The world economy.

Barrell, R.J. ; Gurney, Andrew

CHAPTER II. THE WORLD ECONOMY

The overall outlook

Our February forecast suggested that developments in the short term would be dominated by fears of accelerating inflation and policy responses to them. This has indeed been the case. In Japan, Germany and US wholesale prices have begun to rise relatively rapidly. Although commodity prices, especially of metals and minerals and of developed county foods, have fallen in recent weeks, at least in dollar terms they remain high and oil prices appear to have hit temporary peaks at the beginning of the quarter. These developments are the result of demand pressure. Our equations for real commodity prices, which were reported in the August 1988 issue of the Review, do have rather strong influences from world industrial production in them. As commodity prices are more timely than figures for demand and output they have often been early indicators of rising demand and we believe that they are currently, and correctly, filling this role.

Once again interest rates have risen sharply in the last three months, especially in Germany but also in France and the US, and to a lesser extend in Japan. These interest-rate rises have been a clear, and publicly claimed, response to fears of rising inflation. Short rates are now 3 per cent higher in Germany than they were a year ago, and they have risen by 2-1/2 per cent in the US and by 1/2 to 1 per cent in France and Japan. The rise in long rates has been considerably less, suggesting that the financial markets expect the contractionary monetary stance that is being adopted to be successful in reducing inflation in the long run. Chart 1 plots recent interest-rate rises in the major four economies over the recent past.

Table 1 summarises our forecast. Inflation will rise strongly into this year, but the response of the monetary authorities in raising interest rates will reduce inflationary pressures in 1990 and beyond. In recent months capacity utilisation in all of the major seven economies has been running at the highest level observed for over a decade, and price rises, both in consumer and producer markets, have accelerated markedly. Chart 2 plots recent inflationary developments. We are forecasting that consumer price inflation will peak during 1989, and gradually decelerate thereafter.

Strong growth has continued into the fourth quarter of 1988 and the first quarter of 1989. Although GDP growth has been faster than we estimate is the rate of growth of productive potential, we are not anticipating much slowdown into 1989 and even 1990, except in the cash of the US. Tight monetary policy in the US is expected to slow growth to 1 per cent in 1990, and the fiscal policy risks are clearly not in an expansionary direction.

World trade growth was remarkably strong in 1988, and manufactured exports from the major industrial economies appear to have grown at over 10 per cent. Some of this growth has been associated with developments between the US and Canada where strategic position-taking prior to the 'free trade agreement' has influenced the patttern of trade. Total world trade in all commodities, which also includes the trade of the smaller OECD countries and of non OECD members, probably rose by 8 per cent in 1988 after a reasonably strong growth of 6 per cent in 1987. The combination of tighter monetary policy and unconducive exchange-rate growth in 1989.

The recent rises in oil and commodity prices will have improved the terms of trade of the smaller and poorer countries of the world, and this is likely in the short run to produce a higher growth rate of total world trade than we might have otherwise seen. A given value of commodity exports will now buy more manufactured imports, and hence, at least in the short run, world trade will rise. Trade growth will, however, slow down slightly in the long run, and we project that it will be around 5-1/2 per cent per annum in the early 1990s and that it may even slow somewhat thereafter as real commodity prices stop rising.

Exchange rates and interest rates

The recent and continuing strenghtening of the dollar may be a little hard to explain in terms of current account related flows, and it has had a short-term depressing influence on the growth prospects for the US economy. The strength of the dollar may reflect the market's recognition of the anti-inflationary resolve of the US monetary authorities. If interest rates rise in the US relative to elsewhere, as they have, and markets perceive that rates will remain higher for a long period of time, than the dollar is likely to strengthen in the short term both because the resulting contraction of the US economy will lead to a more rapid reduction of inflation there than elsewhere, and also because a higher interest-rate differential may require a more rapid decline in the value of the currency toward its equilibrium level. This latter is an implication of the 'uncovered interest differential' approach to forecasting exchange rates. In our forecast we assume that exchange rates head towards sustainable, or fundamental levels, but that the rate of approach to that level is determined by arbitrageurs who take up positions until the change in exchange rates that they expect is approximately equal to the observed interest differential between the two currencies involved.

Our whole exchange-rate forecast, which is set out in Table 3, is coloured by the implications of the open arbitrage condition. If a country has a higher than average interest-rate forecast for the future in table 2 then its exchange-rate path, as set out in table 3, will imply a faster than average rate of depreciation. We have also made allowances for risk premia and in particular we have attempted to allow for an increasing premium on Japanese holdings of US assets as Japanese portfolios become saturated with dollar denominated credits.

Recent changes in exchange rates and interest rates have not been conducive to a process of orderly adjustment of current account imbalances toward more sustainable levels. Exchange rates have probably been moving away from their equilibrium levels. The appendix to this chapter sets out the basis of our analysis of fundamental, or sustainable, equilibrium exchange rates, and these must be associated with equilibriating structural capital flows. The equilibrium exchange rates that we calculate are the path of the exchange rate that would produce a current account of the balance of payments that is exactly supported by structural capital flows. We have based our estimates on the structural characteristics of our world econometric model, GEM, and the equilibrium exchange rates differ from observed market rates for a number of reasons. A relative tightening of monetary policy may push the exchange rate above its equilibrium in the short run, as can large-scale speculative purchases of a currency. However, departures from the equilibrium rate will encourage equilibriating pressures both from trade flows and from the capital account. Although these may take time to work through they are an important influence in the market, and hence some knowledge of the location of the equilibrium exchange rate is important.

The dollar has strengthened considerably recently, and both the yen and the deutschemark (along with the franc) have weakened. Chart 3 plots exchange-rate developments over the recent past and for our short-term projection. In the short term there appears to be no tendency toward adjustment of the structure of current account imbalances toward what we perceive as a more sustainable pattern, but in the longer term there may be more orderly, if slow, progress. Chart 4 plots our forecast of the current account to GDP ratios for the major three economies over the period 1977 to 1997. Our presumption that interest rates will remain relatively high in the US, along with a belief that there must be some tightening of fiscal policy, produce a path that eventually looks sustainable. Very high Japanese demand growth, along with projected moves toward a more open economy, also produce a slowly declining Japanese current account surplus. However, on currently announced or anticipated fiscal and monetary policies, along with the characteristics of the German economy as embodied in our model, there seems to be little prospect for a rapid decline in the German current account as a per cent of GDP. This throws light on the major risk in our forecast. The rather smooth adjustment paths we are projecting depend upon the willingness of the rest of the world to accept continuing large German capital inflows. any sign of an unwillingness to borrow from or to sell assets to the Germans is likely to see a rapid and, we project, a substantial rise in the deutschemart exchange rate. An unsmooth exchange-rate adjustment of this sort is likely to have deleterious effects on the smooth evolution of the world economy.

Although interest-rate rises do appear to be a moderately effective way of reducing world inflation they do not necessarily help the process of adjustment to a more sustainabel set od current accounts. We have undertaken a set of simulations on our model, GEM, reported in box A, of the implications of a rise in world interest rates. We have allowed exchange rates to erspond to these changes in line with the equations on our model.

The United States

Monetary policy in the US has tightened considerably over the last year, and short rates are now 2-1/2 per cent higher than they were a year ago. This rise has been largely in response to signs of very strong growth in the economy and also to fears of rising inflationary pressures. Preliminary estimates for 1988 as a whole suggest that GDP rose by 3.9 per cent, but that the drought in the mid-west reduced output by 0.3 per cent. First estimates of 5.5 per cent annualised growth in the first quarter are misleading high, because the treatment of farm output clouds the underlying picture. The staticians have naturally assumed that the weather will be somewhat average in 1989, and hence that agricultural output will recover. The seasonally adjusted estimate of output growth for the first quarter therefore includes estimates of the complete year's GDP contribution from both agricultural production and food restocking by the Commodity Credit Corporation. When this statistical artifact is removed, it appears that underlying growth in non-farm output has fallen from the last quarter of 1988 to around 3-1/4 per cent in the first quarter of 1989.

Inflationary pressures are becoming obvious in several parts of the economy. Producer prices rose by 1 per cent in each of January and February, although they slowed somewhat in March. In the first three months of 1989 the PPI was 5.4 per cent higher than it had been in the same period of 1988. The index has risen at an annual rate of 9.1 per cent from the last quarter of 1988. This acceleration in inflation is also visible in the consumer price index, albeit in a more muted way. In the first three months of 1989 the CPI rose 5.4 per cent at an annual rate from the last quarter of 1988, and in March it was 5.1 per cent higher than a year previously. The rise in producer prices has come partly from higher prices for raw material inputs such as metals and foodstuffs, but it has been moderated by the effects on import prices of the rise in the dollar since the second quarter of last year.

There are signs, however, that high and rising interest rates have begun to slow down output growth. Industrial production in March was unchanged since January, although it was up 4.7 per cent from a year earlier. Construction spending in February fell in real terms from January, although this may have been affected by the weather, but the value of construction contracts awarded has been falling recently. All these are reasonable leading indicators of demand, but the effects of high interest rates have not yet passed through to consumer spending or to the labour market. Rises in US interest rates raise the saving rate, but as US consumers are net interest recipients they also raise disposable income, and hence the immediate effects of interest-rate rises on consumption are unclear. Civilian unemployment was 5.0 per cent in March, 0.6 per cent lower than a year earlier and slightly below the February level. Payroll employment growth has been decelerating, but in March the growth rate was still 3.3 per cent over a year previously. All the signs are that the Federal Reserve's tight monetary stance is beginning to have an effect, but that so far there is little evidence of a sharp contraction in demand and output growth.

Our output forecast for the US is given in table 4. We are projecting a slowdown in growth in 1989 and into 1990 of around 1 to 1-1/2 per cent in each year. Domestic demand growth slackens somewhat in our forecast of 1989, and then decelerates sharply in 1990. The February 1989 chapter on the world economy in the Review set out some simulations of the effect of interest-rate rises on the US economy. These suggest that a rise in interest rates such as that experienced since our last forecast should slow output growth by about 1/4 per cent in 1989 and about 1/2 per cent in 1990. These effects, along with the further feed through of earlier rate rises, are the major reason for the slowdown in the economy. On our model the main channel of interest rates is through investment growth, and we are projecting that business investment growth will slow by 10 per cent between 1988 and 1990. The effects of interest rates on consumption are rather small in total and our simulations suggest that the share of consumption in total demand rises when interest rates are higher.

Probably half of the projected fall in output growth in 1989 comes from the effects on both exports and imports of a higher dollar. The rise in the US effective rate over the last year may be partly a consequence of rising US interest rates. Table 5 reports our trade and payments forecast for the US. We are projecting that export growth will slow sharply in 1989, partly because of the slowing of the world economy reflected in the decline in market growth, but largely because of the decline in competitiveness as measured by relative export prices. Our model equation for US exports suggests that the 9 per cent cumulative decline in competitiveness in 1988 and 1989 will eventually reduce exports by 5 per cent below what they would otherwise have been. Our exchange-rate forecast, which is based on the open arbitrage condition linking exchange rates and interest rates, is that the dollar will resume its decline, and that this will exert a positive influence on US exports.

Our model equation for US non-oil imports has recently proved unsatisfactory, and we have replaced it. The research is summarized in box B. We are now forecasting faster import growth in 1989 and 1990 than we were in our last forecast. About half of this comes from worse competitiveness and higher TFE growth than we had anticipated, but the remainder comes from the effects of using our new, probably more pessimistic, equation. This model change, along with the recent appreciation of the dollar, has led us to forecast an only moderate improvement in the US current account deficit in the short term. As table 5 shows, the improvement in the non-factor services balance is approximately offset by the decline in net property incme from abroad that flows from the cumulating effects of past deficits. In the medium term we expect the visible deficit to slowly improve as slower US growth and improving competitiveness once again stimulate US exports and retard imports.

Our forecast of the US budget position is given in table 6. We have not changed our fiscal assumptions since our last forecast, as we still feel that the outcome of budget negotiations between the Bush administration and the Congress will produce a relatively contractionary package. The April budget proposals seem to us to be rather optimistic in that they neither properly allow for the slowdown of the US economy in response to higher interest rates, nor do they take account of the resulting fall in tax revenues. In particular, the short-term prospects have worsened somewhat as higher interest rates feed through to payments on the accumulating stock of debt, but the combination of our projected 40 billion dollar federal package in 1989/90 and the rising state and local government budget surpluses leads to a clear tightening of fiscal stance in the forecast period. This can best be judged by the contraction of the overall public sector deficit rather than by the path of the more widely noticed federal deficit. The public sector deficit is projected to fall as a per cent of GDP, implying a relatively tight fiscal stance in a period when GDP is probably growing well below its trend rate. Even with our package, however, there is little change of the deficit achieving Gramm Rudmann guidelines. Our February Review contained a set of simulations analysing the effects of our projected package on the US economy and on the deficit. Although the projected deficit has risen the effects of the package on it are unchanged, and readers shold refer to those simulations for further details.

Japan

Economic growth in Japan in 1988 was not only the fastest among the G7 but it was also the strongest experienced for over a decade. Investment, both in housing and in business, was the mainspring of growth. Housing investment growth has declined from its peak in 1987, but that of business investment has been at its strongest since the early 1970s. This investment performance was accomplished at the same time as consumption grew rapidly, and combined with buoyant government investment and positive stockbuilding, it produced a rate of growth of domestic demand of over 8 per cent. Unemployment fell through the first half of the year, and temporarily stabilised at around 2-1/2 per cent before falling further to 2.3 per cent in January.

Growth in the fourth quarter was slightly below the average for the year, and there are clear signs of emerging capacity constraints and inflationary pressures. The former have been somewhat eased by the expansion of capacity associated with investment, but the inflationary pressures have been either located in the labour market or associated with external developments. The low growth in consumption in the fourth quarter, which appears to have been extended into the early part of this year, reflected in part the self restraint being practised in deference to the late emperor's illness and death. We anticipate that there will be a strong rebound in the second half of 1989. The non-consumption components of domestic demand have continued to grow strongly, as private machinery and business construction orders have continued to rise. However housing starts in January fell, reflecting emerging capacity constraints in the construction industry. We believe that housing investment will as a consequence of these constraints decline in 1989.

The relative weakness of the yen in recent months, along with very strong oil and commodity prices, have been major factors behind emerging inflationary pressures. Import prices rose by over 2 per cent in February and wholesale prices rose by 0.3 per cent from the previous month. Our model equation for wholesale prices contains effects from import prices, from trend unit labour costs and from capacity utilisation. All are exerting positive influences at the moment, and we expect wholesale prices to rise strongly this year. Over the last three years trend unit labour costs have been falling as earnings growth has been relatively low, but we expect this trend to be reversed. Unemployment is exceptionally low, and the job offers to applicants ratio is at its highest level since 1974, and as a consequence monthly earnings of regular workers are currently rising at 6 per cent per annum, well above the 1 per cent growth experienced in the first half of 1987. Average earnigns are forecast to rise strongly in 1989 and 1990, partly because of the rise in inflation, but also because of the tightness of labour markets. Real earnings growth in 1989 and 1990 together, will as a consequence, be slightly higher than the average of 1987 and 1988. Capacity utilisation indicators are now probably higher than at any time since 1973 (although their relationship to actual capacity utilisation may have changed after the severe recession of the early 1980s). These factors will all feed through to unit labour costs, and hence to wholesale prices and we are expecting inflation to rise in Japan, but only to moderate levels.

Table 7 gives our forecast for the Japanese economy. Growth is expected to slow into 1989, but strong investment, along with good export growth, indicate that output growth will probably only fall by 1/2 to 1 per cent in the next year. The strong export growth has been aided both by the weak yen and faster than anticipated market growth, and we expect these effects to continue through 1989. Our trade forecast is given in table 8. Low Japanese wholesale price inflation in the recent past, along with the fall in the yen, has recently improved Japanese import competitiveness, and we are now forecasting that import growth will be significantly lower in 1989 than in 1988.

We are not forecasting that Japanese interest rates will rise in the near future, and this not only sustains domestic demand in Japan relative to the rest of the world, but also produces a faster rate of appreciation fo the yen than we had previously projected. The Japanese authorities are worried about the effects on their economy of a world slowdown, especially one led by the US, and have been cautious about forcing rates up further. This monetary stance is likely to support business investment though 1989. The path of government policy is unclear in the wake of the Recruit scandal and the consequent resignation of the Prime Minister. Tax cuts at the end of 1988 are likely to support consumption growth in 1989 and 1990 as will the change in the tax treatment of interest income, but the introduction of higher indirect taxes will cause a rise in consumer price inflation.

In this forecast compared to our last, the long-run path for the yen is stronger, and this accounts for much of both the forecast of more rapid growth of imports and the faster loss of export market share. These trends have been compounded in the recent past by factors that we have not been able to take account of in our model. The residuals on both our exports and imports equations have shown distinct trends since around 1984. Exports have been growing less rapidly than our equation would suggest, and imports more rapidly (see chart 5). We suspect that this recent break in structure, which would be econometrically very difficult to model, is associated with the growth of Japanese subsidiaries in the smaller east Asian economies. We have projected a continuing decline in Japanese exports relative to our equation, but our judgement is that the imports residual has risen to a new level rather than developed a new trend. These elements of judgement cannot be empirically evaluated until the underlying process has continued for several more years, but they are clearly central to our forecast that in the medium term the Japanese current balance will decline as a per cent of GDP.

Table 8 gives our forecast of the elements of the Japanese current account. A slowly improving visible trade surplus is largely offset by a rising non-factor services deficit. The net income receipts on Japanese assets and liabilities rise only slowly in dollar terms as the appreciating yen raises the foreign currency value of yen denominated IPD debits. Cumulating surpluses, however, make the process of orderly balance of payments adjustment more difficult.

Germany

Since the last Review complete figures for German GNP in 1988 have become available. These show that overall GNP growth was 3.4 per cent. Domestic demand growth was largely as we had forecast, apart from a surge in stockbuilding in the fourth quarter. Net exports remained approximately constant as a proportion of GNP. Exports of goods continued to grow strongly, while import growth accelerated in the latter half of the year. Industrial production grew by 3.5 per cent in 1988, and has continued to grow strongly in the first months of this yar. In the year to January industrial production of investment godds rose by 9.8 per cent, of intermediate goods by 4.9 per cent, while consumer goods fell by 0.4 per cent. The GNP figures reveal the counterpart to this in strong growth of investment and exports.

Survey evidence suggests that business confidence remains high, although the index constructed by the IFO in Munich has declined slightly in January and February from its peak in December. The IFO survey also indicates a steay rise in capacity utilisation through 1988. The labour market shows some signs of tightening, with unfilled vacancies up from 174,000 in February 1988 to 216,000 in February this year, and the unemployment rate down from 8.8 per cent of 7.9 per cent over the same period. There has been little effect on manufacturing earnings, however, which continued to grow at around 4 per cent per annum.

There has been a marked tightening of German monetary policy in the last six months. The latest move was a surprise increase in short-term interest rates of 1/2 a percentage point at the end of April. This appears to have been prompted by Bundesbank fears that recent strong growth may result in higher inflation. In the year to January wholesale prices rose by 5.1 per cent, up from 3.5 per cent in the year to December. Consumer price inflation is estimated to have reached 3 per cent in April, although this partly reflects an increase in consumption taxes at the start of the year.

Price pressures have not been helped by the recent weakness in the deutschemark. The IMF effective exchange-rate index shows that the currency was 3.5 per cent weaker in March 1989 compared with the fourth quarte of 1987. This is not an especially large fall, but is a reversal of the deflationary impetus provided by an appreciating currency between 1985 and 1987. The recent rise in interest rates can be seen as the Bundesbank providing a strong signal that they wish to maintain downward pressure on inflation and keep the exchange rate strong. The weakness of the deutschemark results partly from the strength of the US dollar. This effect has been compounded by the withholding tax introduced in Germany at the end of 1987 whic appears to have generated substantial capital flight. The government has announced that this tax will be abolished in July. This should provide further assistance to the German authorities in their attempt to support the currency and hence reduce inflationary pressure. However, we feel that the weakness of the deutschemark in the last year has also been influenced by the increasing probability of the formation of a European monetary union. A union will imply that the deutschemark will be tied irrevocably to weaker currencies, therefore its value in the long run will be lower than it would otherwise have been. If current interest rates have not adjusted to reflect the implied lower long-run value then the current exchange rate will also be weaker.

Table 9 presents our forecast for GNP. Consumption growth is expected to slow in response to indirect taxes introduced at the beginning of the year, and last year's investment boom is also expected to moderate. As a result domestic demand is expected to grow by only 2.1 per cent. It should be another good year for exporters, as they continue to benefit from comparative deutschemark weakness, and from strong growth in external markets. Our forecast of muted growth in domestic demand means that imports do not grow as quickly as in 1988, so that net exports make a positive contribution to GNP growth of 2.8 per cent.

For 1990 we expect that domestic demand growth will increase to 2.8 per cent largely because of the stimulus to consumption provided by the final stage of the 1986-90 tax reforms. This should cut taxes by around 20 billion deutschemarks at the start of the year. Inflation should also edge down. A number of forces are at work here: firstly we expect the deutschemark to appreciate along the path suggested by the interest-parity condition; secondly commodity price inflation is expected to moderate; thirdly the blip in inflation associated with the consumption taxes introduced at the start of this year will have worked its way out. As inflationary fear recede we anticipate that the Bundesbank will be willing to reduce interest rates. This will provide a further stimulus to economic activity. Our model suggests that the effects of an interest-rate change on GNP in Germany are modest, and that the growth path of the economy will not be greatly raised. (See Appendix to world economy chapter, National Institute Economic Review, no. 127).

Our medium-term projections suggest that growth into the 1990s will remain at around 3 per cent per annum. Domestic demand growth is expected to strengthen as inflation settles at 1.6 per cent per annum. German GNP growth is also supported by continued growth abroad, although net exports are expected to decline slightly as a proportion of GNP, as the growth of imports outstrips that of exports. This also contributes to a gradual reduction in the size of the current account surplus in relation to GNP. There remains a risk that currency adjustments will be more violent then the interest parity path we are projecting, and thus would tend to endanger the optimistic prospects shown here.

France

French GDP continued to grow strongly in the fourth quarter of 1988, leading to growth of 3.3 per cent for the year as a whole. Domestic demand grew by 3.5 per cent, led by investment and stockbuilding. There was strong export growth of 7.4 per cent as market share increased, but imports also grew strongly, leading to a small decline in net exports. For the year as a whole industrial production rose by 4.5 per cent; this growth has continued into 1989 with an annual increase of 5.2 per cent recorded in January.

Business survey evidence collected by INSEE indicates a bullish assessment of prospects for the industrial sector, which was maintained in the most recent survey (February 1989). Capacity utilisation is estimated to have risen throughout 1988 and increased further in January 1989. Despite the strong growth rate, labour market indicators do not reveal substantial pressures. Seasonally adjusted vacancies stood at 73,000 in February, up from 61,000 a year before, while unemployment had declined from 10.2 per cent to 10.0 per cent. Hourly wage-rates had increased by only 3.1 per cent in the year to the fourth quarter of 1988.

Consumer price inflation edged up through 1988, reaching 3.4 per cent in January and February 1989, compared with 2.5 per cent a year earlier. Domestically generated inflation has remained moderate, but prices of imported goods have risen as a consequence of the weakness of currencies in the EMS block. Within the EMS the franc appreciated from a low of 2.93 deutschemark/franc in the fourth quarter to 2.95 deutschemark/franc at the end of April. This reflects more French franc strength than is at first apparent because German short rates have risen by 3/4 per cent more than those in France over the same period.

We expect GDP growth in 1989 to slow slightly to 2.6 per cent. Most of this slowdown can be attributed to a reduction in stockbuilding which was especially buovant in 1988. Slower growth in domestic demand should result in a lower increase in imports, while exports should continue to grow strongly in response to buoyant external demand. Consequently net exports are expected to make a positive contribution to GDP growth. We anticipate that inflation will remain at around 3.5 per cent. As in Germany the main stimulus to inflation lies in the weakness of the EMS block, and in higher commodity prices than last year.

The most important policy assumption we have made about France concerns the course of monetary policy. This is largely dictated by our views on the development of the EMS. Theory suggests that French interest rates should tend to converge towards German rates as the credibility of the EMS increases and hence currency risks decrease.

Table 2 shows that the interest-rate differential has indeed decreased over the last year, as the French government have made a strong commitment to maintaining EMS parities. The inflation differential has also decreased, but not sufficiently to enable French exporters to maintain their competitiveness. We therefore anticipate that the franc will eventually need to be devalued against the deutschemark, but do not expect such devaluation to be imminent nor to be more than 2-3 per cent.

Our exchange-rate assumption means that medium-term prospects for France are similar to those of Germany. GDP is expected to grow by just under 3 per cent per annum, mainly due to strong domestic demand supported by low inflation and steady external demand. France, along with the other EEC countries, may also derive additional benefits from extra activity generated by the run-up to 1992. The size of such potential benefits remains difficult to assess, and we have not made any special provision for them in our medium-term forecast.

Italy

Information on Italian GDP in the fourth quarter of 1988 is not yet available, but the third quarter data appears to be consistent with GDP growth of 3.7 per cent. We estimate that domestic demand grew by around 3.3 per cent, led by consumption growth of 4 per cent. Net exports made a positive contribution to GNP growth as the rate of export growth accelerated, and the import growth rate decelerated compared with 1987. Even so Italian exports appear to have experienced a loss of market share.

Consumer price inflation for 1988 stood at 5 per cent, and increased in January to 5.5 per cent. Italian inflation remains the highest in the EMS block, and this could eventually lead to pressure for a realignment. At present however the advantages of the EMS as an anti-inflationary discipline appear to outweigh the disadvantages of a rising real exchange rate. It is therefore unlikely that there will be a realignment in the next six months. Indeed the lira strengthened against the deutschemark in the first quarter and has held steady through April at around 735 lira/deutschemark. This may in part be because in March the Italian discount rate was raised from 12.5 per cent to 13.5 per cent opening up an even higher interest differential against the deutschemark. The authorities' stated aim was to encourage domestic saving and hence reduce the current account deficit. The interest-rate increase was however expected to add 700 billion lire to the public sector deficit. This continues to be major economic and political problem, with some speculation that the current coalition government will collapse, possibly following the European elections in June.

Table 13 presents our forecast for Italy. In 1989 we expect GDP growth to slow to 3 per cent. Consumption and investment are both expected to grow more slowly than in 1988. The recent rise in interest rates contributes to this. External demand should remain strong, however, and net exports are expected to again contribute towards GNP growth. Inflation is expected to average 5.2 per cent for the year. An increase in imported inflation is largely offset by our forecast of a decline in trend unit labour costs.

Our medium-term forecast is shaped by our assumption that Italy remains committed to the EMS, with a gradual depreciation against the deutschemark to alleviate loss of competitiveness. The interest-rate differential is not expected to decrease by much, as we anticipate that there will be a continued risk-premium due to concern over Italy's large public debt. The continued success of the EMS strategy will require further reductions in the growth-rate of Italian earnings. If this materialises inflation could stabilise at 3-4 per cent, enabling medium-term GNP growth of around 3.5 per cent per annum.

Canada

Canadian real GDP rose by 4.5 per cent in 1988, probably the second strongest performance in the G7 and also the highest growth experienced since 1984. The major factor behind this growth, as in the other major economies, has been a rise in domestic demand of over 5 per cent. Business investment growth has been particularly strong, especially in the middle of the year, and has been concentrated in machinery and equipment rather than non-residential construction, suggesting that a considerable amount of retooling and replacement has been undertaken. Consumption growth has slackened somewhat since 1987, although the saving ratio has continued to fall. The combination of strong domestic demand and an appreciation of the currency in particular against the US dollar has led to very strong import growth and a decline in both the visible balance and especially in the balance on investment income.

In the fourth quarter of 1988 GDP growth slowed somewhat as the rise in the currency through the year produced stronger import growth and falls in the level of exports. Consumer expenditure rose more rapidly in the last quarter than earlier in the year as real personal disposable income rose at an annual rate of over 6 per cent. This was a consequence of income tax deductions temporarily falling as part of the current tax reform programme. Residential construction picked up strongly in the last quarter, with housing starts in the early part of 1989 at their highest sustained level since mid-1987. The strong growth in the fourth quarter was accompanied by a decline in the trade balance, but the current account deteriorated more sharply because of one-time payments of large dividends to non-residents worth 3.2 billion Canadian dollars.

There are clear signs of inflationary pressure building up in the Canadian economy. The GDP deflator rose at an annual rate of 6.3 per cent in the fourth quarter, and in January and February consumer prices were rising at an annual rate of 4.8 per cent. Employment has been growing rapidly, with 141,000 jobs created over the November-February period. However the unemployment rate has stayed relatively stable at around 7.6 per cent. The Canadian monetary authorities have perceived these inflationary pressures for some time, and began to tighten monetary policy in early 1988, well before the rest of the G7, and the Governor of the Bank of Canada, Mr Crow, has stressed the importance of the anti-inflationary stance. This has been in strong contrast to the behaviour of the federal fiscal authorities who have been running a government deficit of over 5 per cent of GDP a year. Despite IMF warnings, there were no significant attempts at budget tightening before the November 1988 election which centred on the issue of free trade with the US. The prospects for genuine reductions in expenditure after the April 1989 budget remain poor, but fiscal stance is expected to tighten somewhat in the next year.

The combination of tighter fiscal and monetary policy colours our forecast which is given in table 14. We project that in response to higher interest rates business investment growth will slow down and that housing investment will fall in 1989. Lower government spending growth will also reduce growth, although some of the effects are likely to be absorbed by a fall in the saving ratio as consumers take time to adapt to lower growth in their real incomes. We also project that rising wholesale prices and unit labour costs will lead to a loss of competitiveness and a decline in the real trade balance. As a consequence of all these factors we expect the economy to slow down through 1989 into 1990 before recovering some of its buoyancy thereafter.

Commodity prices

Nominal commodity price rises have moderated somewhat recently, but price levels, as measured by the Institute's indices plotted in chart 6, remain high. Metals prices have been particularly high recently, largely because of the strength of demand in the industrial world. However, as we have discussed in the August 1988 edition of the Review, stock levels have been historically low and hence demand increases have fed through rapidly to prices. Even tin prices have risen in recent months. The effects of increasing demand on metals prices have been exacerbated by high levels of purchases China. Chinese industrial production has been rising strongly recently, raising demand, but stockholding by centrally planned enterprises, both of metals and of fibres such as wool and cotton, have risen recently. This is a common phenomenon in centrally planned economies when financial constraints are relaxed, and its implications are discussed in Bowles (1989).

Food prices have eased recently as US weather prospects improved and as southern hemisphere grain crops began to be brought in. We do not anticipate any significant rises in food prices in the short run, but as chart 7b shows we do expect real food prices, at least on the free market, to rise somewhat as levels of agricultural support in the EC, especially for grains, decline and hence reduce surplus production and excess dumping onto world markets.

Oil prices have been very strong recently, largely as a result of strong demand and declining stock levels. There is some evidence that these oil price rises hae been at least partly associated with temporary supply difficulties, both in the North Sea and in Alaska, each of which have probably reduced the free world flow supply by one or more per cent. However, in a market for a homogenous and easily transportable fluid commodity dominated by stocks, the importance of these supply developments should not be exaggerated. As real activity slows in the next two years we expect oil prices to weaken somewhat. Chart 7a gives our forecast for real oil prices, and table 15 gives our complete commodity forecast.

World trade and the balance of payments

World trade grew strongly in 1988 as a result of economic growth in the industrialized world. The UN series for trade in all goods suggests an increase of around 6 per cent, while the main manufacturing countries' trade in manufactures grew by around 11 per cent. Import volume growth was especially strong in Canada, Japan, the UK and the Far East, and appears to have been weakest for OPEC countries, as oil prices fell.

The growth in world trade is likely to decline both this year and next as tighter monetary and fiscal policies restrain the growth of demand in industrialised countries. Our forecast suggests growth in trade of manufactures will be around 6 per cent this year, and under 5 per cent in 1990, and thereafter trade will grow at around 5-5.5 per cent per annum. The stronger oil price should enable an increase in OPEC's imports, and other developing countries should benefit from further increases in real commodity prices.

Table 16 shows our forecast of world trade percentage shares. The US is expected to increase its share of trade as the dollar weakens further, and the Asian economies are also likely to increase their share. The main losers are Japan, the non-Asian LDCs and Canada. Table 17 shows the implications for world current balances. We anticipate that the main current account imbalances can be reduced, but that progress over the next couple of years will be slow.

APPENDIX

Fundamental equilibrium exchange rates for the G3 by R. J. Barrell and Simon Wren-Lewis

The analysis of 'fundamentals' in discussions of the exchange rate is a popular passtime amongst economists. The concept of Fundamental Equilibrium Exchange Rates (FEERs) was developed by Williamson in 1983. He defines the FEER as the real exchange-rate path that produces an equilibrium, or trend, current account of the balance of payments that is exactly matched by 'structural' capital flows. In charts A1 to A3 we provide estimates of FEERs for the G3 over the period 1971-88 using the characteristics embedded in our Global Econometric Model, and compare them with real exchange-rate paths over that period.

It is possible to derive relationships between the actual and the equilibrium current accounts and the real exchange rate. The actual current account depends upon the current and past values of the factors affecting trade in goods and services, such as the real exchange rate, the level of world income and the level of domestic income, as well as real commodity prices. It also depends upon the income flows derived from existing stocks of overseas assets and liabilities. These stocks are largely the result of the past path of the current account, and hence the current account depends upon its own history. The current account may differ from its equilibrium value not only because of the dynamic effects of past events but also either because the real exchange rate is not at the FEER, or because one or more of the factors determining the current account are not at their trend, or equilibrium, values.

In our analysis we do not presume that the only equilibrium trend current account is a balanced current account. Structural capital flows may exist for long periods of time, and they will influence the sustainable exchange rate. Structural capital flows are those that would take place if there were no speculative flows. As such they are rather hard to observe. If there is a change in portfolio preference toward holding a higher proportion of total wealth as overseas assets, or if the rate of return on real domestic assets falls relative to overseas returns, the asset holders will not wish to maintain their pre-shift allocations. The shift will generate a sequence of capital flows that will be maintained until a new portfolio equilibrium is attained, and this may take many years. Our estimates of these structural capital flows, which are contained in table AP1, are schematic, but we do not think they are unrealistic, and they do allow us to produce an estimate of the equilibrium real exchange rate.

The body of our analysis in Barrell and Wren-Lewis (1989) considers the trade accounts of the G7 countries, using econometric estimates embodied in the National Institute's world model--GEM. We have estimates for income and price elasticities for trade in goods and in services, and we also have relationships determining the influence of world prices on each country's export and import prices. These estimates, along with trends derived from the data for the period 1970-88, are used to produce estimates of the relationship between the balance of payments on goods and services and the real exchange rate. We then consider appropriate assumptions for determining the level of internal balance, which will be affected by the real exchange rate. Changes in the real exchange rate or in real commodity prices, may affect the wage bargaining process, and this will in turn affect the equilibrium level of output both in each economy separately and in the world economy. When we look at the world economy as a whole we take into account interaction between world output, trade, oil and commodity prices.

Before we can produce estimates of the FEER we have to analyse the influence of income flows from overseas assets and liabilities on the trend current account. These flows are significant, and the home currency value of receipts from abroad will be influenced by changes in the exchange rate. These income flow estimates are combined with the factors affecting the trade account to produce an external balance relationship between the trend current account and the real exchange rate.

The core results are reproduced in charts A1 to A3. The real exchange rate R, is 1 in 1980, and a decline in the rate, on our definition of world prices compared to domestic prices, implies an appreciation. Our sensitivity analysis, concerning alternative assumptions about exogenous variables or possible structural shifts in trade relationships, strongly suggests a wide range of possible values for FEERs. The model suggests that the US FEER represents a devaluation of the dollar compared to current (89Q1) levels ranging from 5-15 per cent. However, an even larger range for the dollar may be prudent partly because of the unusually high, and unexplained, level of US imports over the last three years. The FEER for the yen represents an appreciation of 5-15 per cent, with the major uncertainty here concerning the structural outflow of Japanese capital. The FEER for the DM is also 7-17 per cent above current levels. Given our present state of knowledge about the determinants of trade flows, internal balance and structural capital flows, we think it unlikely that empirical analysis would enable these ranges to be narrowed. (Indeed, it would be quite easy to combine quite reasonable assumptions to take FEERs outside the ranges presented here.)