The world economy.

Barrell, Ray ; Anderton, Bob ; Veld, Jan Willem in't 等

Recent developments and short-term prospects Business cycle developments in the major seven economies have not been particularly well synchronised in the last two years. The US, Canada and the UK have been in recession, whilst Germany and Japan have been growing strongly. We are predicting that this asynchronicity will continue over the next eighteen months, with growth slowing in Japan and Germany whilst a mild recovery takes place in the English speaking world. However there are some signs that the recovery, at least in the US, is likely to be rather slower than we had anticipated in our August forecast. The Federal Reserve in the US has clearly been concerned about the most recent signals from the American economy, and it encouraged a cut in interest rates on 6th November. The Japanese authorities have also cut interest rates in this quarter, and we anticipate that there will be no substantial increases in German interest rates.

The rapid response of the US authorities is a major factor in the construction of our forecast which is set out in table 1. Our forecast for 1991 is based on three quarters of data for the US and two quarters for each of the other major economies. It would appear that it will be difficult for the US to avoid a year on year recession in 1991 with growth of -0.2 per cent, although some recovery is underway. The first half year has been strong in both Germany and Japan, but this in part reflects some measurement difficulties which make the seasonally adjusted first quarter data for these countries difficult to interpret.(') However, we are expecting growth to slow sharply in Japan and in the fourth quarter in Germany we expect that output will only just regain its first quarter level. The quarterly output paths for the 4 major economies are given in chart 1, which includes our forecast up until the end of 1992.

Capacity utilisation rates have been falling in the major 4 economies. The US reached its peak in 1989, and as chart 2 shows, the index has declined 10 per cent in the last two years. French capacity utilisation peaked rather later, but as with the US its peak can be seen as a clear cyclical turning point. The German and Japanese indicators reached their peak more recently, and in both cases the interpretation of the signs has been harder. Investment growth in Japan averaged almost 15 per cent a year from 1988 to 1990, and this has inevitably raised the capacity ceiling. German business investment growth has also been strong, and it will probably average over 9 per cent a year over the period 1989-91.

The slowdown in activity in 1991 has not been confined to the major three economies. Growth in France is predicted to be in the range 1 to 1 1/2 per cent this year, and our forecast for Italy suggests that growth there could fall below 1 per cent. Growth in the UK and Canada is likely to be negative in 1991. As a result of these widespread developments we are expecting growth in the major seven to slow down from 3.2 per cent in 1989 and 2.4 per cent in 1990 to only 11 per cent in 1991. We do expect some recovery of growth into 1992, and output in the major-seven is expected to grow by 2.2 per cent, whilst in the major four it will grow by 2.3 per cent. Both are below the average growth rate experienced over the period 1982 to 1987, indicating the slow pace of the anticipated recovery. Growth in most of the major economies is expected to be stronger in 1993 and 1994.

Our current forecast of growth for 1992 is more pessimistic than that published in our August Review. We have revised down our projection for major seven growth by over 1/2 a per cent. A longer.. although shallow, period of below capacity output has led us to revise down our inflation forecast for both 1991 and 1992. Chart 3 plots recent inflation rates in the major four economies. These appear to have passed their cyclical peaks in the US and France (and less clearly so in Japan), but tax changes introduced in Germany in July have kept the inflation rate high in that country. We are forecasting an inflation rate of around 3 per cent a year in the major 4 economies in 1992 and 1993. The major seven also includes some of the more inflation prone countries, and the average rate in this larger group will fall to around 3 1/2 per cent a year.

The slowdown in output growth is reflected in the other aggregate statistics in table 1. OECD industrial production is inevitably considerably more cyclical in its behaviour than is total output and we are expecting a fall in industrial production in 1991 followed by a rather more rapid recovery than that displayed by total output. World trade growth, however measured, is also more cyclical than output. World trade in manufactures (as measured by OECD exports at least) grew at around 9 per cent a year in 1988 and 1989, whilst total trade of market economies grew by more than 8 per cent a year over this period. We are expecting manufacturing trade growth to slow to around 2 1/2 per cent in 1991, and to recover only slowly thereafter. Total world trade growth as measured by the UN index of exports and imports of all market economies is likely to grow slightly more rapidly. These indices are not consistent with that now published by the IMF. The September World Economic Outlook forecast total world trade growth of only 1 per cent in 1991. The new IMF index now contains intra-eastern bloc trade, and the disbandment of COMECON and the collapse of planned trade flows has distorted the IMF measures considerably.

We expect world trade growth to recover only slowly. This in part reflects the sluggish nature of the upturn in output, but also reflects developments outside the OECD. We estimate that the imports of the formerly centrally planned economies fell in 1990 and early 1991, and will rise slowly over the next two years. This growth in part reflects the opening up of the smaller eastern European economies, and also depends upon an increase in grain deliveries to the former USSR. In the longer term we anticipate that the former centrally planned economies will be major recipients of capital, and this will allow their imports to grow more rapidly than their exports. We are forecasting import growth rates of over 10 per cent a year from 1993 onwards. However, this cannot be seen as a major support for world trade growth in the next five years. Developments elsewhere are more important. OPEC imports also appear to have been low in 1990 and 1991, in part because of the disruption caused by the Gulf war. The only strong source of trade growth outside the OECD appears to be the Far East. Both China and the rest of the Pacific Rim have been growing rapidly in recent years, and import growth has been strong at around 10 per cent a year. We expect that this rate of growth will continue, and that this will help world trade growth to recover.

Export-led recovery has been predicted for most of the European economies. However, these predictions (2) are difficult to reconcile with the forecasts produced using large scale multicountry models such as NIGEM. It is possible that Europe could experience such a recovery, but this seems rather unlikely in the current conjuncture. Import growth in 1992 is likely to be weak in North America and Japan, and European trade with the Pacific Rim is limited. An export led recovery could also be generated if the European economies were either gaining world trade share because of increased competitiveness or because they have a tendency to gain share in the long run. The former source of growth is unlikely to be available, whilst research at the National Institute, reported in Box A, suggests that amongst the major seven the UK, Germany, Italy and Canada are the only countries that have ever had a tendency to autonomously gain world trade share. interest rates and exchange rates Interest rates have been easing in recent months, especially in the US and Japan. Chart 4 plots recent 3 month rates on a weekly basis for the last year, and the recent reduction in the US is particularly noticeable. A wide gap has opened up between rates in the US and in Germany. This in part reflects the policy worries of the authorities, but it also indicates the scale of the increase in the demand for capital in Germany that has emanated from unification with the former GDR. The authorities in the US have been concerned with failing output, whilst those in Germany have been worried about rising inflation. German rates have therefore risen. As table 2 shows, rates in France (and in the rest of the 'hard' EMS countries) have risen in line with those in Germany. We are forecasting that French (and Belgian) interest rates will be the same as those in Germany by the second quarter of next year. This strong coupling of interest rates, which is discussed in Anderton, Barrell and in't Veld in this Review, now extends to the Netherlands, Austria and Denmark. It results from the commitment of these countries to a fixed parity with the D-Mark and to a potential monetary union in Europe. It does however make the German authorities' task more difficult.

Our model suggests that the deflationary impact of this rise in interest rates will come through two channels, directly through demand and indirectly through the exchange rate First it will reduce domestic demand directly. Consumption may fall, as might housing investment. If long rates rise then business investment would fall. All these channels are present on NIGEM, but they are rather weak for Germany. A rise in nominal short-term interest rates accompanied by a rise in long rates would cause bond prices and equity markets to be lower than they otherwise would have been, and this would reduce net financial wealth. This in turn would reduce consumption. Once again these channels are present on NIGEM, but as Barrell, Gurney and in't Veld (1991) demonstrate they are weak in Germany. This would suggest that in order to produce a given fall in inflation the German economy requires a larger rise in interest rates than its European partners.

If the D-Mark were not tied into the Exchange Rate Mechanism then a rise in interest rates would be expected to cause the currency to appreciate. This would directly reduce inflation through its effects on import prices and indirectly reduce it through the effects on aggregate demand of lower exports and higher imports. The ERM prevents the D-Mark floating up against its close trading partners because interest rates have become strongly coupled. Although, as chart 6 shows, the D-Mark has appreciated over the last two years it is now 5 per cent above the level achieved in mid 1989.

The combination of a hard ERM and high German interest rates has been a major factor behind the slowdown in Europe. We argued in the February 1991 Review that the effects of the high level of German interest rates more than outweighed the effects of unification which led to strong demand for exports throughout Europe. Our conclusions were that tighter German monetary policy in the last year has reduced growth in the rest of Europe by almost a full per cent, whilst the demand stemming from unification raised it by only 1/4 of a per cent.

The exigencies of the current ERM regime do, however, bring benefits for its members. if, as we anticipate, the Maastricht summit in early December 1991 produces a plan for monetary union, then in most European countries the prospects for inflation will improve. Our medium term forecast for exchange rates, which is set out in table 2, is predicated on a belief that exchange rates will become irrevocably fixed within a few years. We still believe that the italian authorities will be forced into one more realignment before 1995, but thereafter exchange rates will be fixed. By 1997 a monetary union with a common currency is assumed to be in place. This assumption in turn has implications for our forecast of long-term interest rates. In a monetary union (or indeed in any fixed exchange-rate system without capital controls) the rate of interest, net of individual specific risk premia, must be the same everywhere. Hence long rates must be the same everywhere in Europe. Convergence of long rates has to proceed more rapidly than that of short rates. Long rates are the forward looking weighted average of expected future short rates (plus a liquidity premium). Hence if short rates are expected to converge monotonically then long rates should be seen to converge. As Anderton, Barrell and in't Veld in this Review demonstrate, long rates have not yet converged in Europe. The difference between 5 year and 10 year long rates of interest currently indicate the expected short rate differentials between 1996 and 2001. Short rates are expected to be the same over this period in Germany, France, Belgium, the Netherlands, Denmark, Austria and even in ireland, whilst it is not currently clear that this is the case for the other ERM members. Our forecast in this instance departs from that of the average market participant in that we believe the Italian and UK short rates will have converged by the end of the century, and hence these countries will for some time benefit from lower long rates than the markets currently predict.

Long-term interest rates contain a great deal of information about the future, and the rise in long rates in the last quarter of 1989 is particularly significant in the analysis of the evolution of real interest rates. The ramifications of the collapse of the centrally planned economies are still being digested, but by the 9th of November 1989 (when the Berlin Wall fell) it was clear that in the longer term large scale capital flows would be required in order to enhance the process of redevelopment in the east. The timing of these flows remains uncertain, but we have estimated (National Institute Economic Review no. 137 pp. 40-41) that the scale of flows required will only be forthcoming it real interest rates in the rest of the world rise by around 1/2a per cent. This will reduce investment and consumption in the advanced world. This is necessary if the level of saving transfer required is to be forthcoming. Long-term interest rates are very forward looking, and we judge that they already embody the interest-rate increases that will be required. The implications of this 'jump' in long interest rates is only slowly being taken into account, and we believe that it is a significant factor behind the slowdown in the advanced economies. Although in the long run they will benefit from the expansion of activity and markets, in the short run the higher cost of capital causes some temporary disjunction throughout the world economy.

Our exchange-rate forecast is, as usual based on the assumption that the nominal exchange rate will follow the open arbitrage path. We are expecting interest rates (and inflation) to be lower in Europe in the 1990s than they are likely to be in North America. As a result the dollar is expected to depreciate in nominal terms over the next decade. In Barrell and in't Veld (1991) we calculated that the dollar was currently near its real equilibrium level, and we do not foresee any major real realignments between Europe and North America over the next 10 years.

Oil Markets and Commodity Prices

The price of crude oil has risen from its summer low of $17 per barrel to around $21 to $22 per barrel, and forward markets suggest that this level of prices will be maintained throughout 1992. Given the slowdown in world activity it is a little difficult to interpret this rise in prices. There appear to be two major factors behind the strength of prices, both of which are increasing stockbuilding. First, the OPEC producers, apart from Kuwait and iraq, are producing to full capacity, and second, there are considerable worries about the continued flow of oil and gas exports from the former USSR.

Preliminary estimates suggest that OPEC oil production reached 24.7 million barrels a day in September. This is well above the level of 24.3 million barrels per day achieved in the first half of 1990, despite the loss of Kuwaiti and Iraqi output. The Saudis have increased their output from 5.6 million barrels per day in the first half of 1990 to around 8.5 million barrels per day in the third quarter of 1991. This brings their production up to their estimated sustainable capacity output. High levels of capacity use are also seen in Iran, the Emirates and Venezuela, and no OPEC country appears to be producing below capacity. As a result market participants may feel that stocks in the OECD need to be built up to replace the usual margin of spare capacity that would have acted as a substitute for holding stocks.

Production in the USSR has fallen from its 1987/88 peak of 12.5 million barrels per day to an estimated 10.5 million barrels per day in the third quarter of 1991. This output fall, and increasing transport and distribution difficulties, have led to fears that exports will be significantly curtailed. Soviet exports of oil and products fell in 1990 to 85 per cent of their 1988 level, but evidence for the first quarter of 1991 suggests that exports have stabilised. After the disintegration of the Union it is not clear that exports can be maintained at 1.6 million barrels per day throughout 1992, and commercial stocks in the OECD are being built up in anticipation of the short-fall.

Our forecast is in line with forward markets and we expect prices to stay high into 1993. By then the world economy should be recovering, and demand will be rising, and this will help absorb Kuwaiti and Iraqi output. We therefore expect real oil prices to stay firm, and to rise from around 1993 by 2 to 3 per cent a year. Table 4 contains our oil price forecast.

Other commodity prices have been weak during 1991. Our indices are plotted in Chart 7 and longer-term real prices are in Chart 8 and our forecast is set out in table 4. Metal prices have been generally weak, partly because the USSR is expected to increase its exports of high value metals such as nickel and zinc in order to obtain hard currency. Although tungsten dumping by the Chinese has been somewhat curtailed in the last year prices are still weak. Forward markets indicate that prices are generally expected to continue to fall over the next year. The major exception is in the copper market, where stocks are low and strikes have curtailed output. Agricultural raw materials prices have also been falling, in part because Australia, New Zealand and South Africa have all recently abolished price support schemes for wool.

Food prices have generally been weak, but in recent months wheat prices have picked up sharply, and they are expected to rise sharply up to the next harvest. This is in part due to the expected scale of deliveries to the USSR, but it also reflects the expected reduction of 40 million tonnes in world output in 1991/92. The majority of the shortfall comes from the US where output reduction schemes have been particularly successful. Sugar and coffee prices have been weak in 1991, with non-cane production of sugar rising. in general none of the food commodity markets are expecting an upturn in demand, and forward markets suggest that prices will continue to decline.

Longer-term prospects and convergence in Europe

We are presuming that a European Monetary Union will be formed sometime around 1997 to include all Member States except Greece, Portugal and Spain. The drafts to the Maastricht Summit agreement define, very precisely, what convergence criteria must be met for this to happen. In advance of the Summit meeting however, we do not know exactly how the criteria will be set. We can however, reasonably claim that monetary union will only take place if the following broad guidelines are met. There are four essential parts to the analysis of nominal convergence in Europe. First interest rates must converge, second inflation rates (and price levels) must achieve some sort of long-term relationship. This long-term relationship is only likely to be possible if, thirdly, fiscal deficits are kept within bounds, and fourthly if the costs of convergence in terms of unemployment are not excessive. The degree of convergence achieved so far is discussed in Anderton, Barrell and in't Veld in this Review.

Our forecast of the medium term assumes that governments in the Community are serious about their long-term commitment to union. We are assuming that interest rates in Europe converge by 1997. This alone will produce some degree of inflation convergence as the resulting system of fixed exchange rates will put pressure on wages and prices in the traded goods sector of the inflation prone economies. Chart 9 plots our projections for inflation in the major four economies. We are anticipating that UK and French inflation will be consistently below that in Germany over the next decade.

in Barrell and in't Veld (1991) we argued that the D-Mark was undervalued, and that this problem could only be ameliorated either by a revaluation or by faster inflation in Germany than in its partners. We believe that the second path is the more likely. The effects of German unification have loosened the German fiscal stance and have raised the rate of inflation in the Federal Republic. We expect that inflation will remain moderate in Germany, but that lower rates will be achieved in France and the UK. This is in part the effect of the initial small misalignment of the ERM system. if France and the UK are overvalued then exports will be lower than they otherwise would have been, imports will be higher, and the resulting balance of payments deficit will be causing real wealth to drop increasingly far below its equilibrium trajectory. This will put downward pressure on demand in the overvalued countries, even if the deficit is easily financible. Chart 10 plots our projections for the balance of payments of the four largest ERM members over the next decade. The pattern of deficits does not appear to be a potential source of stress in the Community.

The convergence of inflation will inevitably mean a gradual change in fiscal stance in some Community countries. The current version of NIGEM contains wealth effects in consumption, a full set of capital accounts and a set of public sector models for the major economies. This allows us to forecast public sector deficits in Europe over the next decade. Chart 1 1 plots the deficits in the major 4 economies. The public sector deficit in France is not currently excessive, and we expect consolidation to continue. Both the UK and West Germany have suffered recent deteriorations in their public sector balances and both might fail the Maastricht test next year. The deterioration in the UK is largely the result of the sudden downturn in activity, whilst that in Germany is the result of the costs of unification. Both countries can be expected to reduce their deficits in the medium term.

The italian situation is somewhat different. We are assuming that the process of fiscal consolidation that has been underway since 1985 will continue. As a result the italian deficit including interest payments (as a per cent of GDP) will fall from the current level of 9.5 per cent, and from its peak of 14 per cent in 1985, to the more manageable level of 2 per cent by 1999. This is a brave assumption, because we believe that this will only be possible if direct taxes are progressively raised. However the authorities are committed to reducing the deficit. In this forecast we are assuming that both public absorption of resources and public transfers will fall by 1 1/2 per cent of GDP each whilst direct tax revenue rises from 25 1/2 per cent of personal income in 1990 to 27 in 1999.

Chart 12 plots our forecast of unemployment in Europe, and in some countries it stays stubbornly high throughout the decade. These economies may eventually return to full employment but the transition costs are not negligible. The italian authorities believe that low inflation is a prize worth gaining, and high unemployment is a price worth paying. Our model, NIGEM, now has a complete model of the Italian personal and public sectors, and this allows us to quantify the effects. We have undertaken a simulation of our model where we do not tighten fiscal policy. This would make union impossible, and inflation and interest rates would have to be higher. A looser fiscal stance in italy would over the medium term raise growth by 1/2 of a per cent a year, inflation would be 1 per cent higher and unemployment would be slightly lower. In the long run, however, we believe that the real equilibrium of the economy may be little affected.

The United States

The US recession appears to have ended in the third quarter of 1991, but the signals on the strength of the recovery are very mixed. Real GNP grew at an annual rate of almost 2 1/2 per cent in the third quarter, after three successive quarters of decline. Consumer spending has been rising for two quarters, and it is estimated to be 3 per cent above the level of the first quarter. Durables expenditure was particularly strong in the third quarter, with motor vehicle sales in the quarter rising for the first time in a year. Investment also rose strongly, the first increase for two years, and residential investment was 5 per cent above the level of the previous quarter but still 7 1/2 per cent below the level of 1990Q3. Business investment was 1.6 per cent above the level of the first quarter but remained 2 1/2 per cent below its level in 1990Q3.

imports of goods and non factor services continued to rise strongly, whilst exports were weak. Domestic demand grew by over 1 per cent in the third quarter, but the combination of a declining balance in goods and services and the effect of the appreciation of the dollar on the value of net property income from abroad produced GNP growth of only 0*6 per cent in the quarter. The dollar has appreciated some 10 per cent in the last year and the effects on trade have begun to become apparent.

The durability of the recovery will., it appears depend to a large extent on the strength of consumers expenditure. This increases the significance of the sharp drop in consumer confidence in October. The Conference Board of New York index has been falling for four months, and in November it had almost dropped to its Gulf War level. Consumer credit outstanding has also been falling since at least May. it appears that the prospects for increased consumption expenditure are not good, and other indicators point to a slow recovery. The index of leading indicators fell in September, and had been flat for the two previous months. Private housing starts also fell in the same month and are 6 1/2 per cent below the level of a year ago. The unemployment rate rose in October, reflecting in part the lack of any increase in industrial production and manufacturing employment dropped 32,000 in the same month. Capacity utilisation eased down a little in September, although it is still above the cyclical trough achieved in March 1991. The widely quoted National Association of Purchase Managers Index also fell in October and it is just above its 9 year low point.

All the signs are that the recovery is faltering and the Federal Reserve has reacted promptly. The discount rate was cut by 1/2 a per cent to 4 1/2 per cent on the 6th of November. Commercial Banks cut their prime rates at the same time. This response also reflects worries about the slow growth of the broader monetary aggregates in recent months. We believe that the response of the Federal Reserve should be sufficient to stave off a second recession, but we have incorporated some of the recent pessimism into our forecast judgments. Chart 13 plots the residuals on our consumption, business investment and housing investment equations. In each case we set the value of the residual over the medium term approximately equal to its historical value, but in the short term we believe outturns will be worse than an equation based prediction would suggest.

Output in the US grew by 21/2 per cent in 1989, well below its peak in 1988, and the slowdown continued into 1991. We are expecting this slowdown to be reversed. Our forecast for US GNP and its components is set out on table 5. We expect output to grow by 21/4 per cent in 1992, largely driven by domestic demand. We expect growth in both housing and business investment. Short interest rates are expected to stay around 5 per cent for much of 1992 and should rise only slowly there after and long rates can be expected to fall if our short rate predictions are correct. A lower long rate should raise the value of net financial wealth and hence stimulate consumption indirectly as well as having a direct effect on business investment. Lower short rates should help to raise demand in housing markets.

Our short-term interest-rate forecast reflects our view on inflation. The consumers expenditure deflator rose by less than 1 per cent in the third quarter. This is well below its peak rate increase in 1990, and below the 5 per cent increase in 1990 as a whole. We are expecting the inflation outturn for 1991 to be below 4 per cent. This reflects weak wholesale prices, a strong dollar, and the rather limited growth in the pressure of demand. We are forecasting that inflation will fall further in 1992, and interest rates will be held low. Given the slow speed of response of output to interest rates we do not expect strong growth next year, although the average from 1992 and 1993 is very close to our estimate of long-term output potential growth.

Our forecast for the US current balance and for trade is set out in table 6. The US current balance deficit is likely to be only around $15 billion in 1991. This reflects in part the large scale of transfers to the US from Germany, Japan and the Saudis. These transfers are a payment for the Gulf War and probably exceed $35 billion dollars. Although, as Keynes stressed in the 1920s, transfers on this scale are always hard to make effective, the openness of capital markets and the lack of exchange controls should mean that there is little effect on the exchange rate. Net property income from abroad has also been strong in 1991, and we have projected part of this increase in our forecast of the rate of return on assets. This should help sustain the US current account. We are forecasting that the IPD balance will remain positive until 1994, and then cumulating current account deficits will push it into deficit itself. The improvement in the current account is expected to be sustained, and we do not expect the overall deficit to return to its mid 80s levels. The slow recovery from recession will keep the deficit below 1 1/2 per cent of GNP until at least 1994, and we do not expect it to rise above 1*8 per cent in the foreseeable future. This level of deficit is, we would judge, financible.

Our public sector forecast is given in table 7. Our forecast for the deficit uses our model of the US public sector, and our equations are based on data from the National Income and Product Accounts. The forecast deficit therefore excludes any temporary or permanent additional borrowing directly necessitated by the collapse of the Savings and Loans Institution.

We expect that the Federal and total Public Sector deficit on an NIPA basis will expand in 1992 to $188 billion and $175 billion respectively. The public sector deficit will reach 2.9 per cent of GNP in 1992, its highest level since 1986. However during a recession some deterioration in the balance, and especially the balances of states and local authorities, can be expected. We expect the slow reaping of the peace dividend to reduce the level of spending over the medium term, and by 1997 we are expecting the public sector deficit to drop below 1 per cent of GNP.

Japan

Forecasts of output growth over the last 4 years in Japan have generally been lower than the eventual outcome. (3) We have had to slightly revise up our forecast for 1991 in the light of the albeit weak second quarter data, but we have revised down our forecast for 1992, in part because we are expecting a less favourable external environment but mainly because domestic prospects appear to have deteriorated. The Japanese authorities have been worried about rising inflation, and interest rates have been progressively raised. High short and long rates appear to be affecting both consumption and investment.

Real GNP rose 1/2 a per cent in the second quarter, and in the first half of 1991 it was 5 1/2 per cent higher than in the same period a year previously. Housing investment has been the weakest component of domestic demand and in the second quarter it was over 10 per cent below the peak it reached in the third quarter of 1990. Business investment was also lower in the second quarter than in the first, but it has been very strong recently. We are expecting GNP growth to slow in 1991 to just over 4 per cent. This will largely be driven by domestic demand which is forecast to grow at only 3 1/2 per cent in 1991.

Current indicators suggest that the prospects for 1992 are not good, and although we expect the Japanese economy to avoid a recession (by Japanese definitions this means growth below 2 per cent) we are expecting the lowest growth in GNP since the early 1980s. This forecast of 2 1/2 per cent growth follows largely from an automatic setting of residuals at their historical averages, and it reflects the large but slow acting effect of high interest rates on the Japanese economy. However, the forecast is supported by short term indicators. The index of leading indicators fell in August to its lowest level for 15 years, and has been below 50 for twelve months. Capital expenditure is expected to slow sharply, partly in response to high interest rates and the slowdown in activity, but also because equity financing has become more difficult since the fall in the stockmarket in 1990 and because BIS reserve ratio rules on banks has limited their capacity to lend. The generally reliable Bank of Japan survey in August suggested capital spending would rise 8 per cent in the current fiscal year compared to 17 per cent in fiscal 1990/91. Intentions are particularly low in small and medium size firms.

Our forecast for output prices in the Japanese economy is given in table 8. We have seen a small reduction in interest rates this year, and we expect furtherfalls into 1992. We therefore expect some sort of recovery in housing investment but this will be mild, and it will not be sufficient to offset the projected decline in investment growth. Stockbuilding was high in the second quarter of 1991, but we expect that the unplanned component of stocks will be reduced in the near future, and as a result the growth of domestic demand is expected to be reduced. In the longer term we expect growth to increase to around 3 1/2 per cent in 1993 and to just above 4 per cent thereafter. We expect that the recovery will be generated by lower interest rates and by a faster rate of growth of public spending. As far as we can ascertain the Japanese public sector as a whole has been running a surplus for some time, and taxes on business are being increased to pay for the Japanese contribution to the Gulf War. However, in the longer term we expect the budget surplus to decline slowly.

The tight fiscal and monetary stance has reflected fears about rising inflation. The consumers expenditure deflator was flat in 1987 and 1988, and then rose increasingly rapidly to a forecast peak of 2.7 per cent in 1991. Consumer price inflation has been higher, and the index is likely to have risen by 3.3 per cent in 1991. These figures are high by Japanese standards and in combination with a tight labour market they were seen as seriously worrying. However, labour market pressures have begun to ease, and unemployment rose to 2.2 per cent in September from its low point of 2 per cent in May. Employment has fallen below its peak of 64.8 million in June, and the ratio of unfilled vacancies to job seekers has fallen from its 16 year peak of 1.47 to around 1.34. The prospects for inflation look better. Wholesale prices fell in September, as did import prices. However these falls have in part been the result of the strength of the Yen, and this in turn has added to the deflationary pressure on demand. We are forecasting that wholesale prices will rise by 2.3 per cent in 1991 and by less than 1 per cent in 1992. Consumer price inflation will slow to 2 1/2 per cent and the CED is expected to increase by less. The short, sharp slowdown will have successfully reduced the rate of inflation.

The slowdown in activity will contribute to a notable improvement in the current account in 1991. This in part also reflects a J curve effect from the real appreciation of the Yen that took place between the second and fourth quarters of 1990, and it is despite the scale of transfers to the US. The current account is expected to show a surplus of $65 billion in 1991 despite the fact that export volumes are expected to fall and import volumes are expected to rise. The improvement in the surplus is expected to be largely sustained as the real exchange rate is forecast to stabilise at around its current level. The continuing pressure to open up the economy will, we expect, produce faster growth of imports than exports, and especially of imports of services. This will cause the surplus to stabilise, but the cumulating stock of overseas assets and the resulting income flow will impede the fall in the current account as a per cent of GNP.

Work reported in Box A in this chapter and in the World Economy Chapter of the August 1990 Review suggests that the nature of Japanese trade has changed. Up until the mid 1980s the Japanese appear to have been gaining world trade share at constant competitiveness. Since then they have not been doing so. Import demand, both for goods and for services appears to have changed structure in the late 1980s producing a more rapid growth in import penetration. These long run trends are likely to reduce the Japanese current account surplus to around 1 per cent of GNP in 1990. This does not mean that Japanese firms are losing market share, but rather that they have been moving their plant to low cost locations around the Pacific Rim and also into areas such as the European Community in order to avoid tariff and non tariff barriers to trade.

Germany

The process of unification has given a strong boost to growth in Germany. The 1980s were a period of low growth and low inflation in the Federal Republic. Domestic profitability was low and the government ran only small deficits. As a result saving flowed abroad and the current account was in persistent surplus. Germany shared in the world upswing in 1988, and GNP growth rose to 3 1/2 per cent. However much of the upswing was externally generated, and the balance of payments improved to a surplus of 4.2 per cent of GNP. Growth accelerated to 3.8 per cent in 1989 despite a slow down in domestic demand. The current account surplus improved to 4.8 per cent of GNP.

The shock of unification began in 1990, and it produced a surge in domestically generated growth. The current account deteriorated to 3-2 per cent of GNP and domestic demand rose by 5.2 per cent, whilst GNP grew by 4.7 per cent. This marked change in the mainspring of German growth was accompanied by a sharp deterioration in the public sector accounts. The complete public sector displayed a small surplus in 1989, but this deteriorated to a deficit of 3.5 per cent of GNP in 1990. Most of this deficit occurred in the second half of 1990, and it was driven by large transfers to the East. Domestic investment also grew very strongly in 1989 and 1990. As a result of the increase in demand generated by unification the German economy was operating near its capacity ceiling, and investment grew in part to relieve the pressure of capacity shortage.

Inevitably the domestically generated growth began to put pressure on prices in Germany and inflation started to accelerate, and the Bundesbank began to become concerned about inflation. interest rates rose from 4.25 per cent in 1988 to 7.1 per cent in 1989, and they continued to rise throughout 1990. By the end of the year interest rates had risen to 8.8 per cent, and ex post real interest rates were 6.2 per cent. These high interest rates were a reflection of worries about emerging inflation, rather than a response to high inflation. Rates rose further through 1991 to peak at around 9 1/2 per cent in the fourth quarter. The rate of inflation has risen in 1991, but this in part reflects the increase in indirect taxation that took place in July. Wage increases were seen as more worrying. Economy wide compensation per head rose by 3.2 per cent in 1989, by 4.7 per cent in 1990 and are expected to rise by 6.6 per cent in 1991. Manufacturing earnings growth has also accelerated from 3.8 per cent in 1989 to an anticipated 5 per cent in 1991.

The Bundesbank's increasingly tight monetary policy had been having very little effect until the last six months. Fiscal expansion was in part offsetting the effects of high interest rates and as we have noted above the German economy anyway responds rather weakly to high rates. However, the Ketchup appears to have just come out of the bottle (4) and after a very strong set of first quarter GNP figures we are expecting GNP to fall in the third quarter and only return to its first quarter level in the last quarter of the year.

There have been a number of signs of this slowdown. Industrial production fell in every month between June and September. Export volume growth has been very low. Unfilled vacancies have fallen from a peak of 344,000 in May to 320,000 in October. Capacity utilisation in manufacturing fell to 87.9 per cent in the second quarter. Although this is still relatively high, it is below the peak level of 90 per cent recorded in the fourth quarter of 1990. Business surveys by IFO suggests that order books are expected to shorten, and business confidence surveys undertaken by regional Chambers of Commerce indicate a marked weakening.

it is very difficult to interpret the second quarter data because a marked seasonal pattern appears to have crept into the Bundesbank's seasonally adjusted data. German research institutes, such as DIW and IW Kiel, seasonally adjust the raw data themselves. The Kiel interpretation of the second quarter data suggests that consumption growth slowed to around 2 1/2 per cent at an annual rate, and he growth of investment in equipment slowed to 2 per cent at an annual rate. Only construction continued to grow strongly in the second quarter. Their adjusted data has GNP rising by 3 1/2 per cent at an annual rate in the second quarter, whereas the Bundesbank adjusted data shows a fall.

The interpretation of recent data is difficult but it is clear that the economy of western Germany is slowing down rapidly. Table 10 contains our forecast for national income. We are expecting growth to slow to 3.3 per cent in 1991 and to 2.3 per cent in 1992. Much of this slowdown is generated by a sharp slowdown in business investment growth. This is in part a response to high interest rates, but it also reflects the high short term elasticity of investment with respect to output. Our equation residual has been set below its historical average to reflect poor confidence, and our slowdown in Germany is not entirely generated by our model.

The slowdown in activity will, we believe, inhibit the Bundesbank from raising interest rates further, and as is shown in table 3, we expect them to be putting downward pressure on German interest rates from the middle of next year. in the medium term we expect growth to pick up again in Germany as output and demand grow in the east, and we are forecasting GNP growth in excess of 3 per cent in 1993 and for several years thereafter. This strong growth is forecast to be accompanied by stable inflation of around 3 1/2 per cent a year. This may seem high given the Bundesbank predilections, but it is only the German average for the last 20 years, and it should ease the transition to monetary union in Europe.

The German balance of payments deteriorated sharply in 1991, and we are forecasting a deficit of $21 billion, or 1 1/2 per cent of GNP. This reflects in part the Gulf War transfers to the US, but it was mainly caused by the extremely rapid growth of imports. Our forecast of the German balance of payments is given in table 11. Our trade volume data is still based on west German national accounts, but the overall balance is based on all Germany data. The inconsistencies now mainly show up through the invisibles, and we have made some allowance for this in our forecast. The recovery in the east (discussed in Box B) should limit the growth of imports in the medium term, and we are expecting the slowdown in activity to produce a return to current account surplus by 1994. Increasing export profitability is likely to help maintain export growth, and the process of monetary union through exchange rate discipline should help restore some of the German current surpluses observed in the 1980s. Our long term forecast for the German current account is given in chart 10 above.

Monetary union in Europe will require a tightening of fiscal policy, but we believe that developments in the east, along with the announced higher taxes in 1993 would enable the Germans to meet the draft Maastricht rules with ease. Table 12 contains our model based German public sector forecast. Transfers to the east are mainly encompassed by miscellaneous expenditures. These rose from 140 billion D-Mark in 1989 to 368 billion D-Mark in 1991. They are forecast to stabilise at this level, and hence the normal effects of nominal and real GNP growth on revenue, plus the announced rise in indirect taxes in 1993, should be sufficient to reduce the public sector deficit to 1.3 per cent of GNP in 1999.

France

Output growth in France has reflected the pattern in the rest of Europe. Growth accelerated from 2*3 per cent in 1987 to 4.2 per cent in 1988, with much of the stimulus coming from external demand. Growth was also strong in 1989, and it slowed a little into 1990 as net exports grew less rapidly. The commitment to the ERM has meant that short interest rates rose in line with those in Germany from 7.9 per cent to 10.6 per cent in the first quarter of 1990. The rise was an important factor behind the slowdown in activity in 1990. High interest rates caused the ERM block to appreciate against the dollar, and the subsequent loss of competitiveness exacerbated the slowdown.

The reduction in growth in 1991 is not expected to be as severe as that seen in either the UK or in italy. French inflation has been around 2 1/2 to 3 1/2 per cent for some years, and unit labour cost increases have been moderate. This has meant that the effects of the real appreciation of the ERM block have been ameliorated somewhat. French inflation has indeed fallen below that in Germany recently, as can be seen from Chart 9, and this has been one of the factors that have allowed the French authorities to cut their interest rates from 10.6 in the first quarter of 1990 to 9.6 in the second quarter of 1991. We expect that they will rise slightly in the fourth quarter and then converge on German rates in early 1992. The reduction of interest rate pressure has aided the economy, and it has reduced the severity of the slowdown.

The increase in credibility and the resultant fall in interest rates has not prevented a slowdown in activity, rather it has just ameliorated it. GDP in the first quarter was at the same level as in the third quarter of 1990. Real GDP rose by almost a per cent in the second quarter. Consumer spending remained weak, but net exports and investment recovered somewhat. The recovery in investment was largely in construction and this sector has been aided by lower interest rates. Industrial production rose by 0*8 per cent in July/August, but was still 0*9 per cent lower than a year ago. Manufacturing production also rose, but was 2.7 per cent lower than a year previously. There are signs that a recovery is under way, but it is unlikely to be strong. Unemployment generally lags behind output during the cycle. It rose to 9.5 per cent in August, up from 8.9 per cent at the beginning of the year. We are expecting unemployment to stabilise at around 9.6 per cent in 1992 and 1993.

The May INSEE investment intentions survey saw firms revising their plans down by 7 per cent in value (and 6 per cent in volume terms). Capacity utilisation dropped in the third quarter to its lowest level since the end of 1987, and it is 5 per cent below its peak in the second quarter of 1990. These factors reinforce our forecast of virtually flat investment in 1991, with a small rise in construction offset by a small fall in business investment. We do, however, expect output to recover in the third and fourth quarters. This view is supported by the INSEE October survey of activity which suggested a rise in activity in the third quarter of the year.

Inflation has been below 4 per cent for around five years. It rose a little toward the end of 1990 and in the second quarter of 1991 consumer prices were 3.3 per cent higher than a year previously. High interest rates and a decline in output growth have produced a degree of spare capacity, and inflation is now falling. We expect consumer prices to rise by 3 per cent in both 1991 and 1992. Manufacturing earnings grew by 4 1/2 per cent in 1990, but rising unemployment will help put pressure on wages, and we are expecting increases of no more than 4 per cent this year.

Our forecast for the French economy is set out in table 13. We expect output growth in 1991 to be 1.3 per cent, after 2.8 per cent in 1990. Industrial production for 1991 as a whole will be the same as in 1990. Lower interest rates should eventually aid a recovery in business investment, and we are expecting domestic demand to grow by 2.2 per cent next year. The recent appreciation of the dollar along with the weakness of the Franc within the ERM have produced a small real depreciation in 1991, and this will aid net exports, and GDP growth in 1992 is expected to be 2.4 per cent, slightly above the growth of domestic demand. We are forecasting that the French economy will return only slowly to its equilibrium growth rate of almost 3 per cent. Falling interest rates will sustain the recovery, but the ramifications of slow growth elsewhere will be an impediment.

Moderate inflation and a slow recovery should produce an improvement in the balance of payments in 1992. Our forecast for the current account is set out in table 14. Export market growth is likely to be slow in 1991 and 1992, and exports of goods are rising less rapidly than demand. import growth has also been low, and as a result the visible balance has improved in 1991, and is expected to improve again in 1992. The removal of capital controls in mid 1990 has led to large gross capital flows to countries with less regulated banking sectors, and a large amount of French financial business is now being transacted through Luxembourg. This has caused a slight deterioration in the net balance on interest profits and dividends because the spread between borrowing and lending rates enters the accounts as an IPD debit. in the longer term we expect the French current deficit to stay below 1/2 a per cent of GNP. This will in part be driven by sluggish growth. The commitment to the ERM and to Monetary Union requires a process of gradual government budget consolidation. This not only puts downward pressure on demand but it also changes the composition of financial wealth. A reduction in the stock of government debt as a per cent of GNP will induce wealth holders to increase their stock of assets held overseas and this requires an improvement in the current account. Wealth effects in France on NIGEM are relatively strong, as is shown in Barrell, Gurney and in't Veld (1991), and these portfolio effects are at work in our forecast.

Italy

The Italian economy grew by only 2 per cent in 1990, well below the average of 3 1/2 per cent over the previous three years. This slowdown in activity was common to all of the European economies outside Germany. Charts 9 to 12 above give some details of developments in Italy over the 1980s, and our forecast for the 1990s. The Italian authorities have had a strong commitment to the ERM, and although they have continually realigned they have allowed the real exchange rate to rise over the decade. This had led to an increasing balance of payments deficit. In 1987 the balance of payments was in surplus, but in each year up to 1989 it declined by half a per cent of GDP to reach a 1.3 per cent deficit by 1989. The real effective exchange rate rose by 10 per cent over the same period despite two realignments within the ERM.

The commitment to the ERM has been a major tool in the hands of the authorities in combating inflation. Until 1983 most Italian wages were reindexed once a quarter, and it was only in 1985 that full automatic indexation was removed. Since then italian wage growth, at least in the traded goods sector seems to have moderated, but competitiveness has still continued to decline. By the first quarter of 1991 the IMF measure of trend adjusted relative unit labour costs was 20 per cent above its 1986 average. Wage inflation may have moderated, but not yet sufficiently to pull Italy into line with its competitors. in the August 1991 Review in Barrell and in't Veld we calculated that the italian economy was 10 to 15 per cent above its equilibrium real exchange rate, and that the situation is deteriorating.

Manufacturing earnings rose by an average of 6 1/4 per cent a year between 1986 and 1989, but they accelerated to 7-3 per cent in 1990, and we are forecasting that they will grow by 8.4 percent in 1991. Over the same period average earnings in the economy as a whole were rising more rapidly. Between 1986 and 1989 they rose by almost 9 per cent a year, and they rose by 10 per cent in 1 990. This disparity partly reflects wage rises in the public sector and in nationalised industries. The difference between wages in the traded and non traded goods sectors in Italy is becoming quite marked, whilst labour productivity has been growing slowly in services sector. The OECD Economic Survey of Italy for 1991/92 (page 80-86) makes great play of this increasing disjunction.

The increasing competitiveness pressures on the Italian economy are significant when analysing our forecast, which is set out in table 15. The economy has slowed down in 1991, and we are forecasting that growth will fall to .8 per cent. This slowdown has happened despite the fall in interest rates since the abolition of capital controls. Rates are now 1 1/2 percentage points lower than they were in the first quarter of 1990, just before the abolition of controls. However, we believe that the authorities will have to raise interest rates in the near future, and this affects our forecast for 1992. The economy appears to have reached a relative trough at the end of the first quarter. industrial production in the second quarter was 6 per cent below the peak reached in the last quarter of 1989, but it has been growing slowly in recent months. However business confidence indicators, which strengthened in the Spring, have turned down again, and by August prospects for the whole economy were just as poor as they were toward the end of 1990.

We are forecasting private sector investment growth of only 1% per cent in 1991, and although there are signs of stronger consumption growth supported by rising earnings we expect consumption to rise by only 2 per cent in 1992. As a result GDP is expected to grow by only 1 1/4 per cent in 1992. In the longer term we expect a slow recovery in the Italian economy as competitiveness pressure and slow world trade growth will initially hold down exports. In the medium term the process of budget consolidation is expected to be resumed. The public sector deficit has fallen from 15 per cent of GDP in 1985 to under 10 per cent in 1991, and we expect income taxes to be progressively raised and spending growth curtailed in order to bring the deficit down further.

However, in the short term, the government faces serious credibility problems. Inflation has come down, but only to 6.2 per cent in September, well above the ERM average. The 1992 budget contained an ad hoc package of measures designed to cut the overall deficit by Lit 13 billion to Lit 128 billion. However, much of this improvement comes from tax receipt increases produced by special measures such as tax amnesties for tax avoiders. However, after the 1992 elections we expect the process of true consolidation to be resumed, with lower spending growth and higher income taxes. This will help improve the prospects for the balance of payments. As chart 9 shows we expect slower growth to eventually produce inflation convergence on German rates, but we do not believe that this will be achieved without one more realignment before 1995, the date after which the Maastricht treaty suggests no further realignments should take place.

Canada

Growth in the middle of the 1980s in Canada was very strong, stimulated by low interest rates and a loose fiscal stance. Inflationary worries began to emerge in 1988, and interest rates were raised and eventually fiscal policy was tightened. Inflation only rose to 5.0 per cent in 1989, but the authorities were very concerned about it accelerating further. Monetary policy was held tight until the second quarter of 1990, when interest rates peaked at 13.6 per cent. They have been declining since as signs of the recession have become clear. Output fell throughout 1990, and was only 1/2 a per cent above that in 1989, and 1991 will beyond doubt see a recession on a year-on-year basis.

The recession does, however, appear to be over. The economy grew at an annual rate of almost 5 per cent in the second quarter of 1991 after a full year of quarterly declines. Domestic demand rose even more strongly. The decline in interest rates over the last year has boosted both residential investment and consumption. Total consumer spending was 2 per cent higher than in the previous quarter, and durable goods spending rose by 7 per cent (28 per cent at an annual rate). The Conference Board of Canada's Index of Consumers Attitudes rose strongly in the second quarter. Real incomes have risen and savings have fallen. Residential investment has risen strongly, and after five quarters of decline it is now 5 per cent above the level of the previous quarter.

The recovery has been largely driven by the personal sector. Business investment hardly rose in the second quarter, in part because capacity utilisation rates are low, but it appears largely because the corporate financial position is very weak. Profits are 50 per cent below their peak in the first quarter of 1989. Business inventories fell markedly in the second quarter, and the growth rate of GDP would almost have doubled if they had not done so. We expect some of this destocking to unwind, and this is a factor behind our upward revision to our forecast. Table 16 sets out our forecast for the Canadian economy.

The short term prospects look considerably brighter than they did 3 months ago. Core inflation, after netting out the effect of the Goods and Services Tax increase in January, has probably dropped to 3 1/2 per cent. Wage settlements have begun to moderate. Private sector settlements were averaging 3.6 per cent in July, down from 5 1/2 per cent in the first quarter and public settlements, at 2 per cent, are even lower. Unemployment appears to have stabilised over the summer at around 10.6 per cent. As in other countries unemployment lags behind demand.

We are expecting both consumption and residential investment to keep rising, but business investment is expected to be flat for the next year. Both will fall on a year-on-year basis in 1991 before rising by 2.3 and 5.7 per cent a year in 1992. Domestic demand is expected to grow by 2 1/2 per cent next year after falling this year. GDP growth will exceed the rate of domestic demand growth for some time. Export volumes shrank sharply over the last year, and Canada lost market share after competitiveness declined by more than 4 per cent in 1990. However exports of goods have began to pick up in the second quarter, and we are expecting net exports of goods and services to add half a per cent to GDP growth in 1992.

The 1980s saw a long succession of balance of payments deficits of 2 to 3 per cent of GDP. This reflected both the strength of demand and the over-valuation of the Canadian dollar. We expect demand growth, to be slower in the 1990s than in the 1980s, and fiscal policy will be tighter. Hence the current balance is likely to gradually improve over the next decade. REFERENCES

Ray Barrell, Andrew Gurney, and Jan Willem in't Veld, (1991), 'The real exchange rate, fiscal policy and the role of wealth: an analysis of equilibrium in a monetary union', National Institute Discussion Paper no. 3.

Barrell, R. and in't Veld, JW, (1991), 'FEERS and the path to EMU', National Institute Economic Review, no. 136, August.

Breush, F., Busch, G. and Walterskirchen, E., (1991), Short-term prospects for the European Economies', paper given in Brussels in October 1991, published in Conjunctura Italiana, November 1991.