The world economy.

Barrell, R.J. ; Gurney, Andrew

CHAPTER II. THE WORLD ECONOMY(1)

The overall outlook Developments since our last forecast three months ago have been dominated by fears of accelerating inflation and the policy responses to them. Domestic price developments have remained moderate, but both oil prices and the prices of a number of industrial materials, especially metals, have risen strongly in the last few months. Oil prices have risen from their November low of $12 per barrel to around $16 per barrel in recent weeks, and metals prices, as measured by an index constructed by the Institute, rose by 25 per cent between September and December, although prices have been a little easier in recent weeks. Both price developments are good leading indicators of rising demand. A rising oil market gives the OPEC cartel more of a chance to produce a semblance of unity, and metals stocks have been at low levels recently, and hence increases in demand have been feeding rapidly through to prices over the last 18 months.(2)

Interest rates have risen sharply in the last two months as the monetary authorities, especially in the US, but also in Germany, France and the UK have responded to the emerging signs of accelerating inflation. Short rates are now 1 per cent higher in the US than they were in the third quarter of 1988, and are 1.5 per cent higher than we had anticipated they would be in November. The rise in long rates has been more muted, and they are still below their 1988 peak, either because short rates are not expected to stay so high for long or, more likely, because inflationary expectations of participants in the bond markets have been marked down in response to a tighter than anticipated response by the Federal Reserve to rising inflationary pressures. German and French short rates have also risen over the last three months, and are now around 3/4 per cent higher than they were in the third quarter of 1988.

Our forecast, which is summarised, suggests that, as anticipated by the monetary authorities, inflation will rise into this year, but that their quick response will, if it is maintained, reduce inflationary pressures into 1990 and beyond. In recent months consumer price inflation has been running at an annual rate in excess of 2 per cent in both Germany and Japan, well above the average for the year, and US consumer price inflation peaked at around 4.8 per cent in the last quarter of 1988.

Strong output growth has continued into the third and fourth quarters of 1988, and we have revised up our estimates of growth in 1988 both for the US, where consumers expenditure has been stronger than we had anticipated, and for Japan where housing investment does not appear to have been so constrained by capacity as we had been led to believe. Although GDP growth has been strong recently we have revised down our forecasts of growth through 1989 into 1990 for all the major economies, although the strong performance at the end of 1988 raises the year-on-year growth rate for the US, Germany and France. The slowdown in activity is most marked in the US, where very tight monetary conditions have a considerable effect on activity. (Some simulations are reported below that detail the effects of interest rates on the US economy.) The US authorities appear to be more worried about inflation than are those of France, Germany, and especially Japan, and we project that they will keep their rates at high levels until they are convinced that inflation will fall. Interest rates in Europe and Japan are assumed to begin to fall earlier. This pattern of expectations may be the same as those held by the markets. An equal rise in interest rates in Europe and North America has led to a strengthening of the dollar, which would only be appropriate if the markets also believed that rates would stay higher for longer in the US.

We assume that a positive interest-rate differential between say the US and Germany, such as we observe at present, will be associated with an annual rate of decline in the dollar relative to the deutschemark of around the same percentage with perhaps some adjustment for a premium associated with risk or portfolio saturation. Although the open abitrage condition has not always held ex-post, it may still be a useful tool for forecasting, in that the periods when it has not held in the past may for example have been associated with temporary flow disturbances associated with the arrival of news, producing divergences from portfolio stock equilibrium.

Recent changes in exchange rates and interest rates have not been particularly conducive to a rapid and orderly adjustment of current account imbalances toward more sustainable levels. Current account equilibrium is achieved when the current account for each country matches the net level of long-run sustainable capital outflows or inflows. Low levels of manufacturing profitability in Germany lead to large scale capital outflows; high levels of savings in Japan, along with a desire to increase net holdings of financial assets that is not satisfied by the rate of issue of government debt, lead to structural capital outflows that are currently in excess of the Japanese current account surplus. However, the current pattern of imbalances is also strongly influenced by uneven patterns of demand development, and misaligned real exchange rates have exacerbated these problems.

Interest rate increases appear to be a moderately effective way of reducing world inflation, but they do not necessarily help the world economy adjust to a more sustainable pattern of current accounts.

Prospects for world trade remain good, with total world trade, which includes trade in food and raw materials as well as energy products and manufactures, expected to rise by almost 6 per cent in 1989, after an estimated rise of 7.5 per cent in 1988. We estimate that exports of manufactures by the major OECD countries have risen by 8.5 per cent in 1988, and that they will rise by 7 per cent in 1989, but much of this rise has been associated with the increase in US--Canadian trade. Our estimates suggest that world trade, when weighted by say either Japanese or German export markets has only been growing by around 6.5 per cent in 1988, and we expect the growth to slow down into 1989. Looking further ahead the contractionary monetary stance in the US along with our projected fall in the dollar should, we forecast, lead to a fall in US non-oil imports, which reduces the growth of world trade in 1990. We see no reason to be pessimistic about the prospects for longer term world trade and output growth. The growth of world trade in manufactures should return to its estimated trend growth of around 5 1/2 per cent in the longer run.

The United States Strong growth in the US during the last year has led to a considerably tighter monetary stance than we had previously anticipated. First estimates suggest that US real GDP grew at 3.8 per cent in 1988. However, this may well be an underestimate of the underlying strength of the economy because the drought over much of the US reduced annualised growth in the third quarter by 0.5 per cent, and in the fourth quarter by 1.1 per cent. Provisional figures for the third quarter and preliminary estimates for the fourth quarter suggest that non-farm output is growing at around 3 per cent at an annualised rate. However, this apparent slowing down in the growth rate from the first half of 1988 partly reflects the influence of the strong dollar which has reduced the contribution of the overseas sector to the economy, and domestic demand has been growing more rapidly than earlier in the year.

Inflationary pressures are becoming apparent, with consumer price inflation having risen from 3.7 per cent in 1987 to 4.1 per cent in 1988. The consumer price index was rising at around 4.1 per cent in December, despite the effects of much lower fuel and oil prices. Services prices, which more fully reflect domestic costs, have been rising at almost 5 per cent. Food prices for consumers have been rising at over 5 per cent per annum recently because of the effect of the drought on the price of many of the raw products used in the processing industries. Lower than anticipated rises in finished foods prices have helped hold down the rate of inflation in the last few months, but other indicators suggest that inflationary pressures are still building up. Industrial production rose 4.7 per cent in December when compared to the same month of the previous year, and capacity utilisation in manufacturing is now at 84.2 per cent, the highest level reached since November 1979. Unemployment has been falling and seems to have settled at around the 5.3 per cent level observed in December, down from 5.8 per cent a year previously. But employment growth has been remarkably strong, with numbers in employment rising by around 350,000 a month for three months since November, and in January the non-agricultural workforce was more than 3.5 per cent higher than a year previously. This almost 1 per cent rise in the workforce in three months is a strong, albeit slightly lagging, indicator of the rate of increase in demand in the US. The US economy begins to look as if it is working at or near to full capacity, and a considerable rise in inflation might be expected if there were no policy response.

The Federal Reserve has responded to these increasing pressures by raising interest rates significantly. The Federal Funds rate has risen by almost a percentage point since the end of November, and at the end of January it stood at 9.1 per cent, 2.5 per cent higher than it was a year ago. Our model suggests that a monetary tightening of this magnitude will have a considerable and rapid effect on the US economy, specifically through consumers' expenditure and residential construction. Box B sets out the details of a simulation which allows US rates to rise with interest rates in the rest of the world unchanged. We attempt to separate out the effects of interest rates as they work directly on the economy from the effects that operate through higher exchange rates.

We are anticipating a considerable slowdown in economic activity through 1989 into 1990, with the growth rate hardly rising above 1.5 per cent at an annual rate for two years. This slowdown is entirely domestically generated, with investment growth slowing to almost zero and investment in housing declining once again in both 1989 and 1990. Consumption growth also slows down, but so do both earnings and employment growth, and as a consequence in our forecast the savings ratio does not rise. The slowdown in domestic demand is forecast to reduce stockbuilding in the longer run, but the 1989 stockbuilding figures will be difficult to interpret. The 1988 drought has considerably depleted grain stocks, and wheat and especially corn will not flow into the silos to any significant extent until the third quarter. However, the National Accounts statisticians have to seasonally adjust the stockbuilding data, and hence they will have to notionally allocate some of this grain to the first two quarters. It will therefore be only after we have data for the third quarter that we will have a clear picture of the development of stockbuilding for the early quarters of 1989.

There are two further, countervailing, effects on GNP growth during 1989 and 1990. Firstly we expect the overseas sector to continue to contribute to GNP growth, but secondly we expect the public sector to exert a contractionary influence on the US economy. We are still assuming that the Bush administration will introduce a fiscal package at the end of 1989 that will go a long way to meeting the Gramm-Rudman-Hollings guidelines. Our projected package was spelt out in our last Review chapter. In brief, we are anticipating that income tax will rise by $20 billion, and that expenditure will be cut by $25 billion. However, this package will not be enough to meet the agreed guidelines by 1990, largely because of the increased interest cost of US government debt. US government bonds are on average shorter term than UK bonds, and hence increases in current market interest rates feed through relatively quickly to the average interest rate paid on debt on our model of the US public sector.

The fiscal stance in the US should be judged by the wider public sector deficit, which includes state and local government as well as the public sector social security fund. Despite our predictions that output will grow more slowly than trend, we are projecting that the public sector deficit will decline from 2.3 per cent of GNP in 1987 to just under 1.5 per cent in 1990. This cyclically adjusted contraction is a major factor behind the projected slowdown in economic activity.

Export growth in the US will be held back by the recent strength of the dollar, although we are projecting that the exchange rate will decline gradually from its current heights. US exports have been growing very rapidly recently, largely as a lagged response to gains in US competitiveness in 1985 and 1986, and export orders remain at historically high levels. However, export prospects for non-manufactures, which make up some 20 per cent of US export trade, have been damaged by the decline in farm output in 1989. The drought is not, however, expected to have much impact on imports, and we are anticipating that the combination of the lagged effects of improved competitiveness and the considerable slowdown in domestic demand will reduce or even reverse non-oil import growth in 1989 and 1990. We are expecting services exports (travel, tourism, etc) to continue to rise rapidly in response to the falling dollar.(3) By the early 1990s the US is likely to make net payments on its IPD account. This development lags well behind its reputed move in 1985 into net debtor status in terms of overseas assets and liabilities. The lag arises largely because direct investment assets are sometimes measured on an historic cost basis, and US direct investment assets are generally much older than US direct investment liabilities, and hence the former are considerably undervalued relative to the latter. According to our analysis of the flows, which may give a better picture of the true asset values, the cumulating US current account deficits will eventually make the US a true net debtor, but not until some time in the 1990s.

Japan Output growth in Japan in the third quarter of 1988 reached almost 9 per cent at an annual rate, a marked recovery from the decline in the second quarter. Consumption rose by 1.4 per cent, and was 5.5 per cent above the level achieved a year previously. Investment in plant and equipment continued to grow at an annual rate in excess of 16 per cent per annum, and underlying capacity growth in the Japanese economy must have risen significantly in the last year. The growth in business investment in buildings, especially in mining and manufacturing, suggested that resources were being diverted away from residential construction. However, in the third quarter of 1988 residential construction growth at 6 per cent (25 per cent per annum) was remarkably strong, and was a major factor behind high growth in private sector demand.

Overall domestic demand grew more slowly, with bottlenecks in the construction industry continuing to plague public fixed capital formation, which fell by almost 2 per cent in the third quarter. Government consumption also grew slowly, and in combination with a decline in stocks the resulting annualised growth in domestic demand was only 5.5 per cent. The overseas sector produced its first positive, albeit small, contribution to output growth since the last quarter of 1986, largely because exports grew in response to the combination of a weaker yen in the middle of 1988 and declining unit labour costs which were partly the result of higher productivity that has resulted from new investment in plant and machinery. These unit labour cost declines, are, however, likely to be temporary as overtime hours worked decline from their peaks in the first half of the year, and as wages and especially the end year bonuses rise in response to increasing labour market pressures.

There are clear signs of high pressure developing in the labour market. The job offers to applicants ratio reached 1.09 in October, its highest level since 1973. The ratio has been around this level since June, and the unemployment level also appears to have stabilised since then at around 2.5 per cent. The growth of wage costs also picked up in 1988 with nominal wages rising at around 4 to 4.5 per cent in the second half of the year compared to under 2 per cent in 1987. Bonuses, which make up around one quarter of Japanese earnings, are expected to have risen by 6 per cent in 1988, compared to 2 per cent in 1987, although as they are generally paid in the fourth quarter not all the data have yet been collated. Early evidence from the 1989 Spring Wage Offensive suggests that wage increases in 1989 will be almost 1 per cent higher than in 1988.

These rises should be seen in the light of rather moderate increases in both consumer and producer prices over the year to November. Consumer prices rose only 1.1 per cent between November 1987 and November 1988, although the annual rate of increase had risen to 2.8 per cent in the latest three months. Prospects for consumer price inflation are, however, considerably worse than they might at first appear. The prices of fuel and electricity and imported durables have been exerting negative influences recently; and both these effects are likely to be reversed, or at least reduced in the coming months because of the recent rise in the world oil price and the relative weakness of the yen. The anticipated switch from direct to indirect taxation in 1989 is also likely to lead to higher consumer prices.

We are forecasting that Japanese output growth will slow down into 1989, in part because higher interest rates will reduce investment, but largely because we are anticipating a considerable slackening of the rate of growth of real government expenditure compared to 1987 and 1988. This is in line with the government budget announced on January 24, where nominal government expenditure was forecast to rise by 6.6 per cent in fiscal 1989. The switch from direct to indirect taxes, which we have assumed will take place in our forecast, will expand demand by around 0.5 per cent in the long run, but this will take time as consumers react to their higher disposable income. The net cut in taxation is a reaction to the very high direct tax revenue received in 1988 that resulted from the buoyancy of the economy.

If US fiscal and monetary policy remain as restrictive as we are anticipating, then Japanese growth can be expected to slow further into 1990 and 1991. We do not expect this to be ameliorated by faster growth in government expenditure. The combined effects of the forecast rise in the yen and the sluggish growth in export markets in 1989 and 1990 produce from the overseas sector, for the forseeable future, a negative contribution of around 0.5 per cent to output growth. In the longer run we expect the Japanese economy to return to its trend rate of output growth, which we estimate to be between 4 and 4.5 per cent.

We are projecting that the Japanese current account surplus will slowly decline from its peak in 1987, and that there will be a gradual loss of export market share as a consequence of declining competitiveness. The decline in world investment growth will also cut the demand for Japanese exports more than those of, say, the US or the UK as they tend to be relatively less specialised in this field than the Japanese. We expect the surplus to decline as a proportion of GDP, but much of that decline comes from invisibles. Our research on invisibles suggests that they respond strongly to price so that as the yen appreciates Japanese service imports will rise much more rapidly than service exports. In combination with currently planned increases in foreign aid, the declining services balance outweighs the positive effect on net invisible of the accummulating current account surpluses which add to interest profits and dividends. The growth of this property income is also held back by the appreciating yen.

Our rather optimistic forecast for the Japanese current account requires either a reduction in the scale of long-term private capital outflows or a rise in shortterm capital inflows. Long-term private capital outflows have exceeded the current account surplus since 1984, and in 1987 they were $137 billion, or $50 billion bigger than the current account surplus. This outflow had to be financed by a combination of short-term capital inflows of around $24 billion and by reserve accumulation. This desire on the part of Japanese investors to hold assets overseas puts downward pressure on the yen, and if it continues this could prevent the current account surplus declining. We are assuming, however, that more domestic uses for Japanese funds will be found in the future than has been the case in the past.

Germany

The latest indicators for Germany suggest that GNP growth in 1988 was around 3.5 per cent, as predicted in our November Review. This is the highest rate of GNP growth achieved in the 1980s: aside from this GNP growth has only once, in 1984, exceeded 2.5 per cent during the decade. the factors that came together in 1988 to produce rapid growth are all likely to reverse in 1989. Our current forecast suggests that output growth will slow to between 2 and 2.5 per cent in the medium-to long-term.

The reduction in output growth is already apparent in the data available for 1988. In the year to the first quarter German GNP grew by 4.4 per cent, but over the year to the third quarter by only 2.5 per cent. The period of greatest growth was in the latter half of 1987 and in the first quarter of 1988. Consumption grew by 4.6 per cent in the year to the first quarter, and investment by 10.8 per cent, but these growth rates have not been sustained. By the third quarter consumption had grown by only 1.8 per cent over the previous year and investment by 3.2 per cent. This deceleration in domestic demand was partly compensated by higher growth in exports, whose annual growth was 8.5 per cent in both the second and third quarters. This can be attributed partly to the weakness of the deutschemark through much of 1988, but mainly to the general strength of the world economy. For Germany to repeat its 1988 performance would require a renewed stimulus to both domestic and external demand. This seems unlikely to materialise as policy at home and abroad has moved towards restraining any inflationary pressures arising from last year's growth in demand.

In the last three months the most significant developments for the German economy have been a weakening in the deutschemark, and a tightening of monetary policy. The monetary tightening occurred largely in response to the weaker exchange rate. It appeared that it was enacted by the Bundesbank in the face of some resistance from the finance minister, Dr Stoltenberg. The Bundesbank's concern was that a weak deutschemark would fuel inflation, whilst Dr Stoltenberg appeared to be more concerned that monetary tightening would dampen domestic demand both within Germany itself and within the wider EMS block. Further divergence of opinion is evident in the Bundesbank's belief that the proposed introduction of a withholding tax on interest payments has led to a greater outflow of capital, and hence caused the weakness of the exchange rate. This view seems plausible, but is probably only a partial explanation: the deutschemark has also weakened because of rather higher interest rates available elsewhere, most notably in the United Kingdom and the United States. In our forecast it is assumed that there will be no further monetary tightening either within Germany or elsewhere, and consequently that the deutschemark can be expected to appreciate in line with the interest-rate differential. This appreciation should enable inflation to be held at around 2 to 2.5 per cent per annum, and in due course permit a relaxation of monetary policy.

The tightening of monetary policy that has already occurred is due to be complemented by a tightening of fiscal policy in 1989. In addition to the introduction of interest withholding tax, the government has raised indirect taxes and reduced subsidies. These measures are intended to counter a projected increase in the budget deficit, and partially offset earlier expansionary measures introduced as part of the 1986-90 tax reform programme. The final stage of this programme is due to be enacted in 1990. This is intended to reduce net taxes by DM19 billion. Hence while fiscal policy is likely to be contractionary in 1989, it is then set to be expansionary in 1990.

The strongest component of GNP growth in 1989 is again likely to be exports. The proposed fiscal tightening will dampen domestic demand, but the traded sector should continue to benefit from the current weakness of the deutschemark, and from the strength of demand in the rest of the world, although both factors are expected to become less favourable during the course of the year. Our projected 7 3/4 per cent growth in exports in 1989 falls to under 2 1/2 per cent growth in 1990, but a recovery in domestic demand as a result of the expected tax cuts should ensure that GNP growth in 1990 is around 2 1/4 per cent, following 2 3/4 per cent in 1989.

Inflation is likely to increase from 1 per cent in 1988 to around 2.5 per cent in the coming year partly as a consequence of the rise in indirect taxes in the first quarter of 1989. These increases alone are expected to raise prices by over 0.5 per cent. The increase this year is also largely attributable to the unwinding of the large deflationary effect of the appreciation of the deutschemark between 1985 and 1987. We are assuming a rather more modest appreciation of the deutschemark in the years ahead, and this can be expected to contribute to low inflation as German firms seek to keep prices low in order to remain competitive.

France The latest estimates suggest that the French economy grew by around 3 per cent in 1988. This is the highest annual figure in the 1980s, the next best being 2.5 per cent in 1982. Investment proved to be the strongest element of demand, aided by steady growth in consumption. In the early part of the year net exports made a positive contribution to GDP growth, but a strengthening in domestic demand in the latter half of the year appears to have resulted in rather higher imports. The current account balance for 1988 as a whole is likely to be around FF25 billion, little changed from 1987. Consumer price inflation for the year as a whole is estimated to have been 2.7 per cent.

The main focus of macroeconomic policy continues to be the need to remain competitive within the constraints of the EMS. The last realignment in EMS parities occurred in January 1987. The recent weakness of the deutschemark has led to all countries in the EMS block gaining competitiveness against the other major trading nations. This has probably prolonged the period for which the current parities have been maintained successfully. We expect the deutschemark to strengthen during 1989. If this occurs it could revive the pressures for realignment of the EMS.

The French authorities are likely to resist such pressures. In December the Governor of the Bank of France was reported to have said that realignment of the EMS was not an option since it `would simply be a way for West Germany to export its inflation. We have already imported many things, but we do not have any intention of importing inflation'. The authorities commitment to maintaining the current parities means that French firms cannot expect to gain competitiveness by means of a devaluation, and hence puts pressure on them to maintain competitiveness through control of their costs.

Through 1987 and the first half of 1988 the differential remained between 4 and 4.5 per cent, but narrowed in the latter half of 1988 to around 3 per cent. This narrowing may be interpreted as signifying increased confidence in the foreign exchange markets that the current parities can and will be maintained. In the main, however, one of the costs of maintaining exchange-rate parities is that the stance of monetary policy is effectively determined by the EMS commitment. This was evident in the middle of January when the French authorities were forced to raise their interest rates in line with the 0.5 percentage point increase in Germany. This action effectively fulfilled the commitment given by M de Larosiere a month earlier.

Fiscal policy is likely to be little changed in 1989. The government's target is that the public sector debt to GDP ratio should be stabilised over the medium term, and they estimate that this will be achieved with a budget deficit of around FF70 billion. The 1989 budget projects a deficit of FF100 billion. The government estimate that this can be achieved with a 2 per cent growth in the volume of public expenditure.

Our forecast suggests that GDP growth will slow to 2.4 per cent in 1989. This is primarily due to a decline in the growth of investment as tighter monetary policy reduces the prospects for demand both at home and abroad. Consumers' expenditure grew less rapidly than real personal disposable income in 1988, but is expected to grow slightly faster in 1989, helping to sustain the growth of domestic in 1989, helping to sustain the growth of domestic demand. Net exports are expected to remain unchanged as a proportion of GDP. The weakness of the EMS block will mean that France remains competitive, and may be able to increase its share of export markets, while a decline in the growth rate of total final expenditure should also lead to a decline in the growth of imports.

Italy Italian GDP is estimated to have grown by 3.5 per cent in 1988. Full data are only available for the first half of the year, but it appears that growth of domestic demand was lower than in 1987, because of lower growth in government spending, and much lower stockbuilding. Consumption growth was also slightly slower, although still exceeding the estimated 1.6 per cent growth of real earnings, but investment expenditure seems to have picked up, in line with experience elsewhere in Europe. Net exports benefited from the weakness of the EMS block and from the strong growth in export markets. Inflation stabilised at around 5 per cent.

Italian policy is shaped by the need to control the growth of the public sector deficit, and by the desire to remain competitive within the framework of the EMS. Public sector debt currently stands at over 90 per cent of GDP and presents a considerable financing burden. The public sector deficit for 1988 is estimated to have been 11.3 per cent of GDP. The proposals for the 1989 Finance Act aim to keep the deficit constant in nominal terms for 1989 at 117.4 trillion lire. The immediacy of the public sector debt problem menas that fiscal policy will have to remain tight in the years ahead.

The stance of monetary policy is primarily determined by the constraints imposed by the EMS. The high level of public sector debt, and expectations that the lira will be devalued in the next EMS realignment have contributed to high real interest rates in Italy. The prospects for maintaining the lira at its current EMS parity depend crucially on the performance of the deutschemark in the year ahead. While the deutschemark remains weak the authorities can retain the current parity in order to enhance the credibility of their anti-inflationary stance. Continued downward pressure on inflation should allow further reductions in interest rates, which will also help to alleviate the financing of public sector debt. If however the deutschemark strengthens as we expect, the pressures for a realignment of the EMS would be rather stronger. Higher inflation in Italy than in France and Germany, combined with the weakness of the Italian current account makes the lira especially vulnerable. Additional problems may arise in 1990 when controls on the movement of foreign exchange are lifted. Monetary policy will have to continue to be tight, with real interest rates likely to be in excess of 7 per cent for most of 1989.

A tighter fiscal policy and continuing tight monetary policy will contribute to a deceleration in the growth of GDP in 1989. Private consumption, investment and government expenditure are all expected to grow more slowly than in 1988. Exports should continue to benefit from the strength of external demand. We assume that the lira will weaken against the deutschemark during the year, without necessarily triggering an EMS realignment. Net exports are expected to remain unchanged as a proportion of GNP. Inflation is expected to stay at around 5 per cent in 1989.

Canada The growth of the Canadian economy during the first nine months of 1988 has exceeded the expectations of most forecasters. Real GDP grew by over 4 per cent at an annualised rate, and early indicators suggest that the fourth quarter has also been strong. Real domestic demand grew at an annual rate of 4.2 per cent in the third quarter. This was partly due to a strong growth of 4.5 per cent (at an annual rate) in consumers expenditure and to an increase in stockbuilding. Business investment growth slowed in the third quarter, but it has been remarkably strong in 1988, and we estimate that investment will have grown by around 15 per cent in 1988. Some of this growth may have been the result of `tooling up' to take advantage of the recently implemented US--Canadian free trade agreement which should be of benefit to the Canadian industrial sector.

Domestic demand is estimated to have grown by 5.0 per cent in 1988, but imports, especially of machinery and equipment for investment, have grown more strongly than exports, as competitiveness has declined specifically against the US. However, in the third quarter the real merchandise trade balance improved by Canadian $2.1 billion, as import growth declined. We are projecting that both output growth and domestic demand growth will slow down considerably in 1989 both in response to higher domestic interest rates and as a consequence of the slowdown in the domestic economy.

Interest rates began to rise earlier in Canada than in the US, as the authorities began to worry about signs of rising inflationary pressure. Employment rose 190,000 (or by 1.6 per cent) in the first ten months of the year, and the unemployment rate fell sharply into the beginning of the year but has stabilised at around 7.8 per cent (recorded in November) since then. Price increases, both wholesale and consumer, have been moderated by the sharp appreciation of the Canadian against the US dollar. The CPI inflation rate moderated over the summer, but returned to an annual rate of 4.1 per cent in November. Producer prices excluding petroleum and coal products have been rising by around 6 per cent per annum recently. The strong interest-rate response to these developments reflects the commitment of both the monetary and the federal authorities to the use of a monetary policy to contain inflation. The heavy reliance on a monetary policy reflects the federal constitution of Canadian fiscal arrangements, which reduce their suitability for demand management.

Our forecast for the Canadian economy suggests that growth will pick up again more rapidly than in the US, and as a consequence we see the prospects for the Canadian balance of payments on current account deteriorating. Links with the US economy are strong, and currently some 76 per cent of Canadian exports go to the US, and 68 per cent of the imports come from there. Our research on Canadian trade was reported in the November Review. In particular we investigated the effect on Canadian trade of relative wholesale prices, in that country and in the US. We now forecast lower US wholesale price inflation, and this will cause the Canadians to lose market share over the future. The associated Canadian current account deficits will be financed by long-term capital inflows, most of which (currently 77 per cent) have in the past come from the US. These flows are likely to increase as a consequence of the Free Trade agreement, as US companies increase their industrial and resource investments in Canada.

Commodity prices Commodity price developments have been dominated in the last three months by a rise in the oil price and by unexpectedly strong copper and other metals prices. Metals prices rose 20 per cent between September and January, and oil prices rose from $12 per barrel in November to $15 per barrel in January.

Metals prices began to rise strongly in October of 1988, with copper and zinc prices being especially affected. The price rise was triggered by an industrial dispute in the Peruvian mining industry, but in normal market conditions, when prices are dominated by stocks both in the metals exchanges and in the hands of producers, prices would have been little changed. Peru produces under a tenth of world copper, and for most of the 1980s stocks would have been more than adequate to absorb this shock. However, as our chapter in the August Review detailed, world metals stocks are at low levels, and between September and October copper prices rose 21 per cent, and by December, when the strike ended, prices were a further 22 per cent higher. This latter rise partly reflected some supply difficulties in the nationalised Chilean copper mines, but again these would usually have been absorbed by buffer stock holders. Demand for copper, especially in the newly industrialised economies, was strong in 1988. The investment boom in the whole of the OECD increased demand, as many modern investment goods are electronic or electrical, and require copper and other metals in their production. The ending of the Peruvian strike saw some easing of prices, and by early February they were 10 per cent below their December peaks. However prices can be expected to remain high for much of the year. Copper stocks on the London Metal Exchange and on the US COMEX stood at 68,000 tonnes in December compared to 150,000 tonnes in mid-August. Other metal prices have also risen. Zinc prices have been influenced by the construction boom in Japan and Europe, and nickel, which is used in stainless steel, has been affected by the investment boom.

Developing country food prices, on our index, are heavily influenced by coffee and cocoa prices which have developed in diverse ways. The 1988 Brazilian coffee harvest is now estimated to be just 23 million bags rather than the initial projection of 40 million bags. This has put upward pressure on prices, which rose 15 per cent between November and January. However, in early February prices have been below their peak in the first week of the year as supplies have been released from stockpiles. Cocoa prices have been relatively weak because of a very large stock overhang which has made a workable producer--consumer agreement rather difficult to draw up.

Developed country food prices have been dominated by the continuing effects of the US drought. Wheat prices have risen further in January as fears for the US winter wheat crop have developed. Much of the US is still receiving less rain than usual, and the warm dry weather in December and January has not helped grain prospects. Maize prices also rose during December and early January. Our price index the pick up in prices around the turn of the year. The International Wheat Council estimated in early February that world grain stocks stood at 230 million tonnes, compared to output of 1,220 million tonnes. Stocks had fallen from 350 million tonnes in 1987 and from 400 million tonnes in 1986-7. Low stocks are largely the result of the US drought, and they have helped keep world food prices high. A poor Soviet harvest, some 40 million tonnes (20 per cent) below plans, has also improved the prospects for free market wheat prices.

In the longer term we expect free market food prices to rise slowly in real terms as the planned level of support by the EEC to European producers declines. It may seem paradoxical that free market prices rise as support prices fall, but lower support prices mean lower surpluses to be dumped on the world market. Producers in the LDCs should also benefit as free market sugar prices rise. We may of course, be rather optimistic about the effects of recent changes in EEC price support regimes that are now designed to penalise over-production.

Oil prices have been strong recently. The International Energy Agency estimated that oil demand rose by over 2 per cent in 1988, the highest rise in the 1980s. Some of this measured increase may however have gone into stocks. Strong demand by the industrial countries has helped OPEC behave more like a cohesive cartel, but we feel that recent price rises are partly a consequence of the Saudi decision to reduce their output below quota. We expect prices to weaken in 1989 as world activity slows, but we are assuming that after 1990 oil prices will rise at 3 per cent a year in real terms. This long-run upward trend may be related to the level of real interest rates.

World trade and the balance of payments World trade appears to have been growing very rapidly during 1988, and our widest measure, which is based on a UN series for trade in all goods by all trading economies, has been growing by between 7 and 7.5 per cent in 1988. This has partly reflected a degree of turbulence in world oil markets, and increased trade in industrial materials, but it is largely accounted for by an increase in trade in manufactured goods which we estimate has grown by 8.5 per cent in 1988, (as measured by exports from the major OECD manufacturers). Two years ago, when GATT met in Uruguay, it was widely suggested that the growth of trade in manufactures would slow as most removable trade barriers had come down. In retrospect this appears to have been rather pessimistic. The signing of the US--Canadian free trade deal was preceded by a remarkable upsurge in trade between these two economies as firms have increased investment across the border in order to take up advantageous positions once the agreement has come into effect. There are even suggestions that the investment boom in Europe can be seen as the moves made by a group of oligopolists attempting to gain strategic advantage before the removal of trade barriers in 1992. Although this may be an exaggerated, and rather Eurocentric, view of developments during 1988, it is certainly the case that the move to freer trade within Europe is attracting much attention.

Gradual liberalisation of markets in Eastern Europe is likely to increase the level of trade with the west in both directions more rapidly than previous relationships would suggest, and we have made some allowance for this in our forecasts. Greater profitability and flexibility along with a greater degree of openness is likely to lead to a considerable increase in direct investment in Eastern Europe by western countries, especially West Germany. There are many other opportunities for removing barriers to trade and capital inflows, and we anticipate that some will be taken in the next ten years, and as a consequence we expect world trade to continue at around 5.5 to 6 per cent, considerably in excess of the rate of growth of world industrial production. The loss of share by the US followed by a marked gain is significant, as is the sheer size of the US trade share. This results from the US position as the world's largest single exporter of agricultural products and of industrial materials as well as the size of its industrial exports.

The changing pattern of world current account imbalances is influenced by developments in invisible trade and in patterns of net property income. Our November Review chapter presented detailed results of our research in this area. New data series and the new IPD system on our model have led us to make some changes to our forecasts of world current account imbalances. The most notable features are the slowly declining US deficit and Japanese surplus, and this slow adjustment contrasts strongly with the stubborn refusal of the German surplus to adjust. APPENDIX GEM: Some standard simulations by R.J. Barrell and Simon Wren-Lewis The world economy chapter has regularly featured simulations or forecast variants derived from the Institute's Global Econometric Model, (GEM). However it is two years since a comprehensive set of standard simulations were presented in Wren-Lewis (1987). During that period the model has been developed in a number of directions; for example the endogenisation of service trade and IPD flows, the disaggregation of the LDC sectors, and new wage equations. In view of these changes, and the release of the model for public use, it seems appropriate to re-examine GEM's properties. In this appendix we look at a change in the real price of oil, and policy changes in each of the major four industrial countries.

An oil price increase The nominal oil price averaged $14 per barrels in 1988. What would have happened if instead it had risen by nearly 50 per cent, to $20 barrel, and that this rise in real terms had been sustained thereafter. This simulation is of some topical interest, as one of the important changes between this forecast and the last has been a sharp rise in the price of oil.

An analysis of the effects of the oil price change on the world economy has always depended crucially on how policy in the industrial countries reacts. Higher oil prices will increase inflation, and policy is likely to become more restrictive as a result. In the published version of GEM interest rates in the G7 countries are determined by estimated reaction functions, in which inflation plays an important role. However fiscal policy is effectively exogenous.

With the results of the simulation over the first three years. Interest rates rise strongly in the first year in all three major economies, averaging a one point increase. Inflation also rises, but in GEM it is nominal rather than real interest rates that are more important. The initial effect on world trade is positive. This is because the initial expansionary effect of additional income for OPEC members on their imports outweigh the deflationary effect on industrial country or LDC imports of higher oil prices.

The importance of the monetary policy reaction can easily be gauged by repeating the identical simulation with fixed nominal interest rates. After two years, world trade is up 1.1 per cent with fixed rates compared to 0.5 per cent in the main simulation. The equivalent figures for world output are--0.3 per cent and--1.4 per cent respectively. The rise in interest rates reduces the total price response from 5.2 to 4.4 per cent after three years.

The impact on the current account is greater, as a proportion of GNP, in the US than in both Germany and Japan. (With fixed exchange rates and interest rates, the German current balance deteriorates by 0.57 of GNP, Japan's by 0.43 per cent of GNP and in the US by 0.6 per cent of GNP.) This is largely because Germany and Japan benefit from the increase in OPEC imports by more than the US, and is despite the greater dependence of these economies on oil imports. (Under fixed rates, German and Japanese exports increase by 2 per cent after a year, compared to a fall cf 1.3 per cent in the US.) As a result, there is a tendency for German GNP to rise, while the impact on inflation reduces activity in both Japan and the US.

These impact effects help determine the reactions of both interest rates and the exchange rate in each country. The rise in US inflation generates an increase in nominal interest rates, which leads to an appreciation in the dollar in the first year. Both add to the deflationary effects on output of the oil price increase, so GNP falls by 2 per cent after two years. This helps to hold back consumer prices, which reach a peak of + 1 1/2 per cent after two years, and then fall back to less than 1 per cent up after five years.

The yen falls because of the greater importance of the current account in Japan's exchange-rate equation. To moderate this depreciation, as well as reduce inflation, Japanese interest rates rise sharply in the first year, and GNP falls. However the depreciation, particularly against the dollar, brings subsequent benefits to GNP, so that after three years output has fully recovered and the current account deficit has been substantially reduced.

The deutschemark depreciates for much the same reason as the yen, and there is a rise in German interest rates. (In contrast sterling appreciates and UK interest rates fall). The main difference between the reaction in Japan and Germany is that initially German GNP increases because the boost to exports to OPEC (and elsewhere) outweighs the deflationary effects of higher prices and interest rates. Interest rates have a weaker effect on domestic demand in Germany compared to Japan.

Some analytical simulations In this section we look at simulations in each of the four main industrial countries, where we keep interest rates and exchange rates exogenous. These are not designed to be realistic counterfactuals, but are presented to help us understand the properties of GEM, and in particular the fiscal policy simulations in the final section.

We first look at the fiscal policy multipliers in each country under fixed rates.

The impact effects are much larger in Japan and the US, reflecting the smaller leakage into imports. The multiplier in Japan builds up to almost two, reflecting strong accelerator effects on investment. Only in Japan and Germany is their any noticable crowding out, but in all cases the long-run multiplier is significantly positive, even after ten years. Prices have stabilised at a new higher level everywhere except in the US by this time, suggesting that this positive multiplier represents a new flow equilibrium. The natural rate falls essentially because of the appreciation in the real exchange rate, which raises real take home pay relative to the real product wage.

The fiscal expansion produces a current account deficit, and so there will be downward pressure on the exchange rate.

The most notable feature of these simulations is the absence of a J-curve that lasts more than a few quarters. For Germany this is in marked contrast to results using GEM presented two years ago, where the J-curve was very prolonged. The change largely reflects the endogenisation of service trade and IPD flows.

In all four countries, a rise in activity, inflation and (excluding the US) a fall in the exchange rate will, according to the model's reaction function, lead to a rise in nominal interest rates. The effects of interest rates on activity differ significantly among countries.

Fiscal policy simulations What happens in each country if a fiscal policy increase is accompanied by a tightening of monetary policy with a freely floating exchange rate.

In all four countries a fiscal expansion generates an increase in interest rates, and a nominal depreciation. This second result is not inevitable, and indeed in Japan the exchange rate actually appreciates in the first year. The fact that the current account effect outweighs the interest-rate effect on the exchange rate does not necessarily imply a low degree of `capital mobility' in the model, however. If the world was really like a rational expectations model of the foreign exchange market and GEM's backward-looking equations were trying to mimic this then the initial movement in the exchange rate would depend on other factors besides capital mobility.

In all cases the nominal depreciation also generates a real depreciation. This is sufficient to eliminate the current account deficit by the end of five years in all four countries. (In fact in the case of Japan there appears to be some overshooting.) This real depreciation boosts activity, which provides some offset to the deflationary effects of higher interest rates. In Japan, where interest-rate effects on domestic demand are powerful, activity is lower compared to the path, while the opposite result occurs in Germany.

The response of interest rates is determined by estimated reaction functions, which are described in Wren-Lewis (1987). The model also contains equations for narrow and wide money, and it is of some interest to observe the behaviour of these aggregates in these simulations. In the US, for example, M1 falls slightly, suggesting that an alternative monetary policy of fixing nominal money would produce slightly smaller interest-rate increases. In Germany the opposite result occurs, for both wide and narrow money. France is like Germany in this respect, while Japan follows the US.

Another interesting feature of the simulations is the spillover effect. Does fiscal reflation in one country raise or reduce GNP elsewhere? The notable result is the US case, where other countries--especially Japan--suffer from a fiscal reflation. This may explain why other countries are so keen for the US to reduce its budget deficit. The effect from reflation in other countries is more varied, and generally smaller, partly because of the importance of US interest rates in the global economy. NOTES (1)Readers may obtain, for a small charge, a complete set of our 58 forecast tables. These cover the last four years and forecasts up to 1997 for all the countries discussed in this chapter. They also include a number of summary tables. A copy of our residual settings is also available. (2)The Appendix to the August World Economy chapter in this Review gives further details on recent metals stocks. (3)The World Chapter in our November Review describes the research behind our new invisibles sector. REFERENCES Wren-Lewis, S. (1987), `Introducing exchange-rate equations into a world econometric model', National Institute Economic Review, no. 119, pp.57-69.