6 plots recent exchange rate developments. These major changes in exchange rates are set out in table 3, along with our forecast. They seem to have been only partly influenced by interest rate changes. Table 4 sets out past interest rates along with our forecast and chart 7 plots recent weekly data. Rates have recently risen in both Japan and Germany by almost a full per cent over the last six months. In Germany in particular the adjustments that took place in early November and in February were described as largely technical, and closed a 'round tripping' loophole. The rise in both German and Japanese rates reflects worries about potential inflation. Chart 8 plots recent and prospective inflationary developments. The source of the inflationary fears is in both cases related to the strength of domestic demand, and in the case of Germany these fears are compounded by serious worries about the budgetary position of the public sector. Inflation worries in these countries were reflected in rising long term rates, whilst in the US, as can be seen in chart 9, long rates have dropped very slightly. The tightening in the monetary stance in Japan and Germany led to falls in long rates in both countries. These falls have been rather small in Germany, in part because long rates have risen in France and italy.

Our exchange-rate forecasts are given in table 3. We are assuming that the pattern of exchange rates will continue to depend upon the pattern of interest differentials, and therefore after 1991 the yen will in the longer term continue to appreciate against the dollar as Japanese interest rates fall back to below those in the US. The pressures for European monetary integration are building up, and we are assuming in this forecast that France joins the D-Mark block in the first quarter of 1994 at an exchange rate at the bottom end of its current band. The current inflationary problems besetting the UK and Italy lead us to believe that European monetary integration will proceed in two stages, with these countries, along with Spain joining a monetary union in the first quarter of 1997. This union is only likely to be feasible if some of the excess inflation in Italy and the UK is offset by realignments of their central rates. The number of realignments will depend upon the date at which the UK moves to narrower bands, but we believe that Italy will have to depreciate by more than six per cent before they can enter a union and we assume that the UK will join at the bottom of its current band. The formation of a union should remove most risk premia, but we believe that Italian and UK interest rates will settle slightly above those in France and the D-Mark block, reflecting both the slightly higher risk associated with these currencies and also the portfolio preferences of institutions. The United States The rate of growth of US GNP reached a cyclical peak of 4.5 per cent in 1989, and in the middle of that year capacity utilisation in manufacturing reached its highest level since 1980. Output growth has been slowing since then, and pressure on capacity has been dropping. Signs of a gradual slowdown in activity had been visible for several years. Private housing starts peaked in 1986, and by 1989 they had fallen by 25 per cent. Auto sales also peaked in 1986, and have fallen from 114 million a year in that year to 9.5 million in 1990. However, these falls have been gradual and indicators of the consensus forecast suggest that few commentators were expecting the sharp downturn that we have seen in the last few months.

Preliminary estimates of US GNP growth in 1990 suggest a reduction to 0.9 per cent compared to 2.5 per cent in 1989. The outturn for 1990 is more than 1 per cent below the consensus forecast from June 1990(l). The Institute is projecting growth of only 0.7 per cent in 1991. The decline in growth has been almost entirely driven by lower domestic demand. Stockbuilding was negative in 1990, and reduced demand by 0.6 per cent, and consumption growth of only 10 per cent also had a very depressing effect on output. The external sector made a positive contribution of almost half of 1 per cent.

Much of the slowdown during 1990 can be attributed to the decline in output in the fourth quarter. Real GNP fell at an annual rate of 2.1 per cent, and consumption dropped at an annual rate of 3.1 per cent. Residential investment fell at an annual rate of 15.4 per cent compared to 19.8 per cent in the third quarter. Stocks fell by $16.3 billion, and private non residential investment had it largest fall for some years at 4.6 per cent. Consumer spending on durables was particularly badly hit, falling by 8.6 per cent at an annual rate. Auto sales fell by 800,000 at an annual rate to under 9 million a year. Most of this decline was accounted for by sales of domestically produced cars, and import penetration rose (to around 27 per cent) in the fourth quarter.

It is always difficult to find the cause of a turning point in economic activity. The proximate cause appears to be located in the personal sector. At the beginning of August consumer confidence as measured by survey evidence, turned down sharply. This could be seen as an early warning of the decline in consumption in the fourth quarter, and the change in attitudes appears to have been triggered by the Iraqi invasion of Kuwait. However, forecasters began to trim back their predictions for 1991 before the invasion of Kuwait, and there were already signs that the economy was turning down. The leading and coincident cyclical indicators produced by the Department of Commerce both peaked in June and July of 1990, whilst the lagging indicator peaked in October 1990. The widely used National Association of Purchasing Managers index peaked at 50 in May 1990 and declined steadily to 37.7 in January 1991.

It is of course possible that the sudden reversal of forecasters opinions has influenced the outturn for the fourth quarter. A batch of pessimistic forecasts may have further undermined the confidence of consumers and they may have cut their spending even further. Investment intentions and hence outturns may also have been revised down further (2) . Once a decline is under way it may be difficult to stop and hence the forecasts may become self fulfilling. Given that forecasters may affect outurns, it is difficult to judge just how severe a downturn is to be faced. A survey of investment intentions in October/November by the Department of Commerce showed that most industrial sectors had revised down their expectations since the June/July survey, and expenditure is only expected to grow by 0.4 per cent in real terms. However order books remained strong until November, although new orders have been erratic. In December new orders for durable goods were 7.7 per cent lower than a year previously.

Current indicators do not bode well for a quick recovery. Unemployment rose to 6.2 per cent in January, a full 1 per cent higher than in June 1990. industrial production in December fell for the third month, and was more than three per cent below its peak in September. Retail sales in December were 3.6 per cent higher than a year previously, but were 4.4 per cent at an annual rate below the level of the previous month. Housing starts in December were 22.5 per cent below the same month in 1989. This suggests that the decline in housing starts has been accelerating despite falling interest rates, and 1990 starts at 1.19 million were 13.3 per cent below those in 1989. The stock of unsold houses has been increasing in recent months. The very clear slowdown in demand growth and in the level of activity has been associated with some evidence that the rate of inflation has begun to fall. The consumer price index increased at an annual rate of 3.6 per cent in December but was 6.2 per cent higher than a year previously. This decline in the underlying rate of inflation was also indicated by producer prices, which fell by 6.6 per cent at an annual rate in December. However all of the fall in this latter index was due to lower energy prices. The producer price index rose by 6.6 per cent between December 1989 and December 1990. Employment has been falling since June 1990, with 640,000 jobs disappearing in three months to January, when employment fell by 223,000. Table 5 gives our forecast for the components of US GNP. We are not forecasting a year on year recession despite the rapid slowdown in activity in the second half of 1990. Domestic demand grew by only 0.5 per cent in 1990, and we are anticipating a fall in 1991. We are projecting that the destocking seen in the last quarter of 1990 will continue into 1991, reducing demand by over a quarter of a per cent. Recent and projected falls in employment are likely to reduce the rate of growth of compensation and as a consequence we are expecting real personal disposable income to fall in 1991. The effects of this fall are like to be partly absorbed in a fall in the saving ratio, and we are anticipating that consumption in 1991 will remain at its 1990 level. We are projecting a further decline in housing investment and a fall in the level of business investment. Chart 11 plots our residual judgements on our consumption and investment equations, and it is clear that we have not had to judgementally adjust our equations for these two categories in order to project a downturn. However, the increasing crisis in the US financial system has led us to make a downward residual adjustment to our forecast of housing investment. The Federal reserve has responded to the impending recession by cutting interest rates, and short rates are now 2.5 points lower than their peak level in 1989. This would normally stimulate some recovery in housing investment, but the current situation militates against this.

The crisis in the US financial system started in the Savings and Loans sector, and the widespread collapse of institutions is bound to have had some effect on consumer confidence despite the widespread insuring of deposits. The recent wave of bank failures is the most sustained since the 1930s, and it is more worrying than the collapse of the Savings and Loans industry. A combination of over-exposure in property markets and an increasingly competitive international market cutting operating margins has led to severe capital adequacy problems, especially amongst banks in New England. As a result banks have been more cautious in their lending behaviour, and there is increasing evidence that the differential between borrowing and lending rates has increased. This will allow banks to rebuild their capital bases, and also make it easier for them to meet the BIS capital adequacy guidelines. However, these developments have meant that the cost of capital has risen relative to the indicators we use in our investment equation, and the fall in the stockmarket over the last six months will have reduced the availability of alternative sources of funds.

Despite the problems in the banking sector and its implications for business investment we do not feel that the US is facing a severe recession. There are a number of factors influencing our forecast. The Federal Reserve Board has made it clear that it will not repeat the mistakes of the 1930s, and it is allowing interest rates to fall to offset the effects of the recession. The fall in US rates has been associated with a decline in the dollar, and the external sector is likely to be able to partly offset the effects of declining domestic demand. The onset of war in the Gulf produced a sharp decline in the price of oil. This will reduce inflationary pressures in the US and reduce the speed of the downturn. The change in our oil price assumption between November and February reduces our forecast of US inflation by a full point in 1991, and raises our projection of output by a half of a per cent. Finally the combination of the onset of the Gulf war and the possibility of a severe recession appears to have changed the political urgency of budget deficit reduction, and the US Budget proposals published in early February make it clear that the authorities will allow automatic stabilisers to operate (3)

We are projecting a sharp turnaround in the US current account in 1991 and thereafter. Table 6 presents our forecast. There are three factors behind our projection (Box B attempts to decompose their effects). Firstly some of the $52 billion worth of contributions to the costs of the Gulf war from Saudi Arabia, Kuwait, Germany and Japan will enter the balance of payments in 1991. (Some receipts will then be paid out abroad, and we are assuming that not all will be paid in 1991). Second, the 15 per cent decline in the dollar over the last 18 months has improved US competitiveness, and this will allow the US to regain export market share. Import growth is also likely to slow both because of the competitiveness gain and also because higher oil prices will increase domestic oil production and hence reduce imports. The decline in the dollar will also raise net property income to a higher level than it would otherwise have been. Finally a US recession reduces domestic demand and hence lowers the level of imports. We are projecting rather low growth in the US over the whole of the next four years. This is in part because we feel that the impact of the banking crisis will be sustained. The combination of low growth with improved competitiveness is likely to produce a sharp decline in the US current account that the US is facing a severe recession. There are a number of factors influencing our forecast. The Federal Reserve Board has made it clear that it will not repeat the mistakes of the 1930s, and it is allowing interest rates to fall to offset the effects of the recession. The fall in US rates has been associated with a decline in the dollar, and the external sector is likely to be able to partly off set the effects of declining domestic demand. The onset of war in the Gulf produced a sharp decline in the price of oil. This will reduce inflationary pressures in the US and reduce the speed of the downturn. The change in our oil price assumption between November and February reduces our forecast of US inflation by a full point in 1991, and raises our projection of output by a half of a per cent. Finally the combination of the onset of the Gulf war and the possibility of a severe recession appears to have changed the political urgency of budget deficit reduction, and the US Budget proposals published in early February make it clear that the authorities will allow automatic stabilisers to operate(3)

We are projecting a sharp turnaround in the US current account in 1991 and thereafter. Table 6 presents our forecast. There are three factors behind our projection (Box B attempts to decompose their effects). Firstly some of the $52 billion worth of contributions to the costs of the Gulf war from Saudi Arabia, Kuwait, Germany and Japan will enter the balance of payments in 1991. (Some receipts will then be paid out abroad, and we are assuming that not all will be paid in 1991). Second, the 15 per cent decline in the dollar over the last 18 months has improved US competitiveness, and this will allow the US to regain export market share. Import growth is also likely to slow both because of the competitiveness gain and also because higher oil prices will increase domestic oil production and hence reduce imports. The decline in the dollar will also raise net property income to a higher level than it would otherwise have been. Finally a US recession reduces domestic demand and hence lowers the level of imports. We are projecting rather low growth in the US over the whole of the next four years. This is in part because we feel that the impact of the banking crisis will be sustained. The combination of low growth with improved competitiveness is likely to produce a sharp decline in the US current account deficit over the first half of the decade. We are forecasting that US growth will return approximately to trend after 1995, and that the current account deficit will deteriorate to around 1 per cent of GNP.

Table 7 contains our public sector forecast for the US, and this contains a sequence of large and sustained deficits. There are a number of special factors which we have set out on the table, but the recession is not the major cause of such large deficits. Box B attempts to calculate the effect on the public sector of a 2 per cent increase in GNP, but we do not think that the automatic processes are the only ones at work. Pressure had been building up to tighten fiscal policy in the US, but slow growth will reduce its political momentum. We have made a small allowance for the effects of the Gulf war, but we are still expecting military spending to continue to decline. The acquisition of defence equipment is treated as current government consumption in the years when programme expenditures were undertaken. The construction and delivery of a Stealth bomber has affected GNP for some years in the past. Its destruction, or its use of bombs, has no national accounts counterpart. Programme expenditures may rise in future, with some rising very quickly, but if the Gulf war is short and successful there may not be a great need to raise defence spending. Japan Japanese growth has been strong over the last year, and there are no signs of a marked slowdown. We are projecting that growth will slow from 5.5 per cent in 1990 to under four per cent in 1991. The slowdown is largely the consequence of the rise in interest rates we have seen in the last two years. The Gensaki rate has risen from 4.2 in the first quarter of 1989 to around 8 1/4in the first quarter of 1991. However, long rates have been falling recently, and ten year government bond rates are now below 3-month rates. The authorities have been pushing up short rates because of their fears for rising inflation. Their increased resolve appears to have caused the markets to revise their inflationary expectations downward. This has in turn led to a fall in long rates.

There have been clear signs of domestically generated inflation, and the rise in oil prices will have added to the worries. Consumer prices in Tokyo were rising at an annual rate of 2 3/4 per cent in the third quarter and 3 3/4 per cent in the fourth, and prices in January 1991 were 4 1/4 per cent higher than the same month in 1990. This is the highest rate of inflation seen for ten years. National consumer prices have been showing a similar growth and acceleration. Only a part of this acceleration can be put down to higher oil prices, as wholesale prices were rising at under 2 per cent in the fourth quarter, and around 1 per cent in the third. The strengthening of the yen in the second half of the year has almost completely offset the effects of the rise in oil prices on domestic inflation. Profit margins have been widening as capacity utilisation has been high.

There are other signs of the pressure of demand building up. Industrial production in the fourth quarter was 7 per cent higher than a year previously, although it fell slightly from its October peak in both November and December. Labour markets have also been very tight, and wage increases have been accelerating since 1988, albeit only to 5 per cent a year in the first ten months of 1990. More importantly trend unit labour costs have begun to rise sharply as rising wages meet the declining productivity growth that is always associated with the top of the cycle. There are a number of other indicators of continued labour market tightness. Unemployment, at 2.1 per cent, is lower than for most of the last ten years, and employment is two per cent higher than it was a year ago, although it is slightly below the peak achieved in the middle of 1990. The job offers to applicants ratio reached its highest level for 16 years in June 1990, but it has recently fallen from its peak of 147 to 143 in December. Labour shortages have in particular pushed up the cost of services, and this has been a factor behind the higher rate of increase in consumer prices compared to producer prices.

The third quarter of 1990 may well have been a cyclical peak, with GNP only 1 per cent higher than the previous quarter compared to 1.4 per cent the previous quarter. Consumer spending growth slowed down to 1.4 per cent compared with 1.6 per cent in the second quarter. Government spending has been flat throughout most of the year, but it is likely to be 2 1/2 per cent higher in 1990 as a whole than in 1989. Housing investment showed some recovery in the third quarter, rising 6 per cent, but by Japanese standards business investment growth of 1.8 per cent in a quarter is rather low.

We are projecting slower growth in 1991, especially in the interest sensitive personal income led sectors of housing and consumption. Business survey evidence of investment intentions is not overly pessimistic, and the Bank of Japan's survey of major firms undertaken in November even suggests some upturn in investment in 1991. Although we are forecasting investment growth of almost 10 per cent this will be partially offset by some destocking, and we are anticipating that domestic demand will only rise by 3.9 per cent in 1991, compared to 6.3 per cent in 1990. Our forecast for 1991 and thereafter is set out in table 8.

The sharp fall in the yen followed by its recent appreciation have been rather hard to explain. The effective exchange rate is still almost 10 per cent below its peak in the first quarter of 1989, and because Japanese wages and prices have been rising less rapidly than those overseas there has been an even larger gain in competitiveness since that quarter. Net exports of goods and services, along with net property income from abroad, had a negative effect on GNP in 1989 and 1990, partly as a consequence of the previous appreciation but also because of changes in the structure of exports and imports. Our work on structural change in our Japanese trade equations which we published in the August Review suggests that annual export growth has been around 1 1/2 per cent lower since 1985 than it would have otherwise been, and imports have been higher. The structural change has helped offset the improvement in competitiveness over the last year, and we would judge that the yen is on a sustainable path after firstly appreciating too much over the period 1987 to 1989, and then over-compensating as the decline in the surplus during 1989 left her structural outflows well in excess of the current balance. Table 9 gives our forecast of the Japanese external sector. The slowdown in the rest of the world is likely to reduce the surplus below the 36 billion dollars seen in 1990, and the promise to contribute 3 billion to the US effort in the Gulf will reduce it further. In the longer term we see the Japanese surplus falling from around 1 per cent of GNP to around half a per cent. We have argued in the past that Japanese demographic developments through the 1990s are likely to sustain a current account surplus, and apart from the effects of cyclical differences between Japan and the rest of the world, we are projecting that the need to accumulate overseas assets will be met. There are some risks associated with our forecast of Japan. The recent appreciation of the yen may be associated with the decline in net outward portfolio investment from 113 billion dollars in 1989 to only 40 billion dollars in 1990. This decline has been largely caused by institutions repatriating existing foreign assets. In the November 1990 Review we discussed the risks facing the Japanese financial system. The collapse in the Japanese stock market has eroded its capital base, and the repatriation of foreign assets is in part designed to compensate for this. A further fall in the stock market or a more marked fall in property prices could easily lead to a recession generated by financial factors. Germany The process of German unification brought unique pressures on the German economy in 1990. Never before had a highly developed market economy been joined with a centrally planned economy. However while monetary union and political union were effectively introduced overnight, the process of economic integration will take much longer to achieve, and is likely to influence the stance of German economic policy for much of the coming decade. The experience of the last few months has merely emphasised that at present Germany consists of two distinct economies joined in political and monetary union. In the economy of the western Lander output and employment have risen strongly in response to increased demand from the eastern Lander. However the switch of demand towards west German products has meant a sharp decline in demand for goods produced in the eastern Lander, and consequently output and employment have fallen strongly in the east.

The unification of Germany has yet to be reflected in a unified set of economic statistics. Although this is largely attributable to the logistical problems involved in collecting information about the eastern economy, the contrasting performances of the eastern and western economies means that unified statistics would not be helpful in assessing the German economy. National accounts figures for Germany as a whole would mask both the strength of demand in the west and the depth of recession in the east and hence would be misleading to policy makers and commentators. Until the process of economic integration has eroded the stark contrasts that currently exist, the main focus of interest will inevitably be the western Lander which already have the developed market economy structure which has only just been introduced in the east.

The process of unification provided a strong impetus for GNP growth in the western Lander in 1990. The most recently available data shows GNP growth of 1.7 per cent in the third quarter, giving growth of 5.5 per cent over the same quarter of 1989. GNP growth in the third quarter was boosted by German economic and monetary union which was enacted at the start of July. West German domestic demand increased by only 0.5 per cent in the third quarter, but exports of goods and services increased by 6.8 per cent. For National Accounts purposes west German sales in the eastern Lander continue to count as exports, although the trade data is published on an all German basis.

Since much of the stimulus to West German GNP growth in the third quarter arose specifically from monetary union, we would anticipate that GNP growth in the fourth quarter of 1990 will prove to have been rather less buoyant. Nevertheless GNP growth between 1989 and 1990 is likely to have been around 4.5 per cent. This is the highest growth rate in the west German economy since 1976, and whereas the growth in 1976 followed recession in 1975, strong growth in 1990 followed GNP growth rates of 3.7 per cent in 1988 and 3.9 per cent in 1989. In the face of strong demand growth the west German economy has shown signs of overheating.

The IFO business survey indicates that the rate of capacity utilisation in manufacturing rose to 89.9 per cent in the third quarter for 1990, after a decline in the first half of the year from its peak of 90.0 per cent in the fourth quarter of 1989. New orders in October were 10.7 per cent higher than a year before, with orders from the domestic economy up 16 per cent and orders from abroad up by only 2 per cent. Conditions in the labour market also point to the danger of overheating. Hourly wage rates rose by 6 per cent in the year to November up from 4 per cent in 1989. Wage claims by public service and metal workers have risen to 10 per cent. Meanwhile unemployment has fallen from 7.9 per cent in 1989 to 6.7 per cent in November 1990, despite an increase in the civilian labour force caused by immigration from the east. As yet the increased pressures of demand have not fed through to inflation. Consumer price inflation in January 1991 stood at 2-9 per cent. Nevertheless the continuing high level of capacity utilisation and the escalating public sector deficits have prompted the Bundesbank to tighten monetary policy.

In the last issue of the Review we highlighted the need for a policy response in order to restrain demand growth. At that time the authorities were holding back because of the imminent German elections. With the elections over it has become clear that the government does not wish to tighten fiscal policy and hence at the beginning of February the Bundesbank increased its interest rates by 0.5 per cent. The discount rate now stands at 6.5 per cent, its highest level since 1982. We anticipate that the new level of interest rates will be maintained for at least 3-6 months, to give the Bundesbank time to assess whether the current monetary stance is appropriate.

Our forecast, presented in table 10, is that the current strength of demand will moderate. GNP is expected to grow by 2.5 per cent in 1991. Domestic demand growth is however forecast to be 4.8 per cent, unchanged from 1990. The slower growth of GNP is therefore due to a lower level of net exports. Three factors account for this. Firstly we are forecasting slower growth of demand in the rest of the world. The growth in German export markets is forecast to be only 2.8 per cent this year, down from 5.2 per cent in 1990. The second factor is appreciation of the exchange rate. Between the fourth quarter of 1989 and the fourth quarter of 1990 the effective exchange rate appreciated by 7 per cent, including an appreciation of 17 per cent against the US dollar and 12 per cent against the yen. As a consequence German industry has lost competitiveness vis-a-vis its major competitors. The final factor that accounts for the deterioration in net exports is the continuing strong growth of domestic demand, which combined with the effects of exchange-rate appreciation, will lead to strong growth in imports. Table 11 presents our forecast for the current account and box D discusses the effects of German monetary policy on the rest of Europe, whilst box E looks at the direct effects of unification on the rest of the world.

By the second half of the year we expect to see signs that the rate of growth is slowing down. A number of factors contribute to this. Firstly there is the less favourable external environment described in the last paragraph. Secondly the recent rise in interest rates should begin to restrain domestic demand. Finally the demand stimulus provided by unification was primarily a step boost to German demand, affecting the level of GNP rather than its growth-rate. We expect that there will be secondary effects contributing to higher German GNP growth in the medium-term, but these will only materialise when the economic reconstruction of the east gets fully underway.

Table 12 details our forecast for the German public sector deficit. Economic growth has been strong in 1990, and revenues buoyant. As a result the deterioration in the deficit may not be quite as dramatic as many thought at the end of last year, but we nevertheless expect a deficit of DM 157 billion for 1991. We think that a deficit of this size will require an increase in taxes and have assumed that indirect taxes are raised by DM 40 billion at the start of 1992. Such a fiscal response should also enable the Bundesbank to relax its monetary stance without fear of encouraging inflation.

For 1991 we expect inflation will average 3 per cent. Although domestic pressures remain strong, the appreciation of the exchange rate means that the cost of imported goods may fall. The increase in indirect taxes we have assumed for 1992 may temporarily increase inflation to 4 per cent next year. Should inflationary pressures prove more difficult to contain, we would expect the Bundesbank to adopt a tighter monetary stance. Our present forecast assumes that interest rates can be reduced by 0.5 percentage points in the second half of the year without adding to inflationary pressures.

The prospects that the east German economy will start to revive in 1991 appear slim. Having already absorbed the shock of monetary union in 1990, which resulted in a 40 per cent decline in east German industrial output, the east German economy now faces a decline in demand for its products from eastern Europe as a result of the reorganisation of CMEA trading arrangements in eastern Europe. From the start of 1991 CMEA trade will be conducted in hard currency, and will be more market determined. However the shock of unification has produced a large pool of unemployed and underemployed labour, which should start to attract capital investment in the east, which in turn will signal the start of east German reconstruction. We anticipate however that it will be the middle of the decade before recovery in the east is fully underway. France Real GDP is estimated to have grown by 1.3 per cent in the third quarter of 1990, after 0.1 per cent in the second quarter and .8 per cent in the first. The recovery in the third quarter is due to industrial output, investment and exports that picked up after a particularly weak second quarter. For 1990, real GDP growth is expected to be significantly lower than the growth rates achieved in 1989 and 1988, which were at a record high since 1976. The economic slowdown can be attributed to slower growth of consumer's and public expenditure as well as to a widening trade deficit. This latter factor is due to a combination of a world wide deceleration of demand and the impact of the rise of the ERM currencies on French competitiveness. Capacity utilisation fell back to 84.6 per cent in the fourth quarter, from 85.7 per cent in the third and the second record high of 86.3 per cent in the second quarter of 1990. Unemployment fell by 5,900 to 2.53 million in December, but has been rising over the year as a whole. The jobless total was up 0.8 per cent on a year ago. The unemployment rate averaged 8.9 per cent in 1990, the lowest since 1983. Consumer prices fell by 0.1 per cent in December and 0.2 per cent in November due to lower energy prices. Excluding energy costs, prices rose by 0.2 per cent in November and 0.3 per cent in October. Including energy, the annual inflation fell to 3.4 per cent in December, from 3.6 per cent in November and 3.9 per cent in October. On average, inflation dropped to 3.4 per cent in 1990 from 3.5 per cent in 1989. Excluding energy, it fell to 3.1 per cent from 3.4 per cent in 1989.

The trade deficit has deteriorated and is estimated to be FF51.4 billion for 1990, the worst since 1982, and compared with FF44.7 billion for 1989. This reflects the decline in exports due to the slowdown in overseas markets and the rise in the ERM currencies against the dollar. The widening trade deficit led to a further decline in the current account, which for the first 11 months of 1990 worsened to FF38.5 billion compared with FF21.3 billion deficit in the same period of 1989. This also reflected the sharp decrease in the service surplus and a marked deterioration in net investment income and transfers. The monetary authorities remain committed to a strong-franc policy and have resisted any pressure for a realignment in the ERM. After the cut in official interest rates in November, the authorities have not followed the upward trend in German rates, and this has led to a weakening of the franc within the ERM. However, without rising German interest rates there would probably have been scope for further cuts in interest rates. The government aimed to cut the 1990 budget deficit by FF10 billion to FF90 billion and plans to cut it further to FF 80 billion in 1991 which is 1.2 per cent of GDP. Further tax reductions have been announced, including a cut of the corporate tax rate for retained earnings to 34 per cent from 37 per cent, further reductions of VAT rates as part of the harmonisation of VAT rates within the EEC and cuts in taxes on some forms of savings. To reduce unemployment the government has announced a third package of measures to boost employment by reducing labour costs and offering cheap loans for small firms to encourage investment and job creation. Our forecast for the French economy is set out in tables 13 and 14. In 1991 we expect real GDP to grow by 1.6 per cent and domestic demand by the same percentage. Private investment is projected to grow by only .3 per cent, compared with 3.2 per cent in 1990 and 5.8 per cent in 1989. We expect French exports to benefit from higher domestic demand in Germany and to grow by 5.4 per cent, but imports of goods are expected to grow by only 2.5 per cent. For 1991 we forecast a continuation of the decline in the invisibles balance into a deficit and therefore the current account deficit will rise to 0.9 per cent of GDP in 1991. Inflation is forecast to fall below that in Germany. The combination of higher German imports but a strong D-Mark block based on high interest rates has probably reduced output growth by over half a per cent in 1991. Inflation will also have been reduced albeit by rather little. The strength of the ERM, along with the slowdown in the US are sufficient to explain the slowdown in French activity. Although part of the reduction in growth can be explained by high interest rates the commitment of the authorities to the ERM means that the course of monetary policy is dictated by developments in Germany. Hence we would judge that the decline in growth in 1991 is entirely due to external factors. Italy Italian GDP grew by 2.5 per cent in the first half of 1990, down from 3.2 per cent in 1989 and 4.2 per cent in 1988, which represented a cyclical peak. A number of signs indicate that growth may have slowed further in the second half of last year: industrial production in the third quarter was virtually unchanged from a year earlier; the rate of capacity utilisation in manufacturing fell to 78.1 per cent in the third quarter, its lowest level since the first quarter of 1988; and business survey evidence on prospects for the economy shows a rapid decline in confidence from a balance of +24 per cent as recently as May 1990 to a balance of -34 per cent in November.

The decline in Italian growth can be attributed to the combination of high interest rates and a rising real exchange rate. interest rates rose from 11.3 per cent in 1988 to 13.3 per cent in the first quarter of 1990. Some relaxation proved possible in the second and third quarters of 1990, following the adoption of a narrow band for the lira within the European exchange-rate mechanism, but lira weakness in the fourth quarter forced the authorities to raise interest rates again. The effect of this period of high interest rates has been to curtail the growth of consumer spending and investment, with the result that domestic demand growth has fallen from 4.7 per cent in 1988 to 3.3 per cent in 1989, and an estimated 2.4 per cent in 1990.

In 1990 the lira appreciated 12.6 per cent against the US dollar, around 18 per cent against the yen, and 4.5 per cent against sterling, and depreciated by 1.5 per cent against the D-Mark and 2 per cent against the French franc. However, as the differential between inflation in Italy and inflation in both France and Germany exceeds 3 per cent, the Italians suffered a rising real exchange rate against their major competitors. This loss of competitiveness will be a perennial problem for the Italian economy unless it is able either to achieve inflation convergence with the core ERM economies, or it is able to realign its exchange rate within the ERM. The problems were exacerbated in 1990 by the appreciation of the ERM currencies against both the yen and the US dollar. The rising real exchange rate has put pressure on Italian exporters. Relative export prices rose by 5 per cent in 1990, and this has led to a loss of export market share in 1990, which is forecast to continue in 1991.

Italian membership of the exchange-rate mechanism enabled Italy to achieve a rapid reduction of inflation from 15.2 per cent in 1983 to 5.0 per cent in 1987. However, in contrast to the French economy, it has not proved possible to completely eliminate the inflation differential with Germany, which narrowed to 2.8 percentage points in 1989, but appears to have widened to 3.8 percentage points in 1990, with Italian inflation increasing to 6.5 per cent. The adoption of a narrow 2.25 per cent band for the lira at the start of 1990 was intended to give added impetus to the government's anti-inflationary stance. One way by which inflation may be further reduced is through rising unemployment as a consequence of maintaining the value of nominal exchange rates in the face of adverse inflation differentials. The strategy risks political unpopularity which may bring pressures for the exchange rate to be devalued. in our forecast it is assumed that further devaluation of the lira will occur, but any such devaluation is unlikely to fully compensate the loss of competitiveness, in order to ensure that the government's anti-inflation stance maintains credibility. in order to reduce inflation toward German levels we believe that the authorities will have to raise taxes and reduce consumption.

Our forecast, presented in table 15, details the consequences of this strategy. GNP growth is likely to slow further in 1991, as high interest rates continue to bear down on domestic demand growth, and a rising real exchange rate causes further deterioration in net exports. As growth slows down unemployment will increase back towards 12 per cent. This should however alleviate upward pressure on wages and allow inflation to fall below 5 per cent this year, with a further decline to 4 per cent in 1992. As inflation comes down the deterioration in the real exchange rate will lessen and the authorities should be able to lower interest rates, providing a stimulus to domestic demand. This should allow some recovery of GNP growth in 1992, but growth will have to be held down for some years as the budget deficit is reduced from 10 per cent of GNP to around 5 per cent by the time we presume monetary union is formed in 1997. Canada The Canadian economy has contracted further over the last few months. Real GDP fell by 10 per cent in the third quarter of 1990, after a 1.2 per cent decline in the second quarter. Available data indicate a further decline in the fourth quarter. Tight monetary conditions, reflected in high interest rates, have led to a fall in domestic demand. Investment has fallen particularly sharply. The decline in output in the third quarter was exacerbated by strikes in steel, pulp and paper, and automobiles industries. Inventories have been declining in line with output, leaving the inventory sales ratio approximately constant. This should help moderate any subsequent slowdown in production. The current account deficit narrowed to 13.7 billion Canadian dollars in the third quarter from 16.8 billion Canadian dollars the quarter before. The improvement was mainly due to a reduction in net dividend payments abroad from abnormally high levels in the second quarter. Exports and imports of goods both fell in the third quarter, leaving the trade balance basically unchanged. This was in part a response to automobile plant strikes in Canada reducing exports of finished cars and imports of components.

Consumer spending rose by over 1 per cent in the third quarter, but survey evidence from the Conference Board of Canada suggest that consumer sentiment is deteriorating, albeit not so rapidly as in the US. This change in sentiment has been associated with more cautious financial behaviour on the part of households, and the debt-income ratio fell in the third quarter. Residential investment has been declining sharply this year, and the decline accelerated in the third quarter. Housing starts have declined from 223,000 in the first quarter to 164,000 in the third. This indicates a further weakening of demand in the fourth quarter. Business investment also fell for the third quarter in a row.

There are signs that the slowdown in activity is beginning to affect prices. Although the inflation rate increased again in the last three months of 1990, there were indications of easing inflation pressures. Excluding energy, inflation was 4.1 per cent on an annual basis in November. There were signs of a severe squeeze on profit margins developing as activity slows or declines. Corporate profits declined 20 per cent on a year-for-year basis in the third quarter of 1990. This is attributed to weak demand conditions and the continuing rise in labour costs at a rate close to 6.5 per cent, around 2 per cent faster than in the US. As a result the unemployment rate rose to 9.3 per cent in December, with employment losses in manufacturing and service sectors.

The monetary authorities have eased monetary conditions somewhat and reduced the official interest rates in December and again in January. This is consistent with the easing of demand pressure, but unlikely to ease conditions further to prevent any spill over of the recently introduced GST (goods and services tax) into higher inflation. The introduction of the GST could add 1.25 percentage points to the inflation rate in 1991 which we are expecting to reach 5/4 per cent. However, we expect it to decline thereafter, and the recent fall in long rates suggest that this view is widely shared.

Our forecast for the Canadian economy is presented in table 16. Domestic demand is forecast to fall by 0.2 per cent, and real GDP to grow by 0.6 per cent in 1991. We are projecting only a small rise in consumption as the decline in employment bites into real disposable income, and we expect both business and housing investment to fall in 1991. In the longer run we expect growth to pick up again and settle down between 3 and 4 per cent a year. The markets are expecting Canadian interest rates to stay high relative to those in the US and as a consequence we expect the Canadian dollar to depreciate against the US dollar in the longer run. This should help to boost net exports. The deficit in the invisibles balance is expected to worsen and the current account deficit is expected to be around 3 per cent of GDP in the longer term. NOTES (1) See for instance, the average growth rate for the US in Victor Zarnowitz's summary in Economic Forecasts. July 1990. North Holland. (2) Such observer/observed interactions are common in the social world. Karl Popper calls them an Oedipus effect. However we would not wish to claim too much for the Cassandras' of the forecasting world. (3) See the discussion of the US budget in The Financial Times on Tuesday 5th February 1991. BOX A. THE EFFECT OF LOWER OIL PRICES ON OUR FORECAST The onset of the Gulf war produced a drop in world oil prices, and we are now projecting that they will average around $20 per barrel in both 1991 and 1992. In November 1990 we were forecasting that oil prices would be 30.8 per barrel in 1991 and 24.8 per barrel in 1992. There are many factors causing us to produce a different forecast for the world economy this February, and the majority lead us to project lower growth and lower inflation than in our November chapter. The lower oil price will have had a positive influence on growth and a negative one on inflation. Table Al presents an analysis of the size of these effects. We have used our November forecast base and have undertaken a simulation with our February oil price assumption. We have assumed that exchange rates and the fiscal stance are unchanged, but that interest rates follow the rate of inflation downwards, (i.e. real interest rates are held constant). The reduction in inflation and increase in growth depends both upon the importance of oil (and energy) in the economy and on the speed of response to shocks of the wage and price system in the country. The output response is highest in North America, where high energy intensity and free energy markets mean that changes in oil prices have large and rapid effects. Conversely the output response is lowest in the UK. The UK is a major oil producer. and the North Sea is a high cost producer. The tall in oil prices will, we expect. reduce UK oil output, and this will partly offset the positive effects of lower oil prices on output. Inflation responses differ, with the largest impact effects coming in the US and Italy, and the lowest in Japan and Germany. US prices are particularly sensitive to oil prices, and the feed through is almost immediate. The Italian wage price system on our model is still the most responsive to shocks amongst the major seven, despite our attempt to model the changes in the Italian labour market that have taken place during the 1980s. Both Germany and Japan have an extremely good anti-inflationary record, and wage and price setters in these two countries seem to respond relatively slowly to nominal shocks.

World trade in 1991 rises by almost 1 per cent as a result in our changed oil price assumption, but this effect is not sustained beyond the second year as the change in the terms of trade against the third world reduces their ability to finance imports, and trade in manufactures eventually declines. Current balances are generally improved in 1991 as a result of this rise in oil prices, although both the UK and Canada have substantial oil exports (although small net exports of energy), and hence their current balances deteriorate slightly, BOX B. THE US RECESSION AND ITS EFFECTS ON THE WORLD ECONOMY The financial crisis in the US has been worsening banks', and companies', financial positions. It has also worsened the outlook for the US fiscal deficit. We have undertaken a simulation in order to gauge the effects of the slowdown in the US and elsewhere. We have raised US business investment by 2 per cent of GNP in the first quarter of 1991 whilst leaving the variable endogenous. This raises US output growth to 2/2 per cent in 1991 and over 2 per cent in 1992. Table B1 sets out the effects on the US, whilst table B2 summarises the effects on the rest of the world. The recession in the US is likely to reduce US inflation by 1/2 a per cent in 1991 and 3/4 per cent in 1992. The slowdown in activity and the consequent gain in competitiveness that the US is experiencing will improve the current balance by over 1/2 a per cent of GNP a year, or around $33 billion in 1991 and $42 billion in 1992. If there were no slowdown in activity imports of goods would be 7 to 8 per cent higher, but the effect of this on the balance of payments would be partly off set by higher world activity, which would in turn lead to 3/4 to 1 3/4 per cent higher US exports.

The US recession is likely to reduce world trade growth by 2 per cent in 1991 and by a further 34 of a per cent in 1992. As is clear from table B2, Canada would gain most from a US recovery, and Japan and Germany would in the short term benefit more than the rest of the other major economies. In the longer term the benefits would be approximately evenly spread, with output up by a 1/3 to a 1/2 per cent after three years. World inflation would also be slightly higher if there were no US recession. We have assumed that there would be no changes in exchange rates, and that real interest rates would be held constant. A looser monetary stance outside the US would increase the inflationary impact of a US revival, and the increase in inflation in the rest of the world would exceed the 1/5 to 1/3 of a per cent we are suggesting.

The US public and federal deficits would also benefit from an economic recovery. With an unchanged fiscal stance we predict that higher US growth would reduce the public sector deficit by $29 billion per year in 1991 and $45 billion per year in 1992. The Federal deficit would be reduced by $21 billion in 1991 and by $36 billion in 1992. Table B3 gives further details. These figures are rather small but they are in line with those given in the US budget in February 1991 which gives an adjustment of $32 billion to the Federal deficit for the cyclical position. Table 7 in the main text gives our estimate of the underlying structural deficit in the US, both with and without adjustment for all deposit insurance outlays. BOX D. GERMAN MONETARY POLICY AND THE SLOWDOWN IN EUROPEAN GROWTH German unification has turned out to be more costly than had been anticipated, and we are forecasting a public sector deficit of 70 billion D-Mark in 1990 and 155 billion D-Mark in 1991. The deficit, and the associated rise in demand, has put some pressure on German inflation. Although much of the increased demand can be met directly from abroad, the economy of the original Federal Republic is operating near to capacity. The German constitution guarantees the Bundesbank considerable autonomy in monetary policy, and the central bank authorities have become increasingly worried about the inflationary consequences of the emerging deficit.

The Bundesbank have been pressing the government to raise taxes, and have raised Lombard rates twice in the last six months. On both occasions they have been forced into the move by the existence of windows of opportunity for 'round tripping' in the short term money markets. They claim that these developments make it clear that the deficit is putting upwards pressure on the price of funds. The Bundesbank claims that it can only lower rates if the deficit is reduced.

Over the last year German interest rates have risen by almost a percentage point, whilst US rates have fallen by more than that. Over the same period the D-Mark has appreciated by 15 per cent against the dollar partly as a result of this change of the interest differential between the two countries. The other members of the exchange rate mechanism have attempted to maintain their parities against the D-Mark, and as a result their interest rates and exchange rates have been higher than they would otherwise have been. This inevitably involves them in some lost output, but also reduces their rates of inflation. These losses (and gains) should be set off against the expansionary effect of German unification. Those effects are discussed in box E below. We have undertaken some simulations using our global model in order to analyse the effects of higher interest rates and exchange rates. The first, reported in table D1 analyses the effects of a sustained 2 per cent cut in interest rates in Europe. Exchange rates are assumed to be fixed. Output is around 1/4 of a per cent higher in the first year, rising to 1-1 1/2 per cent higher after three years. The output gain from lower interest rates is highest in the UK. Inevitably inflation is also higher, although the assumption of fixed exchange rates leaves it only 1/4 to 1/2 per cent above base. Again the effect is greatest in the UK.

Table D2 reports on a second simulation where we have also changed our exchange-rate path. The 2 per cent decrease in the interest differential against the US is assumed to be maintained for four years, and we have therefore assumed that the dollar exchange rates of each of the European economies falls by 8 per cent in the first quarter, but thereafter rises against the dollar by 2 per cent a year. These assumptions reflect our beliefs that tight money policy in Germany has raised the ERM dollar rates by around 8 per cent, and also that exchange rates can be expected to move in line with interest-rate differentials so that higher rates of return can be expected to be offset by capital losses.

The combined cut in interest rates and the fall in the exchange rate raise output by 2/3 of a per cent in the first year in France and Italy, and by around 1 per cent in the UK and Germany. After three years output is higher throughout Europe by 1 1/3 to 2 per cent. The combination of higher output and a lower exchange rate raises inflation. Amongst the continental Europeans the effect is initially least in Germany, but after three years all three are experiencing inflation just under 1/2 a per cent higher than it would have otherwise been. The effect is greater in the UK, in part because its greater dependence on non-EC dollar denominated trade raises its import prices by 1/2 a per cent more in the first year of the simulation. Our model may be underemphasising the increased integration of the UK in Europe.

One may conclude that the inability of the German authorities to reconcile their differences over policy is probably costing Europe almost 1 per cent growth in 1991 and around 1/2 a per cent in 1992. However, a reduction of inflation by almost 1/2 a per cent has to be set against the output loss. BOX E. THE EFFECTS OF GERMAN UNIFICATION German unification has raised demand throughout Europe. The residents of the eastern Lander want western goods, and some of these can be purchased from West Germany. Some will also inevitably be directly imported. The combination of direct imports and higher demand in the former FDR is raising output throughout Europe. These output gains, along with the associated higher inflation, have to be set off against the output losses that have resulted from the appreciation of the D-Mark and therefore of the whole ERM block which we discuss in box D.

Unification has also resulted in the breakdown of the normal relationship between domestic demand, prices and imports in the old Federal Republic. Our national income and price data are still on an old FDR basis, mainly because all German data is not currently available. However, the trade data for Germany has already been converted to an all German basis. We cannot measure the level of increased demand from the former Democratic Republic, but we can gauge it. Much of the emerging Federal deficit involves direct transfers to citizens in the east, and eastern Lander are also borrowing heavily.

In our forecast we have raised our projection of German imports by 4 1/2 per cent in 1991, by 7 1/4 per cent in 1992 and by 9 1/2 percent in 1993.(1) This adds 3/4 per cent to the level of world trade in 1991 and a further 1/2 per cent in each of 1992 and 1993. The effects of higher German imports are mainly felt in Europe, and we project that after three years French exports would be increased by 2 1/2 per cent, as would exports from the UK, and italian exports would be increased by 3 1/2 per cent.

These increases in exports of goods, along with the effects on services trade, will raise output. Table El gives our projections. The effects are concentrated in Europe, and after three years output is about 1/2 a per cent higher than it would otherwise have been in each of the UK, France and Italy. Higher demand leads to higher inflation, at least in the short run, and table E1 also gives our projections for these effects.

Over the longer term the expansion of the market in Europe may lead to greater competition and specialisation, and hence may lead to a higher level of sustainable capacity output. These are just the sort of expansion effects that are discussed at length in the European Commission's report on the implication of the 1992 single market programme. However, their size is rather difficult to gauge, and we have not extended our simulation into the medium term. TABULAR DATA OMITTED