Supplementary Notes

International Monetary Fund. Research Dept.
Published Date:
January 1987
  • ShareShare
Show Summary Details

Supplementary Note 1

Prices of Oil and Non-Oil Primary Commodities: Recent Developments and Prospects

The markets for both oil and non-oil primary commodities have exhibited considerable weakness in recent years. The price of oil has been declining since 1981 and fell steeply in 1986 to a level that in real terms was about 15 percent below that of 1978, the year prior to the second round of large oil price increases; the real price of oil in 1986 was, however, still more than twice its level in 1972. Real non-oil commodity prices in the 1980–86 period were, on average, some 13 percent below prices in the 1960s and 1970s and had, by 1986, fallen to their lowest level since the 1930s.

The large drop in the real prices of oil and non-oil primary commodities in recent years was partly the result of factors affecting commodity markets in general. The relatively sluggish performance of the world economy and the structural changes in the economies of the industrial countries have tended to dampen demand for most primary commodities. Moreover, prices of both oil and non-oil commodities have been affected during the 1980s by greater tendencies toward substitution—to non-oil forms of energy and synthetic substitutes—and toward savings in the use of raw materials. To at least some extent, these tendencies have been the response to the relatively buoyant conditions of commodity markets in the 1970s and associated concerns about impending resource scarcities. The comparatively high real prices in the 1970s also contributed to increases in productive capacity of both oil and other primary commodities, which subsequently exerted downward pressure on prices.

The markets for oil and other primary commodities have, however, also been influenced by special factors which have been manifested in marked variations in both the degree and timing of price movements. The dissimilarities stem in part from the somewhat different nature of the various markets, particularly on the supply side. Whereas the markets for most non-oil primary commodities have usually included numerous sellers, the world oil market is characterized by a high degree of concentration on the supply side and a limited number of decision makers, mainly the governments of the major oil exporting countries. This has meant that factors affecting supply, particularly the degree of control over it, have generally been a more important determinant of the price of oil than of prices of other commodities. Given the particular characteristics of crude oil, especially its low short-term price elasticity of demand, oil prices have also tended to be subject to large and abrupt changes, at times deviating significantly from underlying longer-run demand and supply conditions. Other characteristics that distinguish oil from most other primary commodities, of course, are its nature as an exhaustible natural resource and the high concentration of world oil reserves in a few countries. These latter factors will be of particular importance for the evolution of relative prices in the longer run.

The relatively similar general behavior of oil and non-oil commodity prices in the recent past, notwithstanding the differences noted above, can be explained partly by the coincidental occurrence of unrelated developments, such as the unusually favorable growing conditions for agricultural crops and the loss of control over the determination of the price of oil by the major oil exporting countries, which in part was the result of both economic and political differences among these countries. While the markets for both oil and other primary commodities are expected to remain subdued in the immediate future, the differences in underlying supply and demand conditions are likely to reassert themselves in the longer run. Primarily because of the declining reserve base of some important producers and the uneven distribution of the remaining world oil reserves, longer-term prospects for the price of oil appear more favorable than for most other primary commodities.

The discussion in the following section of this note on recent developments and prospects for prices of non-oil primary commodities draws on recent studies by the staff of the Fund and the World Bank to be discussed by the Development Committee in April 1987. The final section of this note reviews recent developments and prospects for the price of oil.

Non-Oil Primary Commodity Prices16

Recent Developments

Prices of primary commodities have declined substantially so far in the 1980s.16 In real terms, commodity prices fell by about 30 percent from 1980 to 1986 and were in this period, on average, over 15 percent below the level in the 1970s. The weakness of commodity prices was accentuated in 1986 when the real price index fell by 16 percent to its lowest level since the 1930s. Because of the depreciation of the U.S. dollar, the decline in nominal U.S. dollar terms was considerably smaller (about 1 percent). During the whole period from 1980 to 1986, however, the fall in the dollar price index was commensurate with the drop in real prices (see Charts 19 and 20).

Chart 20.Prices for Groups of Non-Oil Commodities, 1980–86

(1980= 100, in terms of U.S. dollars)

While the weakness of commodity prices has been widespread, suggesting the existence of common causative factors, prices of individual commodities and commodity groups have, of course, been affected by special circumstances. During the period from 1980 to 1986, average prices of both food items and metals fell by about 40 percent in real terms, while prices of agricultural raw materials declined by about 30 percent. Only beverages, among the four major commodity groups, failed to show any appreciable real price decline in the 1980–86 period, but this mainly reflected the sharp rise in coffee prices in late 1985 and in the first quarter of 1986 resulting from extensive drought damage to the coffee crop in Brazil. Average real prices of other commodities declined markedly in 1986.

The prolonged and pervasive weakness of commodity markets has given cause for concern on at least two grounds. One relates to the difficulty in explaining in a satisfactory way the large fall in real prices in the past few years, particularly as it occurred during a period of sustained, albeit moderate, economic growth. This combination of circumstances has prompted intensified research about the behavior of commodity prices and has revived attention on the possibility of a long-term declining trend in real prices. The second, and related, concern is with the effects of the fall in commodity prices on developing countries and its implications for their future economic welfare. The production and export of primary commodities have declined significantly in relation to total output and exports of the developing world, in part reflecting economic growth and diversification, while domestic consumption of such commodities has increased considerably in several countries. Nevertheless, exports of primary commodities are still important for many developing countries, particularly in the groups of low-income and heavily indebted countries; several countries in Africa, for instance, remain heavily dependent upon such commodities. Moreover, with the recent sharp drop in oil prices, some developing countries have become more dependent upon the export of non-oil primary commodities.

Recent studies conducted by the staff of the Fund and the World Bank have confirmed that traditional short-term demand and supply factors have played a major role in the weakness of commodity prices in the past few years. They also suggest, however, that longer-term structural factors affecting both the supply of and the demand for primary commodities, while not a dominant influence on recent price movements, may have been more important than earlier believed. In addition, policies pursued by governments, particularly those of industrial countries, have been found to have had depressing effects on international prices of primary commodities. The discussion below will be limited to an overview of the major factors affecting commodity prices, with emphasis on both the short-term and longer-term dimensions of the issue.

The most recent downturn in commodity prices started around the middle of 1984 and, for most commodities, proceeded almost continuously through 1986. Although a number of factors played a role in this decline, it is clear that short-term factors operating on the supply side had the greatest impact. Production of primary commodities increased strongly in 1984 and continued to expand in the following two years, with new supplies adding to already large stocks. The marked rise in output of almost all agricultural commodities—coffee and cotton being notable recent exceptions—reflected mainly the favorable growing conditions in major producing areas but also, in some cases, lagged responses to earlier price increases. The largest production increases during the 1984–86 period were recorded for food items for which supply tends to have a dominant influence on short-run movements in prices. As food items also have a large weight in the overall index, this supply buildup was the major factor explaining the recent downturn in commodity prices.

The influence of unusually favorable supply conditions tended to be reinforced by relatively weak demand for some commodities in the past two years. Of particular importance in this respect has been the weakness of industrial activity in the industrial countries. Following the strong cyclical recovery in 1984, the growth in industrial output of these countries fell sharply in 1985 and 1986 to only about 1 percent in the latter year. This development had a significant effect on the demand for metals and, to a lesser extent, agricultural raw materials.

The marked slowdown in world inflation in recent years, of course, had a considerable impact on commodity prices in nominal terms. (The supply-induced fall in commodity prices was itself a factor in the general slowdown of inflation.) However, the effect on real commodity prices of the slower increase in the overall price level is not thought to have been important, except possibly by weakening the speculative demand for commodities. While the results of various econometric studies on the impact of changes in interest rates and in exchange rates are somewhat ambiguous, such changes are also not believed to have significantly affected average real commodity prices in recent years.

Beyond the short-term factors discussed above, there are strong indications, even though quantification is difficult, that the sustained weakness of commodity prices has also been the result of structural and longer-term factors. On the supply side, there is considerable evidence that technological progress and the dissemination of new technology have tended to increase output, particularly of agricultural commodities. The introduction of new higher-yielding strains of seeds, in combination with increased use of improved production inputs, has served to raise productivity and yields in many countries. As a result of government policies promoting such developments, some developing countries have not only become self-sufficient in foodgrain production but have also produced export surpluses in recent years. Other developments on the supply side include the achievement of substantial cost reductions in the output of minerals, resulting from a shift to low-cost producers and efforts to reduce costs and raise productivity in existing mines.

On the demand side, longer-term structural changes have perhaps been even more important in explaining the weakness of commodity prices in the 1980s. Demand for primary commodities has tended to be reduced by structural shifts in the economies of several industrial countries away from heavy industry toward light manufacturing (toward the production of electronics, for example) and the service sectors. This shift has coincided with rapid progress in the development of synthetic substitutes and new materials, advances in mineral science, a trend toward miniaturization and downsizing of appliances and consumer durables (such as automobiles), and increased recycling, as well as conservation of metals in traditional uses. Partly as a result of such changes, world consumption of metals remained stagnant in the 1980–86 period, whereas industrial output of the industrial countries increased by about 13 percent in the period. The tendencies toward substitution and savings in the use of raw materials may have been influenced, at least in part, by earlier predictions of resource scareities caused by the sharp rise in energy prices in the period from 1973 to 1981 and tighter supply conditions for other commodities in the 1970s. Although such predictions have not been borne out, the induced research and technological progress may have developed a built-in momentum in some areas. The more recent fall in oil and other energy prices has, at the same time, tended to weaken the prices of some non-oil primary commodities for which oil-derived substitutes are important and readily available. The fall in energy prices has also affected prices of other primary commodities by lowering transportation costs and by reducing prices of important production inputs, such as fertilizers.

It has been recognized for some time that the policies pursued by governments of the industrial countries, and to some extent also developing countries, have tended to depress commodity prices. Policies in the industrial countries, which account for more than one half of world exports of agricultural commodities (excluding beverages), have for several years generally been aimed at subsidizing domestic producers, improving their competitive positions, and insulating them in various degrees from adverse developments in world markets. These policies have included protection from foreign agricultural products and the support of prices and incomes of farmers. In combination, such measures have resulted in increased production of a number of commodities, with surpluses being sold on world markets or being reflected in an accumulation of inventories.

These policies are now being re-examined in light of the excess supply that has existed in world agricultural markets in the past few years. While a reorientation of agricultural policies should over time lead to a firming of commodity markets, at least one important policy initiative is exerting a further downward pressure on world commodity prices in the short run. The U.S. farm legislation that came into effect in early 1986 (the Food Security Act of 1985) attempts to make U.S. farm policy more market oriented—mainly by reducing effective price support levels—and is expected to result in lower supplies in the medium to longer term. The immediate effect, however, has been the release of sizable quantities of stocks, which has tended to depress prices. Measures have also been taken to reduce stocks in other industrial countries, in some cases in order to limit the budgetary costs of price support programs.

In developing countries, government policies have in the past often been designed to maintain, or have had the effect of maintaining, domestic agricultural prices below world market levels for various reasons: for example, to provide budgetary revenue or to keep domestic prices of food items low. In recent years, however, many developing countries have begun to implement policies, often as part of structural adjustment programs, to allow domestic prices of agricultural products to become more market oriented. While such policy shifts are generally desirable, in the short run they have often tended to depress commodity prices by encouraging higher production.


World commodity markets are expected to remain weak in the immediate future and to improve only moderately in the medium term. The staffs current projections indicate that the overall annual index of commodity prices will decline further in 1987, by about 5 percent in nominal (U.S. dollar) terms and by about 14 percent in real terms. In the following four years (1988–91), nominal dollar prices are projected to rise by about 5 percent a year, while in real terms the increase would come to about 2 percent a year. If these projections were to materialize, real commodity prices would not, by 1991, have recovered to their 1986 level and would be about one third below their level in 1980.

A major assumption underlying these projections is that most of the factors that have tended to depress commodity prices in recent years will continue to operate in the short to medium term, with only a gradual tendency for improvement. The prospective global economic environment, discussed in the main body of this publication, is not expected to be conducive to any significant recovery of commodity prices. While industrial output of the industrial countries is expected to expand at a somewhat faster pace in 1987–88 than in 1986, the growth in aggregate real output of these countries is projected to pick up only slightly and inflation is expected to remain low. Moreover, while the tendency toward reduced intensity in the use of raw materials and other longer-run influences, discussed earlier, may weaken somewhat in view of the marked fall in real commodity prices, they will continue to affect prices in the medium term.

In the near future, specific factors affecting the supply of and the demand for particular commodities will, of course, affect conditions in commodity markets. Prices of food items are expected to decline further in 1987 under the influence of high stocks and competition for export markets, but the decline is likely to be considerably smaller than in the past two years, partly because it is unlikely that global weather conditions will continue to be as favorable. Prices of beverages, which had already started to fall in the second half of 1986, are expected to decline sharply in 1987 on a year-over-year basis (by more than 20 percent in nominal dollar terms) on the assumption that normal weather conditions will prevail. Prices of agricultural raw materials and metals are expected to increase moderately, partly because of the anticipated upturn in industrial output of the industrial world. However, any tendency for metals prices to increase will be restrained by the existence of considerable excess capacity.

A major reason for the projection of a moderate rise in real commodity prices after 1987 is the expectation that food prices will recover significantly from their depressed levels of 1986–87. This recovery would result primarily from lagged supply responses to the low prices of most food items in those two years. Moreover, it is expected that the recent U.S. farm legislation, as well as the reorientation of agricultural policies in other countries, noted earlier, will begin to restrict supplies through the operation of market forces. Average real prices of other primary commodities are, however, expected to rise only slightly in the period up to 1991 in view of the anticipated slow growth in demand and increased supplies.

These projections, while not favorable for exporters of primary commodities, are subject to considerable margins of error. It is possible that the growth in world consumption of commodities will be slower than anticipated, even barring the possibility of a recession in the industrial countries, and that further technological progress (such as the introduction of genetic engineering techniques in foodgrain production) could have additional depressing effects on commodity markets. Better economic performance in developing countries—entailing a higher growth in per capita incomes—would, on the other hand, have a positive impact on real commodity prices. A rise in energy prices would also tend to support prices for some other primary commodities. Moreover, prices of metals may begin to firm when the presently idled capacity has been brought back into production.

Perhaps the best prospects for higher commodity prices than envisaged in this report, however, would lie in improved policies of producing and consuming countries. Of particular importance in this respect would be a sustained effort on the part of the industrial countries to dismantle the barriers to trade in primary commodities and to reduce direct and indirect subsidies to domestic producers. Improved farm policies should, in the long run, lead to more efficient production and greater reliance on international comparative advantage. Policies aimed at fostering such developments could be implemented within the framework of the forthcoming round of the Multilateral Trade Negotiations of the GATT. Nevertheless, even with such policy improvements, the medium-term Outlook for commodity prices is not very promising.

Price of Oil

Recent Developments

The international price of oil has been subject to major changes during the past decade. The annual average price of oil exported by the major oil exporting countries17 in nominal (U.S. dollar) terms first increased by 162 percent from 1978 to 1981, then fell back moderately (by about 21 percent) during the following four years (1982–85), before dropping by 48 percent in 1986. As a result, the average nominal price in 1986, estimated at about $13.80 a barrel, was about 7½ percent above the level in 1978. In real terms, however, the 1986 price was about 15 percent lower than in 1978, having fallen by about 55 percent from 1985 to 1986 (see Table 23).

Table 23.Selected Oil Market Indicators, 1979–86
(Changes in percent)1
Price of oil
In nominal terms2162-21107-48
In real terms3109-1186-55
World oil consumption-4½-2½-72
Industrial countries-12½-6½-18½
Oil exporting developing countries2614½44½-1
Oil importing developing contries3
Other countries8-351
(In percentage of total)
Distribution of world oil production
Industrial countries22½26½29½28
Members of OPEC48353032½
Other developing countries11161918
Other countries18½22½21½21½
Share of OPEC in world net oil exports488736467

In examining the causes of these movements, it is important to note that the periods of major price shifts (1979–80 and 1986) were not, on the whole, characterized by large or sudden changes in underlying demand and supply conditions. The explanation has to be sought partly in the lack of effective management of the oil market during those periods and in the particular characteristics of crude oil, especially its low short-term price elasticity of demand, which in turn is related to its strategic importance and the lack of close substitutes in some uses (for example, as a transportation fuel). The periods of relative stability in the oil market have generally been characterized by the management of prices or supplies. During the 1950s and 1960s, this function was carried out by the major international oil companies, and during most of the 1970s, and again in the 1983–85 period, by the members of OPEC.

During the latter period the members of OPEC followed a policy of concerted output restraint under which production quotas were in effect for individual member countries. This policy succeeded in limiting the erosion of oil prices through 1985 in large measure because one OPEC member (Saudi Arabia) had assumed the role of swing producer. However, when some other member countries were unable to maintain output within their quota limits and also began to provide discounts on the agreed official export prices in various direct and indirect ways, Saudi Arabia was faced with a large and continuous decline in its annual crude oil production, which eventually became unsustainable. In the end, therefore, the swing producer feature of the production-sharing arrangement proved to be the proximate cause of its demise. There was a major shift in Saudi Arabia’s oil policies in the latter part of 1985, with the primary aim of restoring the country’s oil production and market share. In addition to relinquishing the swing-producer role, Saudi Arabia’s crude oil pricing policy was changed from the use of official selling prices toward increasing reliance on sales at prices linked to the market prices for refined products (netback prices). A further important development in this period was the decision adopted by the members of OPEC at their meeting on December 5–7, 1985 to “…secure and defend for OPEC a fair share in the world oil market . . . .” This decision was generally interpreted as a de facto abandonment of the previous policy of attempting to administer international crude oil prices through the maintenance of agreed official export prices and concerted output restraint.18

These developments were followed by a considerable, and almost continuous, rise in total OPEC crude oil production which, by August 1986, was running almost one third above the average output in 1985. Mainly as a result of this supply increase, which outstripped the rise in consumption and led to a considerable build-up of inventories, crude oil prices fell precipitously during the first quarter of 1986 and continued to weaken in the following months. Spot market prices for representative crude oils dropped from about $27–31 a barrel in November 1985 to about $10–13 a barrel in April 1986 (see Chart 21). Following a temporary recovery in May, prices fell again to new lows of about $8–10 a barrel for some crude oils in July 1986. Crude oil prices in this period were also highly volatile, being influenced by speculative and transient factors, and the differentials between various crude oils shifted considerably.19

Chart 21.Spot Market Prices for Selected Crude Oils, 1983–87

(In U.S. dollars a barrel)

1OPEC’s marker crude oil up to 1985

The sharp drop in oil prices and unstable market conditions led to increasing calls, from both inside and outside OPEC, for the re-introduction of concerted output restraint to bring about a recovery of prices and the restoration of more orderly market conditions. Increasing contacts were made between OPEC countries and important non-OPEC oil exporters, several of which indicated a willingness to help achieve these objectives. Following unsuccessful earlier attempts, in early August 1986 the members of OPEC agreed to restore previous production quotas for September-October 1986 on an interim basis. In October 1986, this agreement was extended until the end of the year with some minor modifications. Mainly as a consequence of the resulting production cutbacks, spot market prices recovered considerably in the latter part of 1986 and had, by early December, reached some $13–15 a barrel.

At their meeting on December 11–20, 1986, the members of OPEC adopted two decisions of more far-reaching importance. New production quotas were agreed upon for the first half of 1987, entailing a reduction in total output of about 7–9 percent from estimated actual production in the fourth quarter of 1986. (Iraq disassociated itself from the agreement but was allocated a nominal quota.) In addition, fixed official export prices were re-introduced. New official prices were announced for all major OPEC crude oils, based on a reference price of $18 a barrel, representing the unweighted average price of seven medium and lighter crude oils. After allowing for a phase-out period for current contracts (under netback, spot, or other price arrangements), it was agreed that the new prices would come into effect on February 1, 1987.

These agreements began to be implemented after the OPEC meeting, and in January 1987 several OPEC countries had started to cut their production and give notice to customers that crude oil would be sold only at the new official prices. Mainly as a consequence of these actions, spot market prices firmed considerably to some $17–19 a barrel in the second half of January. It should be noted, however, that this firming occurred during a period of seasonal rise in oil consumption, which was accentuated by the unusually cold weather in Europe, and that prices in early 1987 may also have been affected by uncertainties associated with the flare-up of the conflict between the Islamic Republic of Iran and Iraq. Spot market prices subsequently declined in February 1986 but started to recover again in the first half of March under the influence of reduced shipments from the members of OPEC.

The sharp fall in oil prices since the latter part of 1985 contributed to significant changes on both the demand and supply sides of the oil market. Oil consumption in the industrial countries, which had declined by about 1 percent in 1985, is estimated to have risen by about 2½ percent in 1986, despite the moderating growth in real output. Even though part of the fall in crude oil prices was not passed on to consumers in several countries, a considerable portion of the recovery in oil demand in the industrial world last year can be attributed to the effects of lower end-user prices.20 The increase in oil consumption reflected mainly enhanced use of transportation fuels but also some switching back to heavy fuel oil from other sources of energy (particularly natural gas in the United States) in the electricity generation and industrial sectors. As a result, the growth in oil consumption of the industrial countries is estimated to have exceeded that of total energy use for the first time in about a decade. Oil consumption also rose significantly in several oil importing developing countries last year, and some of them have recently taken steps to curtail the rise in demand.

Of perhaps greater importance than these developments, however, were the changes that occurred on the supply side of the oil market. One notable development was that the members of OPEC were able—in line with their stated policy—to reverse partly the previous large decline in their combined market share (see Table 23). The total volume of oil exports from these countries is estimated to have risen by about 15 percent from 1985 to 1986, with most of the increase accounted for by Saudi Arabia and other Middle Eastern producers. Total non-OPEC output, on the other hand, declined marginally in 1986 for the first time in more than a decade. These shifts in market shares partly reflected the more effective marketing policies of OPEC countries in the first half of 1986 (particularly the use of netback pricing) and voluntary output or export restraint by a few non-OPEC oil exporters. However, the fall in non-OPEC output in 1986 also resulted from a reduction in high-cost production, including the shutting down of “stripper” wells (oil wells producing less than 10 barrels a day) and the impact of reduced drilling activity. These latter effects have been concentrated in the United States, where oil production fell markedly during the course of the year. Total output of the oil importing developing countries, on the other hand, continued to rise in 1986.

Another important effect of the 1986 fall in oil prices was the considerable decline in worldwide spending by the oil industry on exploration and on the development of productive capacity (including the use of enhanced recovery methods). Such spending is estimated to have declined by some 30—40 percent in 1986, with the major part of the decline occurring in the United States.


Developments during the past decade provide ample evidence of the instability of oil prices during periods of lack of effective management of the oil market and of the propensity of the price of oil to deviate significantly from underlying longer-term demand and supply conditions. Any projection of oil prices will, therefore, be tenuous. For the immediate future, there can be little doubt that factors affecting supply—particularly the degree of output restraint exercised by the members of OPEC—will continue to be the dominant influence on the price of oil. The main reasons for this contention are the continued existence of considerable unused production capacity among these countries21 and the likelihood that changes in oil demand and supply from other areas will not be of sufficient magnitudes to affect substantially the price level in the short run.

The key issue in the short-term evolution of oil prices, therefore, is whether the recent agreement reached by the members of OPEC can and will be adhered to. Although there remain considerable differences of view about the optimal price of oil among these countries, virtually all of them currently appear determined to achieve oil market stability at the price level recently agreed upon. Considerations underlying this determination include the pressing need for additional revenue for most countries and the widespread recognition, in light of developments in 1986, that lack of sufficient production restraint could lead to markedly lower oil revenue than otherwise would be the case. The possibility of new measures to protect domestic energy producers and to maintain the momentum for energy conservation, in the form of domestic energy taxation or oil import tariffs or fees, on the part of major oil importing countries may also have played a role. For much the same reasons, several important non-OPEC exporters have announced that they will support the recent OPEC agreement, mainly through production or export restraint.

The members of OPEC will, however, be faced with considerable difficulties in implementing the policy recently agreed upon, particularly during periods of seasonal shifts in oil consumption. Among the more important problems are the absence of a single swing producer able and willing to absorb the shifts in demand (as in the 1983–85 period), the difficulties in managing and adjusting crude oil price differentials in an environment of significant shifts in final demand and product prices, the considerable fragmentation of the oil market on the buying side, and the increased role of trading in the spot and futures markets. These circumstances may call for the innovative use of new techniques to adjust both production quotas and price differentials in response to short-term changes in market conditions. Another potential problem in early 1987 is the existence of relatively large commercial inventories resulting from the high level of OPEC production last summer. A rapid drawdown of these inventories could accentuate any renewed weakness in oil prices if the official export prices are not adhered to. Prospective developments in the conflict between the Islamic Republic of Iran and Iraq, and their uncertain effects on the oil market in the short and medium term, represent another major uncertainty in any assessment of future oil price developments.

As noted earlier, other prospective developments do not appear likely to exert a major influence on the price of oil in the near future. Oil consumption in both the industrial and non-oil developing countries is expected to continue to rise in 1987 and 1988, albeit at a slower pace than in 1986, on the assumption that the average price of oil will be in the $15–18 a barrel range. Total non-OPEC production is likely to rise only moderately in both years, signifying a break from the strong upward trend prior to 1986, with a further decline in the industrial countries offset by an increase in developing countries. On the basis of these estimates, world demand for oil from the members of OPEC in 1987 could be somewhat higher than the maximum rate of production agreed upon last December for the first half of 1987, even allowing for a sizable drawdown of inventories from the high level at the end of 1986. (The OPEC agreement provides for quota increases in the second half of 1987 to be implemented if market conditions permit.) Mainly because of the anticipated inventory shift, the annual volume of oil exports from these countries is expected to decline (by some 5–6 percent) in 1987 before rising again in 1988.

In assessing the medium- to longer-term Outlook for oil prices, trends in worldwide demand and supply obviously assume much greater importance than in the appraisal of near-term prospects. Although developments in 1986 provide only limited guidance on the impact of the downward adjustment of the price of oil, prospective longer-term developments point toward a considerable firming of oil prices within the next ten years. The turnaround in oil market conditions would, in essence, result from a gradual rise in demand for OPEC oil and a consequent reduction of the unused production capacity of these countries, a principal underlying reason for their loss of control of oil prices in 1986.

World oil consumption is expected to rise at the relatively moderate pace of about 1–1½ percent a year during the next five to ten years, with somewhat more rapid growth in developing countries than in the industrial world. This assessment is based on the expectation that the gains in energy conservation and the replacement of oil by other forms of primary energy will continue, although at a reduced pace. While the effects of earlier price increases will tend to be offset by the impact of the 1986 price drop, investment decisions will be influenced by the widespread perception among important energy users of future oil price increases. The expectation of further substitution away from oil, following the temporary reversal of this tendency in 1986, is also based on the availability of ample supplies of other major forms of energy (particularly coal in the United States and natural gas in other parts of the world) and a further expansion in installed nuclear capacity, although at a much slower pace than during the past decade.

The relatively moderate medium- to longer-term growth in world oil consumption, discussed above, could easily be accommodated from the existing total resource base. On the basis of proven reserves alone, worldwide consumption could be sustained at the 1986 level for about a third of a century, and additional reserves will undoubtedly be found. Remaining world oil reserves—both those currently proven and those likely to be found—are, however, very unevenly distributed, with almost three fifths of proven reserves located in a limited number of countries in the Middle East, which also offer the best prospects for large new finds. With recent discoveries generally being of modest size and with a number of oil fields reaching maturity, the prospects for sizable future production increases in most other areas are not promising. The diminishing returns on world oil exploration (outside the Middle East) in recent years, and the sharp reduction in investment expenditures by the oil industry in the wake of the 1986 drop in oil prices, lend further support to this assessment. While investment spending is likely to increase in the next few years, the recovery is expected to be slow and gradual unless prices rise to $20 a barrel or beyond.

Supply prospects in the industrial countries are dominated by developments in the United States, which has accounted for some two thirds of total oil production in this group of countries in recent years. Although part of the fall in U.S. oil production during 1986 may be temporary, it is very likely that the long-expected gradual decline in U.S. output is now under way. However, while there is virtual unanimity among experts on the subject that U.S. oil production will decline further in the medium term, assessments differ considerably on the extent of the decline, with estimates for the next five years ranging from ½ million barrels a day to as much as 2–3 million barrels a day. Elsewhere in the industrial world, oil production is also expected to decline in the medium term in the United Kingdom and Australia—where most oil, as in the United States, is produced from maturing fields with a naturally declining reserve base—while earlier prospects for Canadian output from frontier areas and from tar sands have been set back in view of the recent oil price drop. Only Norway among the industrial countries currently offers the prospect of any significant increase in oil production in the next few years; actual output in that country may, however, be influenced by the oil policies pursued by the authorities.

Supply prospects are somewhat more favorable in the non-OPEC developing countries. Significant new discoveries have been made in some of these countries in recent years (in, for example, Colombia, the Syrian Arab Republic, and the Yemen Arab Republic), and production is expected to continue to increase in several other countries. However, the rise in total output of these countries is only likely to match the anticipated decline in the industrial countries. While prospects for the U.S.S.R. (the world’s largest oil producer) remain highly uncertain, longer-term trends continue to point toward a decline in oil exports of that country.

On the basis of these prospective developments, the demand for oil from the major oil exporting countries (or the members of OPEC) seems likely to rise in the medium to longer term, with a consequent reduction in their unused production capacity. While the changes in oil market conditions may initially be slow to materialize, they are likely to result eventually in a considerable firming of prices. During the next few years, however, the price of oil is likely to remain potentially very unstable and will be influenced primarily by the degree of production restraint exercised by the major oil exporting countries.

For the purpose of this report, the staff has adopted the working assumption that the annual average export price of the major oil exporting countries will increase in nominal (dollar) terms by about 9 percent from an estimated $13.80 a barrel in 1986 to $15 a barrel in 1987. Given the projected further large increase in world non-oil trade prices (in dollar terms), this would represent a decline in the real price of oil of about 2 percent. For the years 1988–91), the staff has assumed that the price of oil will remain constant in real terms, implying a rise in the nominal (U.S. dollar) price of 3 percent a year. These assumptions are obviously subject to large margins of error, particularly for the immediate future. The average price in 1987 could be some 10 percent higher than has been assumed if the December OPEC agreement were to be fully implemented.22 If the current OPEC agreement were to break down, on the other hand, crude oil prices could again fall rapidly, possibly to less than $10 a barrel. Prices are likely to be more unstable under this latter scenario and it is unlikely that the average price level will remain below $10 a barrel for an extended period. While medium-term price prospects are also highly uncertain, the price assumed for the latter part of the 1988–91 period would, on balance, appear to carry a higher probability of being on the low side, for reasons discussed above.

Supplementary Note 2

Discrepancy in World Current Account Balances

A troublesome aspect of global analysis over the past decade has been the large and variable imbalance observed when national balances on current account are summed across all countries. In principle, of course, the world current account should sum to zero, with each deficit being matched at the aggregate level by an offsetting surplus. In practice, however, the recorded aggregate balance on world current account has been in substantial deficit for several years (Table A30). From approximate balance in the early 1970s, the aggregate global current account deficit reached about $20 billion in the late 1970s and escalated to over $100 billion in 1982 before receding in the mid-1980s.

Discrepancies in world current account balances of this magnitude tend to undermine the analysis of balance of payments and output developments. Analyses of global savings and investment flows, and especially of the flow of real resources reaching developing countries, are clearly hampered by a global pattern of current account balances such as that prevailing in 1982–84 (Table 24). From a resource transfer point of view, the first question prompted by this table is: Which group of countries transferred resources to the other? For 1982, conventional notions of a proportionate sharing of the discrepancy suggests that the industrial countries were in surplus in that year. But do those same notions carry over to 1984 when developing country balances were much closer to, and industrial country balances correspondingly further from, equilibrium? Had the resource transfer turned negative in that year? Given the size of the discrepancy, the answer is difficult to ascertain.

Table 24.Global Payments Balances on Current Account, 1982–84(In billions of U.S. dollars)
Industrial countries-22-22-62-40
Developing countries-87-64-342353
Other countries3341
Source: Table A30.

Table 24 also illustrates how the discrepancy can hamper the analysis of short-run developments as well. The improvement in the current account balances of developing countries in 1982–83 is matched by a reduction in the global discrepancy, with no matching deterioration in the industrial country balance. Does this mean that the improvement in developing country balances was purely statistical, these countries having uncovered previously unrecorded net credits? Or does it reflect an underlying improvement in their balances that is partly masked by deficiencies in the balance of payments statistics of industrial countries?

Surveillance over the exchange rates of the major industrial countries is also hampered by the discrepancy, since it masks the possibility that one or more of these countries may account for a sizable fraction of the discrepancy. Uncertainties in this respect make analysis of imbalances and exchange rates more imprecise than they would otherwise be. On a different plane, the discrepancy has also given rise to concerns that the bias toward recording deficits in national balances of payments may have given national financial policies a deflationary bias as national authorities sought to strengthen their external positions. By the same token, however, the increase in the discrepancy from 1979 to 1982, which is a measure of the shortfall in recorded net exports, suggests that the growth of global output over that period may have been underestimated by as much as 1 percentage point.

Consequently, the Fund established in September 1984 a Working Party, supported by a technical staff, to investigate the discrepancy and to recommend changes in statistical methodology that might help reduce the size and the variability of the discrepancy. The Working Party completed its work in October 1986 and forwarded its final report to the Fund’s Executive Board in early 1987. The report makes a number of recommendations, the broad thrust of which was accepted by Executive Directors. Many of these recommendations pertain to desirable adjustments to balance of payments compilation procedures, to the need for a revision to the Fund’s Balance of Payments Manual, and to suggested lines of further research. A good part of the report, however, deals with matters that are of more immediate interest to global analyses, and these are briefly reported on here.

The report confirms and amplifies the observation that the discrepancy is highly concentrated in particular types of current account transactions (Table 25). The main points are as follows.

Table 25.Aggregate Current Account Transactions of Industrial and Developing Countries, 1979–851(In billions of U.S. dollars)
Balance on current account-17-31-65-106-82-89-70
Trade balance203126-13108
Balance on services-28-49-79-93-76-89-65
Shipment and transportation-30-35-40-37-32-31-22
Investment income6-8-24-41-35-48-43
Of which:
Reinvested earnings141311281021
Non-direct (portfolio)
investment income-10-11-22-34-32-43-51
Other services-4-7-15-14-10-11
Transfers, net-9-13-12-14-9-10-13
Source: Balance of Payments Statistics Year book (Washington: International Monetary Fund, 1986).

(1) The discrepancy is to be found primarily in the service and transfer transactions. While the statistics on merchandise transactions have deficiencies, the biases at the aggregate level are reasonably well known (see the memorandum items in Appendix Table A30 and associated footnotes) and have not, apart from year-to-year fluctuations, contributed to the pronounced widening trend in the discrepancy over the past decade.

(2) Within the service account, the largest and most rapidly rising discrepancy has been with respect to portfolio income. This discrepancy—which was the primary focus of the Working Party’s efforts—was found to be due to two factors: the emergence of a large body of cross-border assets recognized by the debtor countries but not by the creditor countries; and the rapid rise in international interest rates from the late 1970s. As a result, the discrepancy on portfolio income rose from $10 billion in 1979 to $51 billion in 1985—or three fourths of the increase in the overall discrepancy over this period. The discrepancy on portfolio income basically carries over to the entire investment income account, although it is partly offset by an opposite discrepancy on reinvested earnings where debtor rather than creditor countries tend not to record the transactions.

(3) Another major but relatively stable discrepancy within the service account is in the shipping data. The Working Party’s main finding here is that, whereas countries are generally successful in compiling the amounts paid to foreign-operated carriers for the freight on imports, the corresponding freight earnings of some of the economies with large maritime interests, as well as the earnings associated with ship expenditures, are not fully captured in balance of payments statistics.

(4) There is a large, but again fairly stable, asymmetry on official transfers, with payments reported by donors exceeding credits reported by recipient countries. The Working Party was unable to account for all this discrepancy, but did find that about a third of it stemmed from the omission of international organizations, which are a major channel for these transfers, from global balance of payments statistics.

Although these statistical explanations of the origins of the discrepancy provide essential guidance if compilers are to improve the statistics, they do not directly address the issue that is of greatest analytical concern—the geographical distribution of the discrepancy. This is an inherently more difficult task, requiring the allocation of the discrepancy for any one type of transactions to particular countries or groups of countries. Progress in this area has been harder to achieve. Indeed, the Working Party finds that the best solution would be for national compilers to adjust their estimation procedures to make them more consistent with one another so as to absorb the discrepancy gradually into recorded country estimates. Nevertheless, the Working Party’s research on the investment income account does lend itself to geographic disaggregation, and some initial rough estimates of the adjusted pattern of global current account balances can be reported.

Table 26 and Chart 22 present estimates of the global pattern of current account balances both as reported in the standard World Economic Outlook tables and after adjustment for those components of the investment income discrepancy that the Working Party felt reasonably confident could be attributed to industrial or developing countries. As noted earlier, this attribution stems primarily from the acknowledgement by debtors of liabilities to creditor countries for which no counterpart claim is recorded in the statistics of the latter countries. Accordingly, the Working Party corrected the statistics of the creditor countries to incorporate estimates, based on presumed interest rates, of the investment income on those claims.

Table 26.Summary of Global Payments Balances on Current Account—Recorded and Adjusted, 1979–86(In billions of U.S. dollars)
Industrial countries
Developing countries
Recorded-17-32-69- 109-86-96-79-57
Other countries-3-2-333431
International organizations23333455
Total, adjusted-22-18-42-57-52-50-36-34
Note: For 1979–84, adjustments are based on the Final Report of the Working Party on the Statistical Discrepancy in World Current Account Balances, January 12, 1987. For 1985–86, adjustments are based on preliminary staff estimates.

Chart 22.Balances on Current Account, 1978–86

(In billions of U.S. dollars)

As can be seen, the adjustments attributed to the industrial countries account for about three fifths of the total and those attributed to developing countries account for the remaining two fifths. Moreover, the adjustments seem to move in tandem. Thus, although the adjustments go a long way toward limiting the global discrepancy (compare the total recorded and adjusted lines in Table 26), they do not greatly affect existing views about the global pattern of current account balances or, as the Working Party put it, invalidate the existing analyses of world economic developments.

Nonetheless, this work has been useful and it is important that the effort initiated by the Working Party be carried forward in the interested international organizations and, especially, by national balance of payments compilers. This usefulness is highlighted by the clarification afforded by the Working Party’s work in the interpretation of global developments in 1982–84 described earlier. As will be recalled, considerable uncertainty was attached to developments in global current balances in 1982–83 because the change in the discrepancy was as large as the improvement in the current account balances of developing countries. This picture is substantially clarified in the adjusted statistics in Table 26, where the swing in the discrepancy in 1982–83 is a fifth of its former size, with the adjusted figures pointing clearly to a significant improvement in developing country balances and a roughly matching deterioration in industrial country balances. Moreover, the adjusted statistics, although far from definitive, do point fairly clearly to developing countries having reached approximate balance on current account in 1985.

    Other Resources Citing This Publication