The world economy.

Anderton, R. ; Caporale, G.M. ; Veld, J.W. in't 等

is feared that this may exceed DM400 bn and the German debt stock may well rise to 60 per cent of GNP. This will lead to a sharp rise in interest payments and it is highly unlikely that the government can meet its deficit target. The cost of servicing this debt could well exceed DM30 bn a year. The Chancellor has recently warned that some form of tax increase, probably a petrol tax rise, may be introduced in 1993 to finance the cost of integration of services of the railway systems in the two parts of the country. This may have been a threat to the trade unions to agree on a solidarity pact with government and employers, but whatever the outcome of these negotiations, tax increases are no longer ruled out. Our The realignment of the ERM in September, and the subsequent withdrawal of the UK and Italy from the mechanism, have caused a degree of uncertainty in the financial markets. In our last forecast we were anticipating a realignment of the lira at the end of September (National Institute Economic Review No. 141) but we expected a more orderly adjustment of exchange rates than took place. We analysed the effects of a realignment in a note in the August 1992 Review(1), and we expected the consequences to stimulate activity in Europe. The uncertainties surrounding the exchange-rate outlook make our main case less sanguine than it would otherwise have been. The resulting prospects for slow growth in Europe have to be combined with a modest recovery in the US and Canada and continued below capacity growth in Japan. The success of Governor Clinton in the US elections bodes well for growth in the US, at least in the short term, and this may offset the poor performance elsewhere in the OECD. However, the distinct possibility of a trade war following on from the collapse of the GATT round cannot be ignored, and such an outbreak would make our forecast much more pessimistic.

The uncertainties surrounding exchange rates, and the changing prospects for monetary union in Europe, are central to our forecast. For the last three years our forecast has been based on the assumption that the European economies would make steady progress toward monetary union by 1997, but that there would be at least one realignment of the lira. The result of the Danish referendum in June persuaded us to move the projected date of formation of the union back to 1999, and recent events have made us more pessimistic. Our central forecast assumes that the core European economies, Germany, France, Netherlands, Belgium and perhaps Denmark and Austria, will form a monetary union by 1999. The other European economies are assumed to stay outside the union, at least over the horizon considered, even though we expect them to be members of an exchange-rate mechanism.

Our medium-term forecast depends upon the pattern of sustainable, or equilibrium real exchange rates in the advanced world, and upon the policies adopted by the authorities. We have constructed our forecast on the assumption that fiscal solvency is an important constraint on fiscal policy. We have discussed fiscal solvency and equilibrium real exchange rates in 'Wealth Effects and Fiscal Policy' (Barrell and In't Veld (1992)) and 'FEERs and the path to EMU' (Barrell and In't Veld (1991)). The second section of this chapter discusses our most recent estimates of FEERs and their implications for policy.

Short-term prospects

Our forecast for the short to medium term is set out in Table 1. There are now signs of a slow recovery in the US and in Canada after three years of recession or below capacity growth. Provisional figures for the 3rd quarter outturn for GDP in the US suggest that in that quarter at an annual rate growth was around 2 3/4 per cent after 1 1/2 per cent in the previous quarter. We are now forecasting that US output growth will be 2.0 per cent in 1992 after a fall of 1.2 per cent in 1991. The recovery is, however, moderate. Interest rates, at least at the short end, have fallen noticeably over the last two years in the US and Japan, and this should produce a boost to output. The extent to which lower short rates have fed through into lower borrowing rates and lower long rates is however, limited. Long-term interest rates have not fallen anywhere near as sharply, as can be seen by comparing Charts 1 and Charts 2. Borrowing rates have remained high in the US, and as a result the recovery in both housebuilding and business investment has been sluggish.

Chart 3 plots recent growth rates in the major four economies, along with our forecast. The pattern of the slowdown in activity in the early 1990s is similar in the US and France (and in Italy), but the US actually entered a recession. It appears that it is only in 1992, or late 1991 that the growth in activity has begun to slow down in Germany and Japan whilst recovery may have begun elsewhere.

Although the increase in the spread between short-term and long-term interest rates has been a world wide phenomenon over the last 3 years it has been most marked in the US and Japan. The Federal Reserve and the Bank of Japan have been attempting to stimulate demand, or at least stem the rate of decline in growth rates, but high public sector deficits in the US and the TABULAR DATA OMITTED difficulties of the financial system in Japan have helped keep long rates high. Although in some ways this increase in spreads can be seen as undesirable, even if inevitable, it has had a beneficial effect on margins in the banking sector. This is particularly important in Japan. The increase in output growth in Japan in the late 1980s was associated with sharp increases in equity, land and house prices. In part because the Price to Earnings (PE) ratio had risen to 50 in the Japanese stock market, it appears in retrospect that the increase in asset prices could not be justified by fundamentals. However, Japanese banks lent on the security of inflated assets, and also introduced them into their own portfolios. The fall in stock market prices over the last two years may have only brought them into line with real equity prices elsewhere, but it has resulted in a sharp liquidity squeeze on banks. If this squeeze were not to be eased then the fragility of the Japanese banking system would be a considerable worry.

The slowdown in the growth of output in Japan in 1991 was partly a response to the fall in asset prices. Housing investment fell sharply, and the fall in financial wealth helped slow down consumption growth by 2 per cent. In the short term we expect that the slowdown in the growth of activity will continue. A substantial fiscal package has been announced in order to aid recovery, but much of this can be seen as an attempt to assist the financial system through delays in the depreciation of capital assets and other related measures that are designed to bolster balance sheets. We expect the fiscal stance to remain tight in Japan, and in combination with a rise in the savings ratio in 1992 this should produce a balance of payments surplus in excess of 100 billion dollars in 1992 and 1993.

The process of German unification has been associated with a major fiscal expansion in Europe, and at least in Germany this has until recently more than offset the effects of high interest rates. The Bundesbank responded to the expansionary effects of the financing of the process of German unification by raising interest rates. The commitment to the Exchange Rate Mechanism on the part of the authorities in France, Italy and the UK required either that they raised their interest rates in line with those in Germany or in the case of the UK it meant that rates did not fall as much as might have been indicated by domestic conditions. In the February 1991 Review No. 135 p43 we argued that the net effect of the policy response to German unification would mean slower growth in activity in the rest of Europe because the effects of the appreciation of the ECU and of the increase in interest rates more than offset the effects of the expansion in trade.

The prospects for the European economies are also influenced by recent changes in the pattern of exchange rates. For some time we have assumed that there would be an orderly realignment of the ERM before a final setting of exchange rates, and we expected that realignment to be led by the lira. In the August Review (Anderton, Barrell, In't Veld and Pittis 1992) we argued that the countries that we expected to realign downwards would benefit from some short-term output gains, and in the short term their inflation rates would be somewhat higher than they would otherwise have been. This latter consequence was seen as temporary, because a once and for all realignment requires a once and for all adjustment of the same magnitude in the price level. An upward realignment of the D-Mark was seen as having a beneficial effect on German inflation, and as a result we felt that the Bundesbank would reduce interest rates, and the ECU would as a result fall against the dollar.

It is relatively easy to introduce an orderly realignment into a large scale econometric model of the world economy, but it appears to have been more difficult to manage in practice. The disorderly nature of the adjustments in exchange rates in September and the clear evidence that some authorities were unable to actualise their desired outcome appears to have had a significant, and deleterious, effect on confidence. Our forecast attempts to take account of the realignment and its consequences. The real D-Mark exchange rate has risen since our last forecast, and as a result, and despite a stronger than expected revised outturn for 1991, we have revised our forecast for West Germany downward, and we expect growth to slow down to about 1 per cent in 1992 and output to rise only slightly in 1993. The slower growth is expected to be associated with a reduction in the rate of inflation to below 2 per cent in each year, and it would be even lower if the VAT increase in the first quarter of 1993 were to be discounted.

The commitment to a fixed parity on the part of the French authorities has meant that the franc has also appreciated. Our work on wages and prices in Europe, reported in our August Review, indicates that inflation in France responds more quickly to shocks than it does in Germany. Although in the long run the effects of the realignment on the price level will be the same in the two countries, we are forecasting that inflation in France will decelerate more rapidly than in Germany, and the deleterious effect on output will also be greater in the short term. We are forecasting that French inflation will be around 1 per cent in 1993, and growth will fall to around 1/2 per cent.

The 14 per cent fall in the real value of the lira over the last three months was partly anticipated in our last forecast. Although we argued that the pressures for realignment had become too intense to resist we presumed that the authorities would wish to retain the effective anti-inflation discipline given by the ERM, and that they would follow the common practice of the mid to late 1980s and only remove half of any real misalignment. We calculated in 1991 that our world model indicated that the lira was 10-15 per cent overvalued, and hence we had built into our forecast a 5 per cent realignment to take place in the fourth quarter of 1992. The turbulence on the foreign exchanges and the subsequent loss of faith in the ability of the authorities has affected confidence, and we believe that the short-term gains from the additional fall in the exchange rate will be limited. Output growth in 1993 is unlikely to exceed 2 per cent, but inflation will rise only marginally compared to our last forecast. In the medium term there now appears to be a stronger possibility of a significant reduction in the scale of the public sector debt and a consolidation of the debt stock. If the policy package of privatisations and expenditure reductions that we discuss below is actually introduced then the process of adjustment to a sustainable exchange rate will have been enhanced.

In the very long term a realignment, which is just an adjustment of the price level, should have no macroeconomic consequences. However the asymmetries amongst the European economies in the speed of response to shocks of wages and prices imply that the short run consequences could be expansionary. The effects also depend upon the remainder of the policy stance adopted, but the two economies that have realigned downward have a tendency to react more rapidly to shocks than those that have realigned upward. In the short term we therefore expect the effects of a realignment to be expansionary. There also appears to have been a significant additional easing of the policy stance in the UK and perhaps in Germany, but the positive effects of such measures may be more than offset by the consequences of the disorderly nature of the process of exchange-rate adjustment.

World trade and commodity prices

The medium-term prospects for the world economy are in part indicated by the forecast for world trade growth given in Table 1. In the short term there is a significant divergence between the growth rate of total world trade which covers all commodities and all countries, and the growth rate of world trade in manufactures, which covers the exports of manufactures of the major OECD economies. This discrepancy in part reflects the growth of trade between non-OECD regions and also exports from non OECD to OECD countries. Although trade with East and Central Europe has declined, the 20 per cent increase in Chinese industrial production in the past year has given a significant boost to that country's exports. There have been a number of factors supporting world trade growth during the slowdown in world activity. The process of German unification probably raised world trade by around 1 per cent in 1990/91, and imports into the UK, the US and Canada have been stronger than would be suggested by their cyclical and competitive position. Given the effects of the depreciation of the dollar and the slow recovery in demand we would not have anticipated that US imports would have grown by 9 per cent in the last 12 months. We are forecasting that growth in US imports will remain strong in part because of the recent appreciation of the dollar, and that the same will be true for Japanese imports.

The current problems with the GATT round negotiations may well be soluable. A systematic trade war would have a significant deleterious effect on world welfare. Production and the pattern of trade are partly determined by comparative advantage, and the introduction of high tariffs would cause substitution into less efficient sources of production. In the long run world output would be reduced, and in the shorter term there could be considerable disruption, and output reductions would be even larger than in the long run. The effects will be less the more fluid the commodity. If the US imports less French white wine its price will fall in Europe, and so will output, but the US will be able to substitute with imports from Australia and New Zealand.

The GATT round negotiations have faltered over agricultural subsidies. If there is a successful resolution then we might expect free market food prices to rise, and the costs to EC consumers to fall. However, most commodity prices remain weak. The slow rate of recovery from the recent slowdown is a major factor behind our forecast for commodity prices. Chart 5 plots real commodity prices over the last decade, and Chart 6 plots more recent developments in nominal commodity indices. In the longer term we expect real commodity prices to continue to fall as improvements in extraction and production technology combine with the effects of a low elasticity of demand for the products. Food prices stabilised in 1992, and are expected to rise only slightly. Grain output has been affected by poor harvests in northern Europe and in Canada, but demand for imports of grains into Eastern Europe and the former Soviet Union has dropped by 5-17 per cent because of the liberalisation of food markets. This decline has particularly affected the price of coarse grains (maize, barley) used as animal feed because price liberalisation has removed excessive subsidies on meat production.

Agricultural non foodstuff prices including those of tropical beverages, have been very weak during the recession. A record harvest in 1991/92 pushed coffee prices down 24 per cent in 1992 to a 20 year low. Cocoa prices were very weak in 1991, but a production short fall in 1992 has increased them. The sugar market is in a similar position with production declining in Cuba as a result of economic disruption. Agricultural raw materials prices fell in 1991, but low cotton production in the US in 1992 has stopped the fall, and drought in Australia has reduced prospective wool production for 1993 and hence prices have stabilised. We expect only a small strengthening of prices in 1993.

Metals prices appear to be strengthening slightly ahead of a recovery in world output. Aluminium prices have risen throughout 1992 in part because of production cutbacks. Copper prices rose because of strong demand in the US and China and because exports from the CIS declined, but in the longer term increasing capacity should reduce prices. Demand in the US also strengthened lead and zinc prices as automobile output began to recover, but despite sharp price rises in the summer months stocks on the London Metal Exchange remain high and further rises should be moderate.

Our forecast for oil prices is given in Table 1 and the remainder of our forecast for commodity prices is given in Table 2. We expect oil prices to remain around 20 dollars a barrel in the short term despite the Saudis' desire to raise the price to the OPEC reference level of 21 dollars a barrel. Demand conditions are not strong, and there has been some competition from other fuels. Coal prices have fallen by 15 per cent in the last year, in part because of very heavy sales from the former Soviet Union. The need for foreign exchange by the former Soviet states has meant that they have kept up oil exports despite declining production, but home consumption has also fallen with the level of economic activity. Additional downward pressure on oil prices has come from the more rapid than anticipated recovery in Kuwaiti production to 1.0 million barrels a day, and also from the desire of the Iranians to demonstrate their sustainable capacity ceiling is 4 million barrels per day. Output has continued to grow, and in October OPEC output was more than 25.1 million barrels a day, at least 2 million barrels above the agreed quota. This should continue to moderate the effect on prices of a slight recovery in OECD industrial production.

TABULAR DATA OMITTED

Interest rates and exchange rates

The realignment of the ERM along with the recent appreciation of the dollar has changed the pattern of real exchange rates in the world economy. The Clinton victory in the US elections brings with it the prospect of higher government spending, and hence higher interest rates. As a result the dollar has appreciated, much as we suggested in the August Review.(2) (Box A analyses the effects of a rise in US Government spending.) The recent strength of the Yen appears to reflect a stronger outturn for the balance of payments than the market had anticipated. The appreciation of these currencies has alleviated some of the pressure on the European economies, where high interest rates had combined with a high exchange rate to slow down growth. However, the depreciation of the ECU basket came too late to remove the internal strains within the ERM.

We have been arguing for some time that the pattern of real exchange rates within Europe did not reflect fundamentals. Our calculations of equilibrium real exchange rates in 1991 suggested that sterling was 5-10 per cent overvalued and the lira was probably 10 to 15 per cent overvalued.(3) Relative inflation developments within the ERM since the start of 1991 probably worsened the situation for the UK and Italy. Over the weekend of 11th September the lira was realigned by 7 per cent, much as we and the markets had anticipated. The attempt to defend the sterling parity is discussed elsewhere in this Review, but the failure of the policy led to a substantial realignment.

Our forecast for exchange rates follows on from our forecast for interest rates, because we believe that in a market without capital controls a high interest rate will be associated with a depreciation, although an increase in interest rates will normally be associated with an appreciation. We have been using the open arbitrage, or uncovered interest parity, condition for forecasting exchange rates since 1988, and we see no reason to change our practice now. However, recent events have persuaded us, and to an extent the markets, to revise our assumptions about the process behind the determination of interest rates. For some years we had been assuming that a monetary union would be formed in Europe by the end of the decade. The Danish referendum in June made us less sanguine about this prospect. The scale of the damage to the process of union in September is unclear. Long-term interest rates have risen in Italy, suggesting to us that the chances that they will be able to join a union are considerably lower now than they were six months ago. The same conclusion cannot be drawn for France, where the successful defence of the D-Mark parity has been followed by rapid falls in interest rates. We now expect French and German short-term rates to converge by the end of 1993, somewhat earlier than we had previously supposed.

The appreciation of the D-Mark that has resulted from the realignment will inevitably relieve some of the inflationary pressures in the German economy, and it will also add to the already emerging slowdown in activity. The authorities have made slight cuts in official rates, but market rates have fallen to under 9 per cent, and rates one year ahead indicate that 3 month rates in Germany are likely to fall below 8 per cent by the end of 1993. Our interest-rate forecast is set out in Table 3. Our longer term projections are based on long-term interest rates on government debt, and these suggest further interest-rate falls in Germany and France.

Table 3. Short-term interest rates

Per cent

 GDP
 weighted
 US Japan Germany France average

1987 6.9 3.9 4.0 8.2 5.8
1988 7.7 4.0 4.2 7.9 6.1
1989 9.1 4.7 7.1 9.3 7.7
1990 8.1 6.9 8.4 10.2 8.3
1991 5.8 7.0 9.2 9.7 7.5
1992 3.7 4.2 9.2 10.3 6.1
1993 3.8 3.5 7.3 7.9 5.2
1994 4.8 4.1 6.8 6.8 5.4
1995 5.9 4.9 6.8 6.8 6.0
1996-9 ave. 6.8 5.2 6.8 6.8 6.4

1991 I 6.7 7.6 9.1 9.8 8.0
II 6.0 7.5 9.0 9.6 7.6
III 5.7 6.9 9.2 9.7 7.4
IV 4.9 6.0 9.4 9.6 6.9

1992 I 4.1 4.9 9.6 10.2 6.5
II 3.9 4.4 9.7 10.3 6.4
III 3.3 3.7 8.9 11.2 5.9
IV 3.4 3.7 8.7 9.5 5.7

1993 I 3.7 3.6 7.9 9.2 5.5
II 3.8 3.4 7.5 8.3 5.3
III 3.9 3.4 6.9 7.4 5.0
IV 4.0 3.6 6.8 6.8 5.0

Long term interest rates suggest that the interest differential between the US and Germany is only temporary, and hence there will be no reason for either currency to appreciate in nominal terms in the longer term. However, in the shorter term we expect the US to continue to appreciate against the D-Mark, as short-term US interest rates are markedly below those in Europe. Long term interest rates are also very revealing for other members of the EMS. Italian and UK rates are significantly above those in Germany, and especially in Italy they rose after the events of September. The disorderly nature of the realignment appears to have cost the authorities a degree of credibility, and we are now forecasting that interest rates in Italy will stay significantly above those in Germany, and as a result we expect the lira to depreciate continually over the forecast period. This in part reflects our belief, and that of the financial markets, that the costs of inflation reduction in Italy are both high and politically difficult to sustain. Indeed, our research reported in the August Review suggested that the Italians suffered from the highest sacrifice ratio in Europe, requiring large increases in unemployment for a one point reduction in inflation. Our exchange-rate forecasts, which are set out in Table 4, reflect our views on interest rates over the next decade.

In the longer term the process of adjustment towards macroeconomic equilibrium will erode away the competitive advantage that is temporarily gained after a realignment. A devaluation improves competitiveness, but the resulting switch into domestic expenditure will put upward pressure on prices, as will the rise in import prices. A devaluation will eventually raise the trajectory of the domestic price level proportionately, and the real exchange rate will eventually return to where it would otherwise have been. Table 5 presents our forecast for real exchange rates in the major economies. There have been significant changes in the recent past, but we are forecasting that they will be partially reversed. However, we believe that if a realignment in Europe had not taken place then eventually UK, Italian and French inflation would have had to have been lower than that in Germany for some years because their uncompetitive position would have cost them export and domestic market shares and hence would have depressed demand and output.(4)

Table 4. Exchange-rate forecasts for the major four

Percentage change in effective rate

 US Japan Germany France Italy Canada

1987 -12.7 7.8 8.4 2.0 1.2 -0.7
1988 -5.5 11.2 -0.6 -1.9 -3.4 6.1
1989 4.8 -4.2 -2.1 -2.1 0.0 6.4
1990 -6.8 -10.7 8.1 7.9 4.7 -1.7
1991 -0.7 8.6 -2.1 -2.7 -2.5 1.9
1992 -3.3 4.5 3.2 3.1 -3.0 -6.5
1993 1.6 6.4 2.0 0.4 -10.9 -3.8
1994 2.5 2.3 -1.1 -0.9 -4.5 -0.5
1995 1.1 1.6 -0.3 -0.2 -3.5 -0.4

 Yen D-Mark Franc Lira Franc Lira

 per Dollar per D-Mark

Nominal cross rates, year average

1987 144.6 1.80 6.01 1296.2 3.34 721.5
1988 128.1 1.76 5.96 1301.2 3.39 741.1
1989 137.9 1.88 6.37 1370.9 3.39 729.7
1990 144.8 1.62 5.44 1197.8 3.37 741.7
1991 134.5 1.66 5.64 1240.0 3.40 747.7
1992 125.9 1.55 5.28 1223.7 3.40 788.7
1993 120.0 1.56 5.37 1376.7 3.45 884.5
1994 119.4 1.61 5.57 1474.7 3.45 913.8
1995 118.6 1.64 5.65 1542.0 3.45 941.0
Table 5. Real effective exchange rates

Percentage change in effective rate

 US Japan Germany France Italy

1987 101.2 138.8 87.3 83.9 109.5
1988 97.2 149.6 84.9 81.9 109.2
1989 102.9 138.8 80.9 78.8 111.6
1990 98.0 121.8 88.8 85.9 121.8
1991 97.4 130.3 85.8 81.8 121.7
1992 94.4 135.3 90.3 84.3 120.4
1993 96.2 142.3 91.2 82.7 111.5
1994 99.6 143.5 87.1 79.6 111.5
1995 101.2 143.7 85.2 77.7 112.2

There are a number of approaches to the analysis of equilibrium real exchange rates, and the approach adopted depends in part on the time horizon over which the analysis takes place. In the long run one would expect trade to produce a world in which the prices of identical goods are effectively the same everywhere, and hence bundles of traded goods should have the same price everywhere. If this were the case then purchasing power parity (PPP) would hold. Although of great theoretical interest it is not clear that PPP has much use as a medium-term guide. OECD calculations suggest that both Germany and Japan were overvalued against PPP during the 1980s, despite their large current account surpluses. In a world of differentiated products and imperfect competition exchange rates may have to diverge from PPP for some time. If domestic considerations require a considerable degree of savings for an aging population, as in Japan, then large current account surpluses may be required as domestic productive investment opportunities are used up. A large surplus may require a real exchange rate below that which would deliver a zero current balance in order that home producers export and domestic market shares are increased. This would not be necessary in a world where home produced and foreign goods are perfect substitutes, but it is not clear that we live in such a world. Indeed, if we did we would expect large estimated trade elasticities in export and import equations, and they are generally agreed to be small.(5)

We have preferred to analyse exchange-rate equilibrium using a more medium-term concept. Williamson (1983) describes the concept of the Fundamental Equilibrium Exchange Rate, or FEER. We have used this in Barrell and Wren-Lewis (1989) and Barrell and In't Veld (1991). The FEER is the real exchange rate that would deliver, in the medium term, an equilibrium path for the current account given that the economy is also in internal balance. Its calculation requires equilibrium trajectories for income, commodity prices and the accumulation of overseas assets. We may write:

CB|R.sub.i~ = |f.sub.i~(|R.sub.i~, |Y.sub.i~, YW, A|S.sub.i~, P, T)

where CBR is the current balance as a per cent of GDP for country i, |R.sub.i~ is the real exchange rate for country, |Y.sub.i~ is its income, YW is world income, A|S.sub.i~ is the asset stock, P is commodity prices and T is time. The functional form, |f.sub.i~, is found using a large macro model, in our case NIGEM. We solve for the FEER by calculating trend values, denoted with a suffix T, for income and commodity prices, and we set target current balances that depend upon demographic and other factors. We can invert the functions |f.sub.i~, and solve simultaneously for all equilibrium real exchange rates R|T.sub.i~ ie:

R|T.sub.i~ = |h.sub.i~(CBR|T.sub.i~, PT, Y|T.sub.i~, YWT, A|S.sub.i~)

The trade and other elasticities that determine |h.sub.i~ are set out in the most recent NIGEM manual (NIESR 1992).

The recent realignment of the ERM has brought the UK and Italy much closer to their FEERs, as can be seen from Chart 7 and indeed in the case of the UK it may even have fallen below it. The same is true for Germany, as can be seen in Chart 8, but the French realignment has probably pushed the economy away from equilibrium, and this will increase the deflationary pressures that are already evident. Chart 9 plots the relationship between FEERs and the real exchange rate for the US and Japan. Recent exchange rate movements have also moved them toward equilibrium. Our results are sensitive to our assumptions, and we have assumed that in the medium term the UK, Italy and France would sustain current account balance, whilst Germany and Japan would sustain surpluses of one per cent of GDP, whilst the US would have a deficit of 1 per cent of GDP. The sensitivity of our results to our assumptions is analysed in Barrell and In't Veld (1991) where we conclude that a 1 per cent change in the German target, which would have to be offset elsewhere, would reduce the equilibrium exchange rate for Germany by 4 per cent, and hence if our estimate of sustainable medium-term outflows is too large because of developments in the East, then it is possible that the D-Mark may be slightly overvalued.

United States

The third quarter figures for GDP for the US make the recovery look more robust. Real GDP grew by almost 2 3/4 per cent (at an annual rate) in the third quarter, compared to only 1 1/2 per cent in the second quarter. Both real consumer spending and exports contributed to the stronger growth. A decline in the savings ratio to 4 1/2 per cent (from 5.3 per cent in the second quarter), rather than a rise in personal incomes, explains most of the third quarter's 3 1/2 per cent increase in consumers' expenditure. Although volumes of exports of goods and services grew by almost 2 per cent, import values rose 7 per cent in the third quarter, after growth of almost 15 per cent in the second. The deficit on trade in goods and services in the third quarter rose to over $50 bn, its highest level for two years. After growing by almost 30 per cent at an annual rate in the previous quarter, gross private domestic investment decelerated to only 4 per cent growth in the third quarter. A decline in residential investment of almost 18 per cent at an annual rate explains most of the deceleration. Although investment in producers' durable equipment slowed down from its second quarter growth of almost 25 per cent, it still rose by over 8 per cent at an annual rate. Both government expenditure (growing by 2 per cent) and stockbuilding (increasing by approximately $15 bn in 1987 prices) made a small contribution to GDP growth in the third quarter.

Several other economic indicators have been particularly weak since the end of June in contrast to the latest GDP figures. Non-farm employment has suffered a set-back, falling by 128,000 and 57,000 during August and September respectively. The ending of the emergency summer jobs programme had cut payrolls by 97,000 in September but had boosted them for August. In manufacturing, both average weekly working hours and overtime hours declined in September. The lack of activity in US industry was also highlighted by the decline in September in the National Association of Purchasing Managers' index from almost 54.0 in August to 49.0. The level of the index is moving closer to 44.0 which in the past has often indicated a contraction of GDP. The drop in the index probably reflects pessimism surrounding new industrial orders. After falling by almost 1 per cent in July, new orders for all manufactured goods declined by a further 2 per cent in August. Consumer confidence also remains depressed. The index constructed by the Conference Board of New York fell for the third consecutive month in September to 56.4 compared to 72.6 in June. Consumer credit outstanding fell again in August as attempts to adjust the debt/income ratio of the personal sector continued.

The annual rate of inflation, at just above 3 per cent, compares favourably with almost 5 1/2 per cent and 4 1/4 per cent for 1990 and 1991 respectively. This is in spite of the recent rises in import prices (a 1 per cent increase in June alone) caused by the fall in the dollar since May, but the latest upward trend in the dollar should have a restraining effect. Inflation has benefited from subdued average weekly earnings growth (in September earnings were only 2 per cent higher than a year ago) which have partly been restrained by rising unemployment which is now at 7 1/2 per cent. However, the sluggish nature of the recovery has prevented a substantial deceleration in unit labour costs.

The Federal Reserve has cut the short-term interest rate from around 10 per cent in the middle of 1989 to 3 per cent in November 1993. The Federal Reserve has indicated that further cuts in short rates are unlikely given that recent economic data have provided mixed signals and that GDP growth of between 1-2 per cent for this year is still feasible. However, monetary growth remains very weak with M2 only growing at an annual rate of approximately 1 1/2 per cent since the end of May which is considerably below the target range of 2.5-6.5 per cent.

Despite the magnitude of the loosening of monetary policy the recovery in output has been rather sluggish. Box B gives details of a simulation of a 1 per cent reduction in US interest rates. The simulation shows how much the economy is expected to respond to lower interest rates and also how long it takes for the effects to come through. Of course, depressed external demand has also subdued export growth and made imports more price competitive. The slowdown in world demand coincided with the transition from a robust US recovery to fears of a 'double-dip' recession. However, the US is not a particularly 'open' economy and the reasons for the fragility of the recovery are probably related to internal factors such as high long rates and personal and company sector debt. Although short rates have decreased substantially, long rates have remained fairly high. A reduction in long-term bond yields would boost the economy by encouraging business investment and also stimulate consumers' expenditure via wealth revaluations on equities and bonds. Long rates are high now because short rates are expected to be higher in the future and because the large public sector deficit is growing again. Although a further cut in short rates may stimulate demand and reduce the public sector deficit, the future possible inflationary consequences may put more upward pressure on long rates, and this may offset the effect of higher receipts. A fiscal package which incorporates cuts in long-term large expenditure items, such as federal health care and pensions, may help to reduce long rates. Private sector debt/income ratios may also be deterring expenditure. The financial liberalisation of the past two decades may have left consumers with higher debt to income ratios than they now wish to sustain. Chart 10 shows that over the past two years, US consumer credit outstanding expressed as a percentage of personal income has fallen substantially but that the ratio is still very high compared to historical levels. Chart 11 shows the US personal sector savings ratio and consumer confidence. Two things are apparent: first, that consumer confidence reached a very high level in the late 1980s. Second, this confidence combined with easy access to credit has resulted in an unprecedently low savings ratio.

Our forecast for US GNP is shown in Table 6. We believe that the substantial rise in consumption in the third quarter is unsustainable in the short term and that personal expenditure will only grow by around 2 per cent for both this year. Therefore, partly because of the degree of indebtedness of the personal sector, we expect the savings ratio to remain around 8 per cent rather than decline as the economy recovers. Investment, in both TABULAR DATA OMITTED the housing and business sectors, should continue to grow in the second half of this year. However, business investment will probably grow more strongly in 1993 due to the combination of the lagged effects of the interest-rate reductions and more optimistic expectations of future growth. Import growth has been strong recently, and we expect it to remain so in part because of the recent appreciation of the dollar. Our equation for US non oil imports has a higher short run than long run income elasticity, and its effect is reflected in our forecast of strong import growth at the same time as export growth is held back by slack world demand. Given these factors we expect US GDP to grow by approximately 2 per cent this year and just over 3 per cent in 1993. The delayed effects of the cut in interest rates and the fiscal package we are projecting for 1993 should raise growth next year. Beyond that year, GDP growth will probably decelerate slightly in response to the tightening of monetary policy we expect to begin in 1993.

The deterioration in the trade deficit in the first 9 months of this year partly explains why we expect the current balance deficit to be roughly 1 per cent of GDP both this year and next. Although the decline in the dollar during 1991 and 1992 resulted in a gain in competitiveness, weak export markets relative to domestic demand have caused imports to rise much faster than exports. In addition, the 'J-curve' effect has made the deficit worse this year. Transfers are also much lower than in 1991 as they were boosted last year by contributions from US allies towards the costs of the Gulf War. Table 7 shows our predicted path for the US current balance deficit. The more circumspect attitude of US consumers, and the resulting stable savings ratio lead us to forecast lower current account deficits in the 1990s compared to the 1980s.

TABULAR DATA OMITTED

Governor Clinton was elected on the basis of an economic programme that promised increases in top rates of income tax and public investment in education and the infrastructure. Furthermore, middle income earners would pay less tax and defence expenditure cuts would be increased. After the election, the Clinton camp have stated they will immediately implement a job creation plan consisting of a $20 bn infrastructure programme and investment tax credits for business. The proposals are designed to be consistent with long-term reduction of the budget deficit as extra jobs combined with an investment led recovery should boost tax revenues.

Our forecast of the US public sector deficit takes the above proposals into account and is shown in Table 8. TABULAR DATA OMITTED Our budget deficit figures are based upon the national income and product accounts (NIPA). These exclude financial transactions, particularly those connected with the rescue of the savings and loans institutions. But in our forecast we do allow for higher interest payments on the debt incurred from the failure of these institutions. Compared to the August Review we are now more pessimistic about the deficit in both the short and long-term. We believe that a Democratic President will find it difficult to implement medium-term cuts in the strong expenditure growth areas such as welfare payments, pensions and other social entitlements. In the short-term the jobs programme will worsen the deficit slightly. We expect the public sector deficit to be around 5 per cent of GDP this calendar year. This is largely the result of the effect of below trend growth upon both tax revenues and expenditure on benefits.

The deficit should be slightly smaller next year as GDP growth continues to recover, but higher interest rates on the stock of debt and increased expenditure will limit the extent of the reduction. In the medium-term the continued dissaving of the public sector may partly explain the steady savings ratio of the personal sector. That is, the expectation of future tax increases to pay for the higher deficit may cause individuals to perceive a higher proportion of their income as transitory. Therefore, income will be saved in order to pay for the expected tax increases.

Japan

The slowdown of the Japanese economy during 1992 has been more severe than we initially projected, and the authorities are clearly concerned, especially with the fragile nature of the banking system. At the end of August the Japanese government announced a 2.25 per cent of GNP package to stimulate the economy by additional public orders and relief for banks hit by the collapse of the real estate and share markets.

Real GNP rose by 0.3 per cent in the second quarter of the year, compared with 1.1 per cent in the first quarter, when it was boosted by a higher number of working days. GNP grew by 4.4 per cent in 1991, but for the current year much lower growth is expected. Industrial production picked up again in the three months to August and rose by 0.9 per cent, compared with a fall of 3.8 per cent in the previous three months. In August production was around 6 per cent lower than a year previously. Industrial output growth slowed dramatically, in 1991, when it rose by only 2.3 per cent, compared with 4.7 per cent in 1990 and 6.0 per cent in 1989.

Consumption declined marginally in the second quarter compared with 0.9 per cent growth in the previous quarter. Retail sales in the third quarter were down from a year ago. Disposable income growth has been less than in previous years, as wage increases fell and the special winter and summer bonus payments rose less than usual. There was also a distinct decline in overtime worked. In addition, households' wealth has been affected by the fall in real estate and equity prices, and we expect them to save more to rebuild their wealth stock. As a result consumer spending is likely to be depressed in the near future.

In the two years up until the end of 1991, equity prices fell by 25 per cent, and the fall continued into this year. In the mid-1980s, asset prices had been boosted by low interest rates and financial liberalisation. Speculative funds had driven an increasing divergence between asset prices and their fundamentals and the price to earnings ratio was very high. Tight monetary policy and high interest rates helped to bring about the downward adjustment that has occurred since then, but even now that monetary policy has been relaxed and interest rates have been cut, the fall in asset prices has continued. Investment has been severely hit by the collapse of the stock market. The fall in equity prices has led to a rapid deceleration in capital expenditure as the cost of capital has increased. Banks have seen their capital position eroded as they had invested large sums in equity and property markets. Banks have to build up their ratio of equity to total assets to meet the new rules by the Bank of International Settlements and as a result both bank lending and borrowing have been restrained.

In the second quarter business investment fell by 2.4 per cent, compared with 0.2 per cent in the first quarter. According to the Economic Planning Agency, manufacturing firms will reduce their investment this year, though investment may rise in the non-manufacturing sector. Other forecasts are more pessimistic and profit expectations have been falling this year. The index of leading indicators fell back to 25, after reaching 63 in July, the first time for two years that it exceeded the level of 50. The Bank of Japan's latest survey showed that business confidence in manufacturing has fallen in the third quarter to -37, its lowest level for over ten years. The indicator had fallen in the second quarter to -24, from -5 in the previous quarter. A similar fall was recorded in the non-manufacturing sector. Expectations for the fourth quarter had also worsened in both sectors.

Housing investment has fallen for more than two years now, in part because enormous falls in property prices have discouraged building activity. Housing investment fell by 7.7 per cent in 1991, but it has been stronger recently, and in the second quarter it rose by 2.5 per cent. There is cautious optimism that the decline in housing investment may have come to an end and housing starts may pick up again. Government investment, which accounts for a fifth of total investment, grew by 7.3 per cent in the second quarter, as public works expenditure was brought forward as part of a fiscal stimulus package.

Inflation rose to 2.2 per cent in September, up from 1.8 per cent in August but still below June's rate of 2.5 per cent. As the economy has slowed down, the labour market has weakened. In the Shunto, the annual spring wage negotiations, the wage increases averaged 4.9 per cent, compared with 5.7 per cent in 1991. The summer bonuses were only 1.7 per cent above last year's, while the winter bonuses increased by 3.6 per cent compared to a year ago. These wage rises are the lowest for four years. The fall in overtime worked also reduced the wage costs. This easing of wage pressure and the recent appreciation of the yen will reduce inflationary pressures.

The Bank of Japan has cut interest rates significantly over the last two years but banks have become more circumspect about their lending. Concern has been expressed about the risk of a monetary contraction in Japan. Money supply growth has been very weak recently and this has put the Bank of Japan under considerable pressure to reduce interest rates in order to stimulate the economy. The money supply, M2 plus CDs, actually fell in September, and with inflationary pressures receding after the recent appreciation of the yen, further cuts in interest rate are likely.

Our forecast for Japanese GNP is set out in Table 9. We expect GNP to grow by around only 2 per cent this year. That would be the lowest growth in Japan since 1974. This sharp slowdown is mainly due to the fall in investment, which, as discussed above, has been severely hit by the fall in asset prices. We expect business investment to fall by 2 per cent this year. The contribution of net exports to GNP growth is relatively weak, as exports have been held back by the slow recovery in Japan's major export markets. We foresee another year of relatively low GNP growth in 1993. Consumers spending will remain low as consumers will like to rebuild their wealth stocks. Furthermore, no strong boom from exports can be expected as Japan's competitiveness has been reduced by the recent appreciation of the yen.

Our forecast for Japanese trade is shown in Table 10. This year, with domestic demand depressed, imports TABULAR DATA OMITTED have weakened. The trade surplus is rising again as exports have picked up. In the 8 months to August, the trade surplus was $25.9 billion, compared with $21.8 billion over the same period last year. The trade surplus in 1991 was $78 bn. There is a trend of falling exports to the US, as many Japanese exporters have set up factories outside Japan to serve the US market. As a result the trade surplus with the US has risen much less than that with S.E. Asia and the EC. We forecast both imports and exports to weaken, as the Japanese economy slows down, whilst the recent appreciation of the yen is likely to offset the boost to exports from the recovery in the US. We forecast stronger export growth next year due to a stronger recovery in Japan's major export markets, but imports of goods are also liable to pick up. The current account will improve less than the trade balance, as the deficit on invisibles increases, due to the rising services deficit.

The stance of government policy is difficult to assess. The government had initially announced an historically small increase in government expenditure for this fiscal year. An additional package, presented in April, was widely regarded as inadequate, but a more substantial supplementary budget, announced in August, will be debated in parliament soon. This package includes additional government investment, a government land purchase plan and loans to encourage private investment. The government will issue debt-financing bonds, which represent more than 10 per cent of total revenue in the current fiscal year, compared with 7.6 per cent in the previous year. Despite this recent package, as can be seen TABULAR DATA OMITTED from Table 9 the fiscal stance in Japan has been relatively tight, and this, in combination with the danger of a monetary contraction, could mean that the prospects of a recovery are not good.

Germany

Output growth in Germany has slowed considerably in the last year after reaching a peak in 1990, when, boosted by unification, it grew by 4.7 per cent. Signs of a slowdown in economic activity became visible in 1991 when three consecutive quarters of negative growth were recorded, but that was after a very strong first quarter. For 1991 as a whole, real GNP grew by a relatively strong 3.6 per cent. More modest growth is expected in 1992.

The government has revised down its forecast for the eastern Lander and now expects real GNP there to grow by 2 to 5 per cent, after huge falls in the previous two years. The restructuring of the new Lander has turned out to be more expensive than initially envisaged and progress has been slow. The transfer payments required to support the east have had serious implications for public sector budgets. The implications for wage settlements and the inflationary pressures stemming from the extra demand have added to the concerns of the Bundesbank. They see their aim of bringing down inflation jeopardized by the strong growth in the M3 measure of money supply and have so far resisted pressure to lower base rates.

Real GNP fell in the second quarter of 1992 by 0.2 per cent, after a strong surge of growth of 1.9 per cent in the previous quarter. Growth in the first quarter was boosted by special factors such as the fourth successive mild winter and a higher than average number of working days. Domestic demand has weakened during the year and with interest rates remaining very high it is unlikely to pick up soon. The combination of slow demand growth in Germany's main export markets along with the recent revaluation of the D-Mark indicate that export growth will be slow in the coming year.

Consumers' expenditure fell by 0.9 per cent in the second quarter compared with 1.4 per cent growth in the first. After high consumer spending in 1990 and 1991, retail sales have weakened this year as spending was hit by tax increases imposed to finance the restructuring process in the eastern Lander. However, consumption is likely to pick up in the second half of this year now that the unification solidarity tax, imposed last July, has been removed. In addition, the full effect of this year's wage increases will come through later than previous years' because of the long duration of the negotiations. Also, spending is likely to be brought forward in advance of the VAT increase in January 1993, when the higher VAT rate will be increased from 14 to 15 per cent to harmonise with those in other EC countries.

Total investment fell by 4.2 per cent in the second quarter, after growing by 8.7 per cent in the first. Mild weather had boosted building activity in the first months of the year, but recent data suggest a decline in construction activity in later months. Capacity utilisation fell in the second quarter to 85.0, the lowest level for 4 years. Orders had risen a little in the first quarter, but have since fallen, reflecting the weakening of domestic demand. Export orders were hit by sluggish growth in foreign markets and by the appreciation of the D-Mark.

Unemployment in the western Lander rose to 1.91m in October, the highest since July 1990. The unemployment rate rose to 7.2 per cent. The unemployment rate in the eastern Lander had jumped to 17 per cent in January, when funding of various short-term working schemes was withdrawn. It has gradually fallen since then to 13.9 per cent in October, largely because of early retirement and special new work schemes. Total unemployment in the east is now 1.10 million, while the number of short-time workers has fallen to 240,000. But it is estimated that a total of around 1.7 million workers are either on job creation schemes and training programmes or have retired early. The dismal state of the eastern labour market is unlikely to improve. Productivity in the eastern Lander is much lower than that in the west, and the program of wage equalisation between the two parts of the country threatens to become a significant disincentive for much needed new investment.

Recent wage settlements in the west have been of the order of 5 to 6 per cent, well below the original wage claims by the unions. The government hopes to agree a solidarity pact between employers and trade unions to facilitate the restructuring process and this should lead to moderate wage increases in the coming years. Unions have asked for tax increase for top income earners in return, something the government has been unwilling to implement. It remains to be seen whether such a solidarity pact can be agreed upon.

The Bundesbank has been mainly concerned with setting its monetary policy in relation to the M3 money supply measure, which includes cash in circulation, current accounts and short-term deposits. The target range for this measure is 3.5 to 5.5 per cent growth for 1992, but actual growth has exceeded this range. In August M3 grew by 9 per cent, up from 8.5 per cent in July. In September M3 grew by 9.1 per cent, far less than was feared after the Bundesbank's heavy interventions during the ERM crisis. The Bundesbank spent a considerable amount to defend the lira, pound and French franc and a large amount of these funds flowed back to the German money markets. The bank's reserves rose by DM80 bn in September, indicating the sums the bank spent to support the ERM currencies. The risk of loosing complete control of money supply growth cannot be ignored because it is very difficult to sterlise such a large sum.

The Bundesbank's policy of focusing on growth in M3 has been criticised. In setting the target range for M3 growth, the Bank appears not to have made sufficient allowance for the effects of the ending of price controls in the east. In addition, the growth of M3 has been influenced by some special considerations, such as the expansion of credit to the eastern Lander. When short rates are above long rates funds are transferred to short-term deposits and this also increases M3. The Bundesbank recently has softened its emphasis on M3 and policy statements of bank council members suggest that the bank is drawing back from its insistence that the M3 measure of money supply should be regarded as the most important indicator of monetary stability. Other factors that would be taken into account are the recent currency inflows and Germany's changed international competitive position.

This shift of emphasis has led to speculation of forthcoming cuts in base rates, which are made more likely because inflationary pressures are declining after the recent revaluation of the D-Mark. Overnight rates are now 1.5 percentage points lower than at the start of September and capital market rates are below 7.5 per cent, the lowest since 1989. In October, the Bundesbank lowered the repurchase rate, a key money market rate. It issued a variable rate securities repurchase tender, through which the bank supplies funds to the domestic money market, at a minimum rate of 8.75 per cent. This rate was previously fixed at 8.90 per cent and has come down one percentage point since it reached its peak in early August at 9.70 per cent. The money market seems already to have discounted an interest-rate cut and our forecast incorporates this. We foresee no steep cuts in base rates as the inflationary risks have not receded completely. The costs of the restructuring of the east are soaring and the implications for public sector deficits are uncertain. High wage increases this year and further anticipated wage rises in the east are considerable inflationary factors that are likely to make the Bundesbank very cautious in reducing interest rates too quickly.

Our forecast for German GNP is set out in Table 11. We expect GNP to grow by only 1.1 per cent this year. This is largely due to lower investment and consumption growth, which have both been hit by the record high interest rates. Although we expect some consumer spending to be brought forward before the VAT increase, this is unlikely to boost consumption growth above 2 per cent for this year. Consumers' expenditure is unlikely to recover next year, and we forecast growth at much the same rate. Real GNP will only grow by 0.7 per cent in 1993, the lowest since 1982. Inflation has edged up slightly in the last few months and reached 3.7 per cent in October. But it is still well below the peak of 4.8 per cent it reached in March this year. The recent revaluation of TABULAR DATA OMITTED the D-Mark will help to moderate inflationary pressures of the restructuring process. We expect inflation to fall to 1.9 per cent on average next year, despite the increase in the higher VAT rate.

Germany's current account remains in deficit this year. The trade surplus continues to fall and the deficit on services and transfers has become larger. In 1991 the deterioration of the trade surplus was caused by the rise in imports after the boost to demand from unification. Now that domestic demand has weakened the trade surplus has been further reduced by the weakening of exports. Imports have fallen by 0.9 per cent in this year so far, reflecting lower domestic demand, and although exports increased over that period, they are expected to weaken later this year as the recovery in Germany's main export markets fails to materialise. Our forecast for German trade is shown in Table 12. We expect exports to be stagnant this year. Germany's exports are heavily concentrated in the high-quality capital goods sector and very sensitive to cyclical trends. Exports have been further hit by the recent appreciation of the D-Mark. In addition to the declining trade surplus, the invisibles deficit has doubled since 1990. Expenditure on foreign travel continues to rise, while net investment income has fallen as Germany's net external asset position has deteriorated and the large interest-rate differential between D-Mark and dollar assets persists.

The total public sector deficit seems likely to rise to TABULAR DATA OMITTED TABULAR DATA OMITTED more than DM110 bn this year. Although an improvement was visible in the federal budget in the first half year, this was mainly due to special factors which delayed expenditure. The one year solidarity tax on income and on corporation tax has now run out, and in combination with the consistently high level of transfer payments to the eastern Lander, the federal budget deficit is expected to exceed DM40 bn this year, but will stay below last year's deficit.

Tax revenues have been strong in the first half year, boosted by the solidarity surcharge, the substantial wage rises in the eastern Lander, which have raised the tax base, and also by the increase in excise taxes. More recently, the slowdown of the economy has led to a shortfall in budget revenues and extra savings have been announced. Despite planned cuts in spending on defence, family benefits and pensions, expenditure continues to grow, mainly due to public transfers to the east. These are now estimated to be approximately DM170 bn this year against DM140 bn last year. Until last month, the government had ruled out new tax increases and stressed that any new spending in the east must be financed by cuts in other expenditure. The total public sector deficit on a national accounts basis, which includes the deficit of the social security funds, is likely to rise this year to DM116 bn. The deficit of the social security funds will increase significantly this year, due to the introduction of west German pension legislation in the eastern Lander and the rise in expenditure faced by the statutory health insurance funds.

The government plans to reduce the public sector deficit to below 3 per cent of GNP by 1994, and this will only be achieved if taxes are raised substantially. Our forecast for the German public sector is set out in Table 13. This table shows the deficit on a national account basis, which includes the social security funds. We expect the deficit to remain around 4 per cent of GNP in 1993 and to see only a very slow reduction after that year. Prospects for the public sector balance sheet are made uncertain by the debt legacy of the former GDR. In 1994 the government must take on all the outstanding debt of the Treuhandanstalt and the GDR Debt Fund. Estimates vary, but it