Chapter II. The world economy.

Barrell, R.J. ; Gurney, Andrew ; Dulake, Stephen 等

The longer-term prospects and the implications of developments in Eastern Europe In the last six months political and economic developments in Europe have moved very rapidly. The democratisation of the East has been quite remarkable, and much heralded, but the change in the political agenda in the West has been almost as important, although much quieter. The drive toward European Monetary Union seems to have speeded up, although there are many important issues still to be settled.

There has been a marked convergence of economic performance in Europe over the last decade, and inflation rates in France, Germany and Italy are now very similar. However, it would be wrong to deduce from this that it would be easy to create a monetary union in Europe. As Britton and Mayes (1990) argue, economic performance may have converged either because policies have become more coordinated or because economic structures in different countries have become more similar. The evidence points to better policy coordination as the source of improvement; the process of structural convergence has only just begun.

Early in the process of the digestion of news from the East it looked as if the unification of Germany would pull the Federal Republic away from the rest of Europe. However, the political agenda seems to have changed. The Germans and the French have recently committed themselves to closer political cooperation, and the drive for European Monetary Union has gathered pace in the last four months, especially with the publication of the EC's paper on `Economic and Monetary Union', (1990). The major question concerning European monetary union now appears to be the date at which the union will be formed. In our forecast we have assumed that exchange rates in Europe will be irrevocably fixed from 1st January 1997. The costs of monetary union are discussed in a note on the EMS and EMU in this Review, which shows how the loss of an independent monetary and exchange rate policy might reduce the ability of the European economies to respond to shocks. These problems have to be set against the gains from enforced monetary policy cooperation and also against the benefits from the removal of the transactions costs involved in the use of a number of disparate and independent currencies. There is some discussion of these costs in the Annex to the EC's 1990 paper, and they do appear large. Recent developments in East Germany and elsewhere have done a great deal to clarity the longer-term prospect for Eastern Europe economies. The unification of Germany within a few years is now certain, and the transition of at least Hungary, Poland and Czechoslovakia into market economies looks highly likely. The outcome of political and economic change in Russia is less certain, but the liberalisation of production and labour markets now seems irreversible. The future of the Baltic states remains in the balance, but they are likely to become both more independent and more economically progressive. The effects of all these changes could be vast, but their impact has so far been small, except perhaps in the Federal Republic of Germany.

Developments in Eastern Europe have also led us to change our forecast and our model in various ways. The Centrally Planned Economies (CPEs) have been running current account surpluses for most of the last decade. We have assumed that they will not do so over the next decade. The increased investment opportunities in the East are likely to induce major capital inflows, and hence large, and increasing current account deficits can be financed. We have assumed that there will be little effect on trade in 1990, but that import volumes will rise strongly in 1991 and 1992. It is, however, important to realise that prospective developments in Eastern Europe, outside of Germany, are likely to have little impact on the West. A doubling of Eastern imports would add only 5 per cent to the level of world trade, even after the multiplier effects are taken into account by our model. Our improved model allows us to take full account of the effects of German labour migration and unification on real wages and employment in Germany, and the combination of lower wages and higher investment that we are forecasting lead to up to 1/2 a per cent a year higher growth in the 1990s. The supply side expansion should mean that this shock can be absorbed with little effect on inflation. The effects of changes in the East should not however, be confined to Germany alone and they are discussed further below. The rest of Europe is expected to benefit, both from higher demand in Germany and also marginally because of increased CPEs imports. However, there will also be benefits from reduced military spending throughout the West. The effects are most important in the US. As we stressed in our last forecast chapter, the reduction in expenditure and the concomitant reduction in the government budget deficit change the longer-term prospects for tax changes. The reduction in military spending that is already underway has changed the political agenda in the US.

The overall outlook

Output growth in all the major seven economies has been strong for some years, and since 1982 the growth rate of the group as a whole has exceeded 3 per cent a year. By the late 1980s output was probably at or above capacity in most of the major economies, and by early 1989 capacity utilisation indices based on business survey evidence were at their highest levels for a decade. Output growth began to rise in the late 1980s for a number of reasons. There was perhaps some delayed response to the fall in oil prices in 1985. More importantly the attempts to defend the exchange rate parities associated with the Louvre accord meant that the Dollar had to be supported. Although this defence failed, and the Dollar fell, in the process the Americans were able to export some of their inflationary pressure to the rest of the world. There may have also been some impact on investment and demand in Europe from the realisation that the 1992 programme might stimulate trade. This build up in demand had consequences for inflation. Inflation rose in all the major economies in 1989, with consumer price inflation averaging 3.75 per cent in the major four and over 4 per cent in the major seven. Chart 1 plots recent and prospective inflation in the major four.

The surge in inflation brought a sharp monetary policy response from the authorities, and especially in North America where output and capacity utilisation have responded. Capacity utilisation rates have, however, not fallen in either Japan or in continental Europe. Chart 2 plots capacity utilisation indices for the major four economies. There has been only a small output response to the tightening of monetary conditions that took place during 1989. Prospects for inflation now look less promising then they did three months ago. Charts 3 and 4 plot both three month interest rates and long-term government bond rates. Short-term rates have not risen noticeably since our last forecast, except in Japan, whereas long-term rates have generally taken an upward step of half a point or more. Yield curves are now upward sloping in all the major economies except France, Canada and the UK. Long rates reflect expected future short rates, and hence a rise in long rates tells us that the expectations of short rates in the future have been revised upward. This may reflect either higher expectations of inflation in the future or higher real interest rates, which would in turn reflect a higher level of capital productivity. We think that on balance the rise in long rates reflects an increase in expected inflation over the future. Our forecast, as set out in table 1, does reflect this. We have in particular revised upwards our forecasts for inflation in Germany and Japan. We think that there will be no further monetary response in Japan or the US, in part because a further tightening of interest rates in the US would put further downward pressure on the Yen. There is a risk that short rates will rise in Germany, and given their commitment to the EMS this would produce a rise in rates in other European countries. However, we feel that the longer-term fiscal prospects for the major seven, and especially the heavy defence spenders such as the US and Germany, have been positively influenced by developments in the East. There should therefore be less pressure on the monetary authorities to raise interest rates as a tighter fiscal stance will soon be exerting some downward pressure on demand.

We are anticipating that the slowdown in US activity that began to emerge in 1989 will not be reversed in 1990, and inflation will begin to fall in 1991 after continuing to rise further this year. The slowdown in US and Canadian activity, along with less precipitate growth in Europe, is expected to reduce the growth of world trade in 1990 to 4.2 per cent for manufactured goods, and 5 per cent for all goods. The combination of stronger growth in Europe, along with a recovery of activity in the US in 1991/2 is expected to produce an increase in world trade growth in 1991. Thereafter there is likely to be a boost to trade growth rates as CPEs imports begin to rise, and also as the 1992 programme begins to produce greater cross border flows in Europe. The slowdown in world trade looks precipitous, because trade growth in 1988 and 1989 averaged 8.5 per cent a year. However, some of this growth came from improved recording of US/Canadian trade, and some from the large increase in Chinese imports from East Asia. Neither should have had much effect on major seven activity levels. The greater than anticipated inflationary pressure being experienced, especially in Europe, has not been reflected in commodity prices. Crude oil prices rose to around $20 per barrel at the beginning of 1990, but the slowdown in activity along with a warm spring and a chronic tendency for all OPEC producers to exceed their quotas has caused a sharp fall in oil prices to below $18 per barrel in the last three months. Overproduction by OPEC has partly been absorbed into stockpiles, but continued overproduction could lead to further price falls. The prospect of major reductions in CPEs hydrocarbon exports seems to be the only factor likely to support the price in the near future. Other commodity prices have in general been weak. Chart 5 plots recent developments, and table 2 sets out our forecast. Metals prices have fallen from their recent peaks, and despite some supply difficulties for copper we expect that in the longer term they will continue to decline as improvements in extraction technology cause potential output to grow more rapidly than demand, which grows only slowly in the longer term as the metals intensity of output is expected to continue to decline.

Prospects for food prices are more uncertain. We are expecting developed country free market food prices to rise only slowly this year and even fall in 1991. This is because US production should continue to recover from the 1988 drought, and also because we are anticipating major falls in CPE food imports in the next few years. Increased efficiency should show up relatively quickly, except perhaps in Russian agriculture, and more reasonable food prices should cut meat consumption substantially. As a result we are expecting low levels of demand for free market grains, both wheat for human consumption, and also for feed grains. This will put downward pressure on free market grain prices during the early 1990s. In the short term prospects for LDC food prices look bleak, especially as the coffee and cocoa cartels appear to be suffering great strain. Prices for their products have dropped in the last year, and we expect them to be weak in the near future. Only sugar prices have been strong, partly because of supply difficulties in Brazil but also because of constraints by the EC on oversupply in Europe. Attempts by the EC to rationalise its support system should help LDC food producers, and after price declines this year we are anticipating rising prices in the near future as free market prices respond to lower EC surpluses.

Exchange rates and interest rates

The last year has seen sharp changes in exchange rates that might seem hard to justify. Despite the large current account surplus the Japanese effective exchange rate fell by almost 16 per cent between the first quarters of 1989 and

1990. The D-Mark has appreciated by 7.5 per cent in effective terms over the last six months, despite the potential inflationary consequences of German economic and monetary union. The other EMS currencies have also appreciated in effective terms over the same period, although the Lire rose by only 3 per cent. These exchange-rate appreciations in Europe were partly the consequence of an increase in interest rates, especially in France and Italy, along with a fall in rates in the US (recent and forecast interest rates are set out in table 3). if the increased differential in favour of Europe is expected to be maintained for some time then European exchange rates will rise as market participants take on board the new information. A higher differential for a given risk premium will produce a greater rate of depreciation (or a smaller appreciation) in the future to offset the higher return, and this depreciation will in general be preceded by a currency appreciation. However, we would not claim that the rise in European rates has been the major factor behind the strength of the D-Mark, and interest rates in italy and France have risen mainly to keep the EMS intact given the strength of the D-Mark.

The D-Mark has risen over the last six months mainly as a consequence of developments in the East. Firstly the influx of a large number of skilled migrants and secondly the prospect of a much increased German labour force after unification have meant that the outlook for German competitiveness is much improved. This large increase in the labour force is likely to put downward pressure on wages, and hence will cut costs. If the D-Mark had not appreciated then the current account surplus would have increased considerably, and would have exceeded long-term sustainable capital outflows.

This argument on the causes of the D-Mark appreciation is in line with the Institute's approach to the evaluation of sustainable exchange rates using FEERs (see Barrell and Wren Lewis (1989)). The FEER (Fundamental Equilibrium Exchange Rate) is the rate at which the current account exactly matches the long-term sustainable set of capital flows. The FEER will not in general be equal to Purchasing Power Parity (PPP). This is partly because PPP is only useful as a very long-term concept, because it applies to both traded and non-traded goods. Even if the law of one price holds for traded goods a bundle of all goods bought in one country may not cost the same as in another because efficiency wages in the non-traded sectors may differ considerably. These differences in efficiency wages will be reflected in non-traded goods prices, and experience even in Europe suggests that it takes a long time to compete them away. PPP appears to be useful in tying down exchange rates in the very long run, but for it to hold rates of prof it and efficiency wages would have to be the same everywhere. We seem to be rather a long way from this situation in the world economy today. We would argue that for medium-term analysis of exchange rates one should use empirically based trade and current account equations and also analyse internal balance using the concept of the NAIRU (see Barrell and Wren Lewis (1989)). Calculations of FEERs give intuitively more plausible explanations of exchange rates than PPP calculations. In particular it is difficult to understand the implications of PPP calculations for Japan and Germany. The OECD in their most recent Main Economic Indicators calculate that these persistent surplus countries are overvalued by 20 to 30 per cent in PPP terms. Achieving PPP would produce even larger current account surpluses, and it is difficult to see how these could be sustained by capital flows in the long run.

There are other problems with using PPP. Countries that run persistent surpluses, such as Japan and Germany, can build up large stocks of net overseas assets, and these will produce income flows that will persist. Even if Germany or Japan wished to return to current account balance they would have an exchange rate that produced a deficit on trade in goods and services, which could be seen as a rate that is above underlying PPP. This is because their property income surplus would have to be offset by a deficit elsewhere, and this can only be generated in a world of imperfect competition by holding the exchange rate above PPP.

We would also argue that countries do not run persistent surpluses by accident. Any group of individuals that wishes to build up their net wealth can either invest in physical assets or acquire financial assets from non-members. For a country this means that its net wealth is either in its physical capital stock or in its set of net overseas assets. The desire to accumulate wealth will depend upon the age structure of the population, the time preferences of the population at each age, bequest motives and the rates of return available. Given these factors, the choice is to either invest in physical assets at home or to build up net claims on foreigners. Low profit countries such as Germany will accumulate wealth by investing abroad, and as a consequence will raise output and improve efficiency abroad, and in the long run this will reduce profits abroad toward German levels.

It is very important to also bear these factors in mind when analysing recent developments in the Japanese exchange rate. Japan has, over the last 30 years, had a rapidly growing population and a greatly increased life expectancy. Hence there is a very large body of savers compared to dissavers in the economy. Gross savings by individuals are therefore high, with the personal savings ratio currently around 14 per cent. Although investment in business is around 25 per cent of GNP, much is internally financed, and even given the relatively high rates of return on Japanese investment, there is still the need to invest a good proportion of savings in overseas assets. An analysis of the factors affecting savings and the equilibrium pattern of net assets, as in the IMF paper Net foreign assets and international adjustment' is important when analysing the FEER for Japan, and this is central to our exchange-rate forecast which is contained in table 4.

Although the Yen has fallen sharply recently, we are not anticipating a sharp appreciation. The reasons for the timing of the fall of the Yen are discussed below, but our central argument is that the current account surplus being generated by the real exchange rate in 1989 was too low for equilibrium capital flows. Hence the exchange rate has had to fall. We are anticipating a sequence of surpluses throughout the 1990s, declining from over 2 per cent of GNP to around 1.5 per cent. These surpluses reflect our view of the sustainable capital outflows from Japan over this period. The level and path of the exchange rate is clearly dependent upon our assumptions on sustainable capital flows. We have looked at this issue elsewhere (Wren-Lewis and Barrell (1990)). if the sustainable level of capital outflows from Japan were 1 per cent of GNP lower than we have assumed then we calculate that the Yen exchange rate would have to appreciate by around 10 per cent. We do consider that this is a distinct possibility but this is not our main case. This calculation uses the GEM trade equations and internal balance relation along with actual GEM forecast of net property income flows in a steady state version of the model. The sensitivity analysis in this paper suggests that it is the relatively closed US and Japanese economies whose real exchange rates have to change most for a given change in structural capital flows.

Our exchange-rate forecast for all currencies is based upon the assumption that desired net overseas asset paths and hence sustainable capital flows have to be reached by the end of our forecast horizon in 1999. These target current accounts and hence real exchange rates are reached by assuming exchange rates change in line with interest differentials. Therefore our forecast of interest rates is central to our exchange-rate projections, as are our assumptions about risk premia. The buildup of pressure for European Monetary Union has changed our assumption in this area. We have presumed that union will be formed in 1997, and that it is anticipated. Risk premia will be much lower inside the union than without it, and as the union approaches we believe that risk premia will fall. At the start of our forecast we have a risk premium on the French Franc against the D-Mark of 0.5 and on the Italian Lire are of 2.0. However, we expect these will begin to fall by the middle of the decade and at least in the case of France we would expect them to disappear by 1997. We anticipate that there will be several EMS realignments in the run up to full union, with the first occurring before the end of 1991. A monetary union in Europe at the end of the decade affects our growth and inflation forecasts. Following the argument in the note on the EMU and the EMS in this Review we have assumed that a successful monetary union will require fiscal tightening by its weaker members such as italy, and higher interest rates in Germany.

The United States

Growth in the US remains strong, and the outlook for inflation in the shorter term has worsened. First estimates of GNP growth in the first quarter of 1990 suggest that output growth was 2.1 per cent at an annual rate, compared to 1.1 per cent in the fourth quarter of 1989. This figure, however, hides some of the strength of the economy. Final sales, excluding stockbuilding, rose by 4.0 per cent and final domestic demand grew by 3.4 per cent both at an annual rate. inventory investment hardly grew, partly because of considerable destocking of automobiles. This destocking was partly associated with strong growth in consumer durables demand, and strong equipment investment also reduced stockbuilding.

Indicators of inflation generally suggest that prices are rising more rapidly now than they were a year ago. Producer prices rose by 6.75 per cent at an annual rate between December 1989 and March 1990, but were only 4.4 per cent higher than a year previously. First estimates of the GNP deflator suggest that it also rose by more than 6 per cent at an annual rate in the first quarter. Consumer price inflation has also risen, averaging over 5 per cent in the first quarter compared to the same quarter a year previously. This is half a per cent above the rate seen in the last quarter of 1989. The CPI rose at an annual rate of 8.7 per cent between December 1989 and March 1990. Only a small proportion of the increase in inflation can be attributed to the fall in the Dollar over the last year. The trade weighted index fell only by 3.75 per cent between April 1989 and April 1990, and given the US's low import propensity the effect on prices is likely to have been small. The major factor associated with rising prices has been domestic employment costs that have been driven by a tight labour market, indicating that inflation has been driven by excess demand in the domestic economy.

Unemployment fell in March to 5.2 per cent from 5.3 per cent, its first fall for nine months. All these factors indicate that inflationary pressure may be rising. There are, however, signs that the tightening of credit in 1988 and early 1989 is having some impact on demand. Unemployment rose again in April to 5.4 per cent, and employment growth has been weak. Housing starts are falling, and are 6 per cent below a year ago. industrial production growth has been strong over the last 2 years, but it has clearly slowed from 5.6 per cent in 1988 to only 0.7 per cent at an annual rate between December and March of 1990.

Our forecast, which is set out in table 5, reflects these recent developments. We are forecasting that output growth will drop to 2 per cent in 1990, well below the 3.5 per cent a year experienced between 1985 and 1989. This projected growth rate is also below our estimate of the growth of capacity, and as a result we are projecting that employment growth in the year will only be 0.5 per cent, and that unemployment will rise into 1991 to over 6.25 per cent. This slowdown in the economy comes both from the low level of business investment growth that is a consequence of high interest rates, and also from the effects of reductions in government spending. We are expecting reductions in defence spending to begin to take effect this year. We have assumed that the 25 per cent cut over five years in real defence spending that is currently being discussed will not all take place, but that there will be major cuts in the growth of expenditure programmes over the next three to four years. We are assuming real government expenditure will grow by O.5 percent per annum from 1991-3.

The projected slowdown in activity is, however, only temporary. The peace dividend from Eastern Europe has reduced spending plans and has therefore reduced the pressure on the executive to raise taxes and cut social welfare programmes. Both should help keep consumers' expenditure buoyant in the early 1990s. We are also assuming that once the process of budget deficit reduction is well under way then there will be less pressure on spending programmes, and expenditure growth will rise again in the second half of the decade, albeit not to the same extent as in the 1980s. Table 6 sets out our forecast for the US public sector.

The overall forecast has implications for our current account forecast, which is set out in table 7 below. There has been a marked improvement in the current account since 1987. The current account deficit appears to have been $106 billion in 1989, and we are expecting a slight improvement to $103 billion in 1990. This partly reflects slow import growth, which follows from our forecast of slower growth in the economy. However, in the medium term we are expecting the current account to begin to deteriorate again as import growth rises and despite strong export growth. This deterioration is driven by two factors. Firstly strong domestic demand growth, especially after the first wave of major defence cuts, will help raise imports. Secondly we are no longer projecting a major decline in the Dollar. Interest rates in the rest of the world have risen more than in the US in the last few months, and inflation prospects seem recently to have deteriorated more in Japan and Europe than in the US. These factors have led us to revise upward our forecast of interest rates outside the US, which in turn leads to a lower rate of depreciation of the Dollar through our use of the open arbitrage condition.

We believe that the current account deficits we are projecting for the 1990s are sustainable. They are around 2 per cent of US GNP, and are therefore considerably smaller than those experienced between 1984 and 1988. They are also associated with a declining share of government spending in GNP and a rising level of business investment. The shift from defence to consumer goods industries in the 1990s that will result from the peace dividend is likely to be more marked in the US than elsewhere in the world, and this is likely to attract large scale capital inflows because levels of overseas direct investment in consumer goods industries are far higher than in defence industries. There is a limit to the size of cumulating US deficits, as the change in national wealth is likely to put downward pressure on consumption, but the effects of the rising net debtor position will be felt only slowly.

Japan

Both short-term prospects and longer-term developments in Japan are dominated by the outlook for the Yen. The exchange rate has been falling sharply recently, and in the first quarter of 1990 the Yen/Dollar rate was 16 per cent below the level of the same quarter in the previous year. Over the same period the effective exchange rate fell by 10 per cent. Over the period 1984 to 1988 the Japanese current account surplus averaged 3.5 per cent of GNP. Despite this in each year long-term capital outflows exceeded the current account by an average of $40 billion. These capital outflows reflected both the savings behaviour of the society and also the progressive removal of outward capital controls in Japan. A combination of pressure from the Americans and a belief that profit opportunities had improved in Japan put upward pressure on the Yen during 1988 and early 1989. As a consequence the current account surplus fell to $57 billion in 1989. We feel that this decline in the external surplus was excessive, and did not leave enough room for the long-term capital outflows likely to be generated over the next decade. The pressure of capital outflows was masked somewhat during 1989 as Japanese corporations induced a capital inflow by issuing $40 billion more securities abroad than they had in 1988. However it is clear that this capital inflow was a temporary phenomenon, resting in part on the very high level of the stock market index. The realisation of the excessive decline in the surplus and the recognition that some capital inflows would be short lived contributed to the decline in the Yen between the middle of 1989 and the first quarter of 1990. We do not feel that these factors alone would, however, be enough to push the Yen below the level it reached in the first quarter of 1990.

Our short-term forecast is heavily influenced by the effects of the fall in the Yen up to the first quarter. However there are two further factors to take into account. Firstly the Nikkei index began a precipitate decline in February, and by the end of April the index had fallen 30 per cent from its end 1989 level. Secondly the exchange rate fell to 158 Yen per Dollar in early May. The fall in the stock market was almost inevitable, with only the timing of the fall hard to predict. It had been widely agreed that the index was overpriced. Not only were P/E ratios, at around 50, well above those seen in other stock markets, but more importantly they were more than double the levels seen on the Tokyo exchange between 1976 and 1986. The fall in the stock market is not unconnected with the further fall in the Yen. Sales of securities abroad have virtually ceased, and some houses are now reallocating their portfolios toward overseas markets This temporarily large capital outflow has put further, temporary, pressure on the Yen, and we think that with the cessation of the fall in the stock market this will be removed and the Yen will rise somewhat. Our short-term forecast is based on the presumption that the Yen will be around 145 to the Dollar in the third quarter.

Inflationary pressures have clearly emerged in the Japanese economy. In the last quarter of 1989 wholesale prices were rising by 3.75 per cent and consumer prices by 2.5 per cent, both at an annual rate. First estimates suggest that consumer price rises increased to 3.75 per cent in the first quarter of 1990, and wholesale prices also continued to rise. The Japanese economy is still relatively closed, with average import penetration well below 20 per cent, and hence the decline in the Yen is unlikely to exert any major inflationary impulse. Most inflationary pressures in Japan reflect domestic demand pressures, and unemployment has continued to fall into 1990, reaching 2.2 per cent in March. The job offers to applicants ratio is no longer rising as fast as it did in 1987 and 1988, but at 1.37 it is at its highest level for over a decade. Wage costs have been increasing, with earnings growth reaching 5.75 per cent at an annual rate in the fourth quarter of 1989. The emergence of some capacity constraints has also led to a decline in the rate of productivity growth, and unit labour costs are now growing by between 2.5 per cent and 3.5 per cent at an annual rate. Wage increases from the Spring Round Negotiations were 5.25 per cent in 1989, up from 4.5 per cent in 1988, and they are expected to be 5.75 per cent to 6 per cent in 1990.

We are anticipating that domestic demand growth will slow down during 1990. The drop in stock market prices will reduce consumer wealth and hence is likely to have a negative effect on consumption. We are projecting that consumers' expenditure will only grow at 2.25 per cent in 1990, well below the average of 4.5 per cent in 1987-9. Housing construction is unlikely to grow much in 1990, partly as a result of the stock market fall, although housing starts were high at the beginning of the year. The fall in the stock market is also likely to affect business investment, in that it raises the cost of capital. However, at the beginning of the year non-manufacturing firms were expecting to increase the rate of growth of their investment and all industries were slightly more optimistic in the Bank of Japan's February `Short Term Survey of Enterprises'. We would expect these firms to be more pessimistic now than they were three months ago, and along with slightly higher interest rates, especially at the long end, this has persuaded us to revise down our investment growth forecast to 12 per cent, some 6 per cent below the growth rate seen last year. Our overall forecast for domestic demand is that growth will fall to 4.5 per cent in 1990 and 3.5 per cent in 1991 as tighter monetary conditions take hold. in the longer term we expect Japanese domestic demand growth to rise back to 4.25 per cent to 4.5 per cent on average.

Table 8 sets out our forecast for Japanese income growth, and Table 9 our forecast for the Japanese current account. We are expecting the Japanese external sector to contribute almost 1 per cent to GNP growth in 1990. Almost one half of this comes from the invisibles, both from services trade and also from the effect of net property income on GNP. There is no J curve effect for property income, and the competitiveness elasticities for services on our model are not only quite large but also fast acting, and we have allowed the effect to feed through. We are expecting some slowdown in world trade in 1990, but the combination of 6 per cent growth in markets and at least a 10 per cent drop in the exchange rate should raise Japanese export growth to over 8 per cent in 1990, and given the improvement in Japanese competitiveness we expect export growth to be even higher in 1991. The combination of slower domestic demand growth and improved competitiveness should also reduce import volume growth in 1990.

We expect the overall effect of visible trade on GNP to exceed 0.5 per cent in 1990, mainly as a result of the fall in the Yen. We do, however, believe that the relationship between imports and demand has changed in Japan, and we are projecting quite rapid non-oil import growth, most of which will be concentrated on manufactured goods. The Japanese are under much pressure to open up their economy, and we expect that the reduction in barriers, combined with lower export growth, will result in domestic demand growth exceeding GNP growth from 1992 onwards. This process of opening the Japanese economy will be aided by the prospects for higher inflation. Excess demand and the fall in the exchange rate are expected to produce inflation of 2.75 per cent in 1990. We are expecting the Japanese economy to continue to operate near capacity throughout the 1990s and hence there is no domestic factor causing inflation to fall. We are therefore forecasting Japanese inflation of 2.5 per cent a year for most of the 1990s. in the long term this will reduce Japanese competitiveness, albeit only marginally, and will contribute to the longer-term decline in the current account surplus.

Germany

The West German economy approaches economic and monetary union with East Germany on July 2nd against a background of two years of strong GNP growth, accompanied by rising, although still low, inflation. This background has led to a tightening of West German monetary policy designed to slow down the growth of demand. West German GNP growth in 1989 was bolstered by net exports which accounted for 1.2 percentage points of overall GNP growth of 4 per cent. The prospects for net exports are however much worse this year, with a projected decline in the rate of growth of economic activity in the US, France, Italy and the UK, together with an appreciation of the D-Mark, notably against the Japanese Yen, the US Dollar and against Sterling. A tighter monetary policy and bleaker external factors will both act to restrain economic activity in West Germany. Economic and monetary union with East Germany is however likely to provide a renewed stimulus to the West German economy, which may enable it to grow by more than 3 per cent for the third successive year.

The most recent statistics on the West German economy suggest that the monetary tightening which took place last year is beginning to have its intended effect. Wholesale price inflation in February 1990 was only 0.5 per cent, down from 6.5 per cent in May 1989, and consumer price inflation was 2.8 per cent, down from 3.3 per cent in October 1989. The IFO index of business climate, derived from surveys of manufacturing industry, was lower in February than in any month since May of last year, following three especially high figures in November, December and January. This perhaps indicates a revision of business expectations about the likely impact of economic union. The level of the index continues to suggest however that business confidence remains high. The rate of capacity utilisation fell slightly in the fourth quarter of 1989, but also remains high at 89.3 per cent. Labour market figures show a drop in unemployment to 7.4 per cent in February 1990, down from 7.8 per cent in November 1989. It therefore appears that the rise in inflation has been checked without incurring a significant cost in business expectations or output.

Our forecast for West Germany is presented in Table 10. In constructing the forecast we have continued to use our econometric model, GEM, but have made a number of adjustments both to the model and to equation residuals in an attempt to incorporate some of the likely consequences of and responses to economic and monetary union. The changes we have made to the model are outlined in Box 1.

The first such residual adjustment is on the equation for consumers' expenditure. We expect consumers' expenditure to grow by more than our equation would otherwise predict for two main reasons. Firstly there is the effect of immigration into West Germany. These immigrants will have brought their savings to the West, some of which will be spent on consumption goods. The second effect will largely occur after economic and monetary union in July, and hence have most impact in the latter half of this year and in 1991. The terms of the union have been set to allow East German adults to convert 4,000 Ostmarks into D-Mark at a one-for-one exchange rate. Additional savings will be converted at 2 Ostmarks for each D-Mark. Pensioners will be able to convert up to 6,000 Ostmarks at parity, but children will only be allowed to convert 2,000 Ostmarks at this rate. The scale of East German savings is not known for certain, but is estimated at around Ostmark 180 billion. These savings can be divided into two categories, which we will term voluntary and involuntary savings. Voluntary savings are those which the population would choose as assets in order to facilitate life-cycle consumption. involuntary savings are those held merely because appropriate consumption goods are not available at an affordable price. Considerable price differences exist between the two Germanies. In 1988 a refrigerator cost 1,425 Ostmarks but only 559 D-Mark, and a television set 4,900 Ostmarks against 1,539 D-Mark. It is expected that there will be a boost to consumption following economic and monetary union as some of the East German involuntary savings is spent on cheaper West German consumer durables.

Prices, however, are not necessarily cheaper in the West. The existing East German economic system subsidises many basic items. In 1988 5 kg of potatoes cost 0.85 Ostmarks but 4.94 D-Mark, 1 KWH of electricity 0.08 Ostmark against 0.42 D-Mark, and the monthly rent for comparable apartments cost 75 Ostmarks against 411 D-Mark (source Dresdner Bank (1990)). Details on the treatment of price subsidies have yet to be finalised, but they will need to be phased out rapidly in order to prevent arbitrageurs capitalising on East German subsidies. The cost of living is therefore likely to rise markedly in East Germany, as prices adjust to West German levels. This effect is likely to curb some of the growth in spending on consumer durables that would otherwise have occurred.

The scale of increased demand from East Germany will be dependent on the viability of East German enterprises following monetary union, together with the extent to which the East German workers can be absorbed into the West German labour market. Initially wage-rates are being converted into D-Mark at the one-to-one exchange rate and corporate debt at two-to-one. The viability of enterprises at this exchange rate is difficult to assess: the question hinges on the flexibility of wages once monetary union has occurred. Greater inflexibility increases the likelihood that East German enterprises will prove unviable and hence that unemployment in the East will rise. A second factor will be the extent to which either German government is willing to subsidise any East German enterprise which finds itself uncompetitive. However the risks that uncompetitiveness will entail for East German unemployment need to be weighed against the lower wages of East Germans compared to West Germans, and hence increased employment opportunities for them in the West. In the medium term lower East German labour costs will result in new enterprises being established in the East, but in the short term it is more likely that East Germans will commute to existing West German locations.

Our second residual adjustment has been made to our investment equations. Economic and monetary union will provide a considerable impetus to western investment within East Germany itself. In this respect East Germany should fare considerably better than Poland and Hungary in its transition to a market economy. The latter countries have not benefited as much as they had hoped from inward western investment, largely it would appear because of western wariness about the permanence of the new economic and political systems. It is likely that western firms will also be hesitant about investing in East Germany until they are more confident that it will also enter a political union with West Germany. In the short term they will be able to supply the East German market and still benefit from cheaper East German labour by expanding their operations within West Germany itself.

Our third residual adjustment has been to our equation for employees' wages and salaries. This reflects our expectation that there will be downward pressure on average earnings, caused by the expansion of the labour supply, including many skilled workers used to receiving lower wages. This reduction in average earnings has a beneficial effect on German employment and also reduces inflation in the short-to-medium term, allowing the West German economy to expand faster than it would otherwise have done without incurring inflationary bottlenecks.

Table 10 shows that with these residual adjustments we are forecasting West German GNP growth of 3.1 per cent. This is lower than many other forecasts which we feel have exaggerated the immediate effects of economic union. We anticipate a surge in consumption associated with spending out of previously involuntary saving, and also strong growth in investment and government expenditure. However the impact of these effects is likely to affect 1991 more than 1990. Within 1990 the expansionary effects of economic union will be partly offset by the deflationary effects of higher interest rates, a higher exchange rate and a deceleration in the growth of export markets. GNP growth should increase to 3.9 per cent in 1991, mainly because of an improved net exports performance, which can be attributed to our forecast of both an increase in export market growth and of a much more moderate appreciation in the effective exchange rate.

We anticipate that inflation will remain under control. In 1990 the appreciation of the exchange rate should assist in reducing the inflation rate to 2.6 per cent. After 1990 the inflation rate is further assisted by the downward pressure on average earnings associated with the increased labour supply. We are also forecasting a gradual appreciation of the D-Mark, in line with the interest parity condition. We anticipate that interest rates will remain at approximately current levels for the remainder of this year, and will be gradually reduced thereafter as the inflationary threat recedes.

Table 11 shows our forecast of the German current balance. This differs from previous forecasts in showing a marked decline in the current balance surplus as a share of GNP. This can be explained in two ways. Firstly we are expecting a higher level of demand in West Germany. Part of this extra demand will be met from within West Germany itself, possibly at the expense of production that would otherwise have been exported. The remaining part will be met by increased imports. The second explanation is in terms of the savings-investment balance within West Germany. As a consequence of economic union and East German reconstruction, we expect that greater investment will occur within West Germany. We do not assume however that West German savings will increase to the same extent, so that some of this increased investment will be financed from abroad. The consequence is an increase in capital inflows into Germany, or equivalently a reduction in current account inflows.

The potential unification of Germany will have little effect on published trade statistics. Trade between the two Germanies was not recorded as external trade because the Federal Republic refused to recognise the division of Germany. increased exports to the East will be equivalent to either higher domestic consumption or investment. Trade between East Germany and the rest of the world is rather limited and we expect it to stay so. We therefore expect our new current account system, reported in Annex 1, to be useful for forecasting purposes for some time.

Table 12 sets out our new German public sector forecasts. Revenues have been very buoyant in the last year, but cuts in direct taxes at the start of 1990 have reduced our projection of revenues for the year. Expenditure growth was low in both 1988 and 1989, and as a result the public sector deficit fell from 43 billion D-Mark to 25 billion D-Mark, and the Federal Government looks like it will have a surplus in 1990/1. (Box 2 discusses the German public sector and explains why we do not model the Federal deficit).

The German public sector as a whole would have been in surplus by 1993 if policies had not been changed and if the events in the East had not occurred. However, the wave of migrants and the prospect of unification have affected our expenditure forecasts. We have assumed that transfers to those resident in the West will rise in 1990 by 10 billion D-Mark more than our equation suggests, reflecting transfers to migrants, and that government expenditure on goods will rise by 12 billion D-Mark more than the equation suggests. We have also assumed that up to 30 billion D-Mark of East German corporate debt will be consolidated into the Federal Republic's debt, and as a result interest payments will be higher than they would otherwise have been. Our assumptions produce an increase in the public sector deficit to 59 billion D-Mark in 1990, but buoyant revenues thereafter will reduce the deficit in subsequent years. We have not presented longer-term forecasts of the debt and deficit as unification will alter the structure of this sector of our model.

France

French GNP grew by 3.7 per cent in the year to the fourth quarter of 1989, giving a growth rate of 3.4 per cent for the year as a whole. This was only fractionally below the growth of 3.5 per cent realised in 1988, and was achieved with only a modest increase in inflation from 2.7 per cent in 1988 to 3.3 per cent in 1989. Domestic demand growth slowed to 3.2 per cent in 1989, but net exports improved, aided by a lower exchange rate and strong demand growth in the rest of the world.

Two years of strong growth are reflected in the survey evidence for capacity utilisation in manufacturing. This showed an increase to 85.8 per cent utilisation at the end of 1989, compared with 83.3 per cent at the end of 1 988 and 80.8 per cent at the end of 1987. Unemployment has fallen, although rather slowly, standing at 9.4 per cent of the total labour force in February 1990, compared to 9.6 per cent a year earlier. The growth of hourly wage rates rose in 1989 to 3.9 per cent from 3.2 per cent in the two preceding years.

Monetary policy was tightened in 1989 in order both to restrain the growth of domestic demand and to maintain the Franc's exchange-rate parity in the face of interest-rate rises in Germany. The Franc has appreciated along with other EMS currencies, notably against the Yen, the Dollar and Sterling, and since December has appreciated also against the D-Mark. Our forecast assumes that the exchange rate moves in line with interest-rate differentials, which means that the appreciation is not rapidly reversed. The recent strength of the Franc has allowed French interest rates to be reduced from their peak in January, but they remain higher than they were in the first three quarters of last year.

Our forecast for France is presented in Table 13. With interest rates and the exchange rate both higher than last year, and with an expected decline in world demand, the French economy is expected to grow more slowly this year. Domestic demand growth is forecast to slow to 2.7 per cent as consumption and investment expenditure are both restrained by higher interest rates. The less favourable external environment is likely to result in net exports reducing GDP growth by 0.6 per cent. The appreciation of the Franc will exert renewed downward pressure on inflation, which is expected to fall to 2.9 per cent this year. The monetary and economic union in Germany should provide France with some extra demand, as the German demand for imports is likely to rise.

Box 3 highlights the French achievement in reducing inflation over the course of the 1980s. The principal policy instrument that has been used is the exchangerate mechanism of the European monetary system. This has meant that French monetary policy has had to be closely co-ordinated with that of other ERM members, most notably Germany. The continued operation of this policy in our forecast means that interest rates are likely to remain close to current levels for most of the year, but will be gradually reduced thereafter. This should provide a stimulus to both consumption and investment, allowing domestic demand growth of around 3 per cent per annum.

The inflation differential with Germany may rise a little in the early 1990s, as France will not share the downward pressure on average earnings exerted by Germany's expanded labour force. We have assumed that the `Franc fort' policy described in Box 3 will continue and that the ERM will be developed into a full monetary union by 1997. We expect that this will exert a continuing discipline on the growth of unit labour costs, and hence check the growth of domestically generated inflation. Control of inflation is required to maintain the competitiveness of French industry, and prevent an unsustainable deterioration of the current account balance. Our forecast of French trade is presented in Table 14. We are forecasting that the current account will move into deficit of around 1 per cent of GDP by the end of the decade. Such a deficit can only be sustained if France is able to attract the necessary capital inflows to finance it. If these are not forthcoming then there will be a need to adopt a more restrictive fiscal policy in order to curb the growth of domestic demand, and hence of imports.

The French government remains keen to develop the EMS into a European monetary union. This is a very different proposition from the German monetary union. On the one hand, transition to European monetary union would be economically simpler, since the European countries are operating fully developed market economies which have already achieved a high degree of inflation and monetary policy convergence. But European monetary union raises questions of sovereignty and new institutional controls which are likely to prove more difficult to resolve than the German monetary union.

Italy

Figures for the third quarter of 1989 suggest that in contrast to Germany and France, growth in the Italian economy slowed appreciably from its cyclical peak in 1988. We estimate GNP growth of 2.9 per cent in 1989 compared to 3.9 per cent in 1988. The slowdown was mainly due to a decline in domestic demand, which in turn was accounted for by much lower stockbuilding and by slower growth in government spending.

In spite of this apparent deceleration in the growthrate, the level of capacity utilisation remains high at 80.2 per cent in the fourth quarter of last year. Survey evidence suggests that business confidence also remains high. Questions on the future tendency of production and the prospects for the total economy produced positive balances of 29 and 14 percentage points respectively. Sales of manufacturing and mining goods rose by 6.6 per cent in the year to the third quarter. Unemployment, which had remained between 12.0 per cent and 12.2 per cent throughout 1988 and the first three quarters of 1989, fell to 11.6 per cent in the fourth quarter. There is therefore some evidence of continued business optimism in the Italian economy.

Inflation signals are rather mixed. Wholesale price inflation peaked at 7.1 per cent in March 1989 and had fallen to 5.4 per cent by December, but wage inflation rose from 5.6 per cent to 6.9 per cent over the same period. Consumer price inflation stabilised at around 6.5 per cent in the latter half of the year.

Italian monetary policy was tightened during the course of 1989 and the discount rate further raised from 12.5 per cent to 13.5 per cent on 3 March of this year. At the start of the year the italian government announced that the Lira would be held within the narrow band of 2.25 per cent around its central EMS parity. At the time of the announcement the Lira stood at 1.33 D-Mark, but had risen to 1.36 D-Mark by the beginning of May, close to the top of its new band. The currency continues to be supported by the high level of Italian interest rates compared to other ERM participants, and the currency's rise suggests that exchange markets consider the new band for the Lira to be credible.

As a consequence of its membership of the exchange-rate mechanism in 1990 the Italian economy will share the deflationary effects of higher interest rates and exchange rates with France and Germany. Coupled with declining demand in the rest of the world these factors are likely to add impetus to the cyclical downturn. This impetus should however be in part offset by the stimuli provided by German economic and monetary union, and by movements towards the single European market.

The Italian economy is however distinguished from France and Germany by two difficulties. The first is the continued problem of its large public sector deficit. In mid-March projections presented to Parliament suggested that the state borrowing requirement for 1990 would overshoot its target of L133 trillion by L14 trillion. The problem of managing the deficit is exacerbated by high interest rates which however are as much effect as cause of the deficit. The second Italian difficulty is its continued inflation differential with its European partners. The recent acceleration in wage inflation indicates that inflationary tendencies have not been completely suppressed. The Government's decision to place the Lira within a narrow ERM band can be seen as an attempt to consolidate its progress in achieving inflation convergence, with its implicit warning to Italian firms that inflationary settlements will not be accommodated by exchange-rate devaluation.

Our forecast for Italy is presented in Table 15. For 1990 we expect that the twin restraints of high interest rates and an appreciated exchange rate will lead to a further reduction in GDP growth to 2.2 per cent. Slower growth of consumption and investment contribute to a decline in domestic demand growth to 2.8 per cent, and the less favourable external environment leads to an increased negative net export contribution. Some reduction in inflation is expected, occurring mainly as a consequence of lower import prices following the recent Lira appreciation.

In 1991 we anticipate that growth will increase to 2.7 per cent. We anticipate that there will be some scope for lowering interest rates, which should aid the growth of domestic demand. Exports should benefit from increased activity in Germany, and the deflationary impact of the recent appreciation should have begun to dissipate. We nonetheless expect further progress to be made in reducing inflation, as wage settlements moderate. A major risk to our forecast is that wage-inflation may not respond so quickly.

Our forecast assumes that the authorities continue to pursue a policy of achieving inflation convergence with Germany through the EMS, and that they further move towards participation in a full monetary union. We have assumed that this is achieved by 1997. As explained in the note on European monetary union in this issue of the Review, italian entry into a full monetary union requires a tightening of fiscal policy in order to prevent inflationary bottlenecks occurring as a result of excessive Italian demand.

Canada

GNP growth in Canada in 1989 is estimated to have been 2.5 per cent, which represents the first year that growth has been below 3 per cent since the recession of 1982. This deterioration in the growthrate is attributable to a marked reduction in the contribution of net exports. Domestic demand growth of 6.1 per cent was higher than any other year in the 1980s, due to continued strong growth in consumption and investment, and a considerable amount of stockbuilding. The deterioration in net exports may be attributed to the rise of the exchange rate over the last three years. During 1987 the Canadian Dollar appreciated 5.3 per cent against the US Dollar, during 1988 7.9 per cent, and during 1989 3.3 per cent. This appreciation was much greater than the inflation differential between the two countries, so that Canada experienced a loss of competitiveness. As Canada is a very open economy, with exports and imports of goods both accounting for more than 30 per cent of GDP, its output is very susceptible to large changes in the terms of trade.

Our forecast for Canada is presented in Table 16. We anticipate that domestic demand growth will slow to 2.5 per cent. Consumption and investment growth are both likely to decline as a result of the higher level of interest rates that have prevailed since the second half of last year. We expect interest rates to remain at their present levels for most of the year. Stockbuilding is also forecast to decline this year from the exceptionally high level recorded in 1989. The outlook for net exports should begin to improve. We are forecasting a depreciation of the exchange rate in line with interest-rate differentials, adjusted for a risk premium which we have assumed initially accounts for 1 percentage point of the differential. The combination of a lower growth in domestic demand and a depreciating exchange rate should lead to a lower growth of imports, while export growth should rise as a result of Canada's improving competitiveness.

The Canadian government is currently in the middle of a tax reform programme initiated in January 1988. Stage II of the reform will be the replacement of the current Manufacturers Sales Tax, by a Goods and Services Tax (GST) in January 1991. The new GST is a more broadly based value added tax, which will be levied at a rate of 7 per cent. It is expected that the medium-term impact of the tax will be to make Canadian exports more competitive and imports less competitive. It should also reduce the cost of capital and consequently boost investment. The immediate effect of the tax is however likely to be a temporary increase in the rate of inflation in 1991, which may dampen economic activity through its adverse effect on the growth of real personal disposable income. GNP growth in 1991 should recover as net exports continue to improve, as a consequence of the new tax and of a continued depreciation of the exchange rate.

We anticipate that monetary policy will continue to remain tight in 1990 and 1991 as the Government attempts to prevent an acceleration in inflation. if this endeavour is successful interest rates can be expected to fall in 1992, 1993 and 1994. This should provide a stimulus to activity in the middle years of the decade.

REFERENCES

Barrell, R.J. (1990), `The EMS and European Monetary Union', National Institute Economic Review, No. 132, May.

Barrell, R.J. and S. Wren-Lewis (1990), Fundamental equilibrium exchange rates for the G7', CEPR Discussion Paper No. 323.

Britton, A.J. and D. Mayes (1990), 'Obstacles to the use of the ECU: macroeconomic aspects', Economic Journal, forthcoming.

Dresdner Bank (1990), `Germany monetary unification', Trends Special, March, Frankfurt.

European Commission (1990), 'Economic and monetary union: the economic rationale and design of the system', BIS Review, No. 62, March.

Horne, J., J. Kremers, and P. Masson (1989), `Net foreign assets and international adjustment in the US, Japan and the Federal Republic of Germany', IMF Working Paper No. 89/22, March.

Wren-Lewis, S. and R.J. Barrell (1990), `Calculation of FEERs using steady state GEM' for John Williamson's 'Equilibrium exchange rates: an update', May.