PART III. ISSUES IN AGRICULTURAL COMMODITY MARKETS

Price developments for basic food commodities
Oil prices and agricultural commodity prices
Tariff peaks in agricultural markets and tariff cutting formulae
Recent trends in deficits and surpluses in basic food commodities in Africa
Impact of the Uruguay Round Agreements of relevance to the agricultural sector: Winners and losers

Price developments for basic food commodities

Introduction

The most striking feature of the developments observed in the international markets of basic food commodities during the last decade is the prominent price spikes that occurred for most of them in the mid-1990s (see Figure 3.1)[5]. Although, the downward trend since then has been reversed for dairy products from 1999, the prices of cereals, meat and oilcrop products continue to be depressed at levels that have not been seen for nearly two decades. Indeed, as measured by the relevant FAO indices, the average annual changes in the international prices since 1995 were -8.8 percent for major grains, -6.6 percent for rice, -6.6 percent for oilseeds, -3.4 percent for meat products and -3.5 percent for dairy products[6].

Figure 3.1: Annual FAO price indices of major food and feed commodity groups

Such strong convergence of price developments during the 1990s[7] is an indication of the commonality of the underlying fundamentals influencing the markets of various basic food commodities, as well as of the strong linkages existing between these markets. Although there were developments unique to each commodity market, these were either exerting pressure that tended to influence the markets in the same direction or not strong enough to overcome the effects of those that were working to influence the markets in the same direction.

Grain markets

Over the past decade, world wheat and coarse grain sectors have witnessed several important developments, some with major implications for international prices. As indicated below, changes in the pattern of trade (i.e. the changing composition of the main importing and exporting countries) and developments in the domestic cereal market of the United States and China have had the most direct impact on the global cereal economy, in general, and on the prevailing depressed state of grain prices, in particular.

Prior to the 1990s, the sharpest year-to-year (as well as intra-year) price swings were triggered by the presence or absence of two dominant, and yet unpredictable, buyers on world markets, namely the former USSR and China. The breakdown of the USSR not only eliminated the world’s largest grain importer from the international scene but also resulted in a significant expansion in unsold grains (hence large inventories) in major exporting countries, which continued to overhang the market through 1995 as other buyers did not fill the void. Between 1990 and 1995, major exporters made greater recourse to export discounts and subsidies in order to reduce their large stocks.

By 1995, however, they had also began reforming their domestic policies in order to deal with chronic problems of over-production. After the sudden price surge in 1995/96, caused by a significant fall in global cereal production, mainly due to weather problems in major exporting countries, export subsidies were no longer necessary and a sharp draw-down of stocks in major exporting countries raised expectations of continued strong prices, a factor which encouraged farmers to produce more grains in the following year. At the same time, a number of net-import producing countries such as Pakistan and India also adopted policies that were successful in encouraging production. In 2000 both India and Pakistan harvested record crops. As a result, both countries shifted from being net importers to net exporters often at relatively low prices which dampened quotations on the world market.

Against the background of strong prices and in compliance with Uruguay Round commitments, the 1995 Farm Bill in the United States, known as the Federal Agricultural Improvement and Reform (FAIR) Act, came into force in April 1996. While the FAIR Act aimed at removing (or decoupling) the link between support payments and farm prices by replacing deficiency payments by direct compensatory payments, the continuing support to the United States farmers in the form of emergency aid, on the top of the compensatory payments, provided enough incentive to farmers to continue growing grains even though the market signals, i.e. prices, were not encouraging.

In the meantime, China, a major wheat importer, continued on its policy to increase domestic production and reduce its dependence on imports. China’s grain policy since the mid-1990s encouraged higher production year after year so that by the late 1990s the country needed to import only small amount of grains. Following a short-lived interval in the mid-1990s, China’s maize production began to exceed domestic demand and this gave rise to a period of export boom. However, driven by the high financial burden of rising inventories along with the country’s bid to join WTO, China started to reform its grain economy, reducing production incentives and lowering its large stocks[8]. These changes resulted in a decline in plantings already in 2000. Beside lower plantings, weather problems also affected production in 2001, which fell well below consumption requirements. However, the shortfall was largely met through the release of stocks rather than by imports. In addition, China continued to export maize, although the decline in maize production was even more substantial. The absence of China as a major wheat importer and its continuing maize sales into the world market have also contributed significantly to the to put downward pressure on international prices.

Rice market

The sharp increases in prices between 1994 and 1996 reflected a sudden rise in import demand by countries that had experienced serious crop shortfalls in that period, including Bangladesh, China, Indonesia and especially Japan, which emerged as the top rice importer in 1994. As a result, the global volume of trade rose substantially and in 1995 surpassed the 20 million tonnes benchmark for the first time. The tightening of the international rice market revived food security concerns among large rice producing/consuming countries, and encouraged Governments to adopt more expansionary production policies. This new policy stance, together with favourable growing conditions, was responsible for the subsequent increases in global paddy production, which reached a new record in 1999.

For many of those countries that had embarked in supportive production policies, gains in output in the second half of the decade largely exceeded growth in domestic requirements, leading to the formation of large surpluses that had to be carried over the next years. As a result, global rice stocks rose substantially, especially at the close of the 1998 and 1999 seasons. Moreover, several of the countries that had been major importers in 1993-96 either cut imports or achieved self-sufficiency and some reverted back to a net exporting position. Purchases by China, for instance, which had soared in 1995 following a very poor 1994/95 harvest, were curbed substantially in 1997, as the country returned to the market as an exporter. In 1998, its exports climbed to 3.8 million tonnes and, have since then remained close to 3 million tonnes, even in 2000 under conditions of falling production, as the country could draw from its very large rice inventories.

While the faltering global import demand and the pressure from large export availabilities had a strong depressing effect on prices in the late part of the 1990s, the high level of protection that characterizes the rice sector did not allow the fall in prices to be fully transmitted onto domestic markets in major producing countries. Instead, Governments continued to shield their country’s rice sectors from falling international prices, by reducing import competition or by promoting exports, either through government-to-government deals, barter arrangements, or credit programmes, all of which tended to exacerbate the drop in prices in 2000. To some extent, commitments taken under the Uruguay Round Agreement on Agriculture hindered the ability of countries like India, the Republic of Korea, Japan or the EC to dispose of their excess supply on the world market other than in the form of commercial sales or food aid. Such constraints did not apply to other major rice market players, such as China, Myanmar or Vietnam, which were not WTO members. However, there is no formal evidence to suggest that these countries subsidized their exports.

Figure 3.2. Monthly FAO index of international rice prices

There are signs, however, that may point to a price recovery soon. Since 2000, global rice production has fallen short of global rice consumption, requiring stocks to be released in a number of countries, especially China. Moreover, a number of countries more exposed to international competition (Argentina, Uruguay) have reacted to the low prices by cutting plantings. In others, Governments are embracing less supportive production policies, especially China, but also Egypt, Japan, the Republic of Korea, Pakistan and Vietnam. Finally, drought conditions currently prevailing in various countries might also support a firming of prices.

Markets for oilcrop products

During the nineties, price developments in the oilseeds complex manifested the usual complex and diverse pattern (see Figure 3.3). Overall, periods of more stable prices alternated with periods of strong price movements. Until the end of 1992 and again between 1996 and mid 1999, prices for oils and meals - linked to one another through the crushing of oilseeds - moved more or less in parallel. By contrast, there have been two exceptional spells, 1993-1995 and 1999-2000, where the oil and meal price indices moved in opposite directions. The market fundamentals underlying price developments during these two periods are outlined below.

In the second half of 1993, international oil prices started rising due to shortfalls in global oil supplies, and decreasing inventories. International oilmeal prices, by contrast, weakened, primarily because of depressed import demand. This situation persisted throughout the following year, when a further tightening of oil supplies and markedly reduced stock levels led to further sharp increases in oil prices, while meal prices remained low due to continued weak import demand particularly in the EC, where agricultural policy changes favoured the use of grains over oilmeals in feed rations, and in Eastern Europe and the CIS, where domestic demand for and imports of oilmeals dropped as a result of low economic growth and of foreign exchange shortages. Eventually, oil prices decreased again in 1995 as, due to improved harvests of high oil-yielding crops, the tightness in global supplies eased and stocks in importing countries were replenished. At the same time, however, meal prices started to climb due to a combination of factors, including a fall in global soybean output, sustained global demand for livestock products and an increase in demand generated by a more favourable meal/grain price ratio following the marked increase in feed grain prices in 1995.

Figure 3.3. Monthly FAO indices of international prices of oils and meals

Toward the beginning of 1998, meal prices, and later also oil prices, came under intense downward pressure. This was mainly the result of a strong recovery in global production of soybeans and other oilcrops (partly related to agricultural support policies applied in certain countries), combined with a slow-down in the expansion of global oil and meal consumption. In the case of oils, prices continued to fall sharply during 1999 and 2000, mainly for two reasons: further good harvests of high oil-yielding crops, and the full recovery of tropical oil production from the adverse effects of El Niño in 1997-98 - both leading to a marked increase in inventories in major exporting countries. Initially, also meal prices continued falling as world soybean production and global oilmeal stocks expanded, and because competition from feed grains increased again following a conspicuous drop in their prices. However, as opposed to oil prices, from mid 1999 onward, international prices for cakes and meals recovered, mainly because the expansion of global oilmeal production came to a halt, with soymeal supplies actually falling short of demand, thereby reversing the situation observed since the end of 1997.

Markets for livestock products

Meat

International meat prices have exhibited a sharp decline during 1997-98 (see Figure 3.4). There were many factors influencing this. They ranged from: the effects of animal productions cycles, particularly for beef; and changing composition of meat trade, caused by changes on the demand side; to high degrees of concentration of international meat imports, making international markets susceptible to shocks occurring in any one of the major importing countries.

Figure 3.4. Monthly FAO index of international meat prices

The analysis of prices in the meat sector, when compared to those of the crops sector, is further complicated by the heterogeneous nature of meat products that currently enter international markets. This makes it difficult to identify representative international prices, which, with the rapidly changing structure of international trade, makes a general meat index quickly lose its relevance for tracking market developments. Keeping these caveats in mind, Figure 3.4 indicates that meat prices have often not followed the same path as those of basic food crops.

An important component of the FAO meat index is the price of beef products, and, hence, the decline in beef prices since the mid-1990s has been an important contributor to the decline in the overall index. Herd liquidation in the United States, the world’s largest importer of beef, beginning in 1996 because of the sharp increases in the prices of feed grains and subsequently oil meals, resulted in reduced import demand and thus contributed to weaker international beef prices. Meanwhile, a stagnant economy in Japan, the second largest import market, led to consumer preferences shifting to lower-value imported beef cuts, such as short plate, brisket, and chuck, adding further downward pressure on average beef prices. This declining trend tentatively reversed itself in 1999 in response to recovery in Asian markets and indications of herd rebuilding in the United States, both of which turned out to be short-lived, maintaining the international prices at their new lows, aided by the re-emergence of the BSE crises in Europe later on.

Additional downward pressure on international meat prices after 1997 was generated by the financial crisis in Russia, in 1997 the world’s largest meat importer and the recipient of more than a quarter of world trade in poultry meat. International poultry prices dropped sharply as a result. Moreover, currency devaluations in Thailand and Brazil, two major poultry meat exporters, contributed to the downward pressure on poultry prices. However, the gradual stabilisation of the Russian economy and animal disease outbreaks in late 2000 and 2001, affecting the availability of beef and pigmeat, could lead to a significant recovery in poultry prices. Similarly, lamb prices, after jumping precipitously in 1996 and 1997 in response to BSE concerns by consumers in Europe, the largest import market for ovine products, are witnessing significant price strength in 2001.

The structure of the global pigmeat market is relatively thin, with trade constituting approximately 3 percent of global production. In addition, it is highly concentrated-two-thirds of world trade determined by only three different import and exports markets - thus making the overall market more vulnerable to individual country market shocks. This was evidenced in 1998 when escalating production in exporting countries at the time of the financial crisis in Russia-the largest meat import market led to a precipitous drop in demand for pigmeat imports, resulting in pig prices in the United States and the EC dropping to 30 year lows. Pigmeat prices have remained low since, probably as a result of restructuring in pigmeat industries in the United States and the EC, with record low prices forcing smaller producers out of business. Lower marginal costs of production as companies increased their size and scale of operations may be allowing product to be sold at lower prices. Some shifting of product preference to lower-value cuts in Japan, may be also maintaining some downward pressure on prices.

Dairy Products

Over the past decade, international prices for dairy products have shown considerable variation (see Figure 3.5). During this period, import demand grew by approximately a third and rose from 5.5 percent to 7.0 percent of world milk production. However, despite the increase in world demand, the amount of milk traded, in relation to total production, remained relatively small. On the export side, the international market for dairy products is characterised by a limited number of exporting countries. While the number of importing countries is greater, a limited group of countries account for a substantial proportion of imports.

Figure 3.5: Monthly FAO index international dairy prices

As a result of the above, relatively small changes in supply or demand in the main exporting and importing countries have been sufficient to produce substantial changes in international prices. Thus, during the latter part of the 1990's, economic adjustments in Asia and the devaluation of the Russian ruble and the Brazilian real were important factors in the lower prices seen during this period. The more recent rise in dairy prices reflects: improved economic conditions in some of the major importing countries; for petroleum-producing countries, increased revenue to fund imports; domestic market conditions in the EC, which led to the reduced availability of surplus dairy products for export; and commitments to reduce the volume and value of subsidised exports arising out of the Uruguay Round Agreement on Agriculture.

Concluding remarks

The prominent spikes in the international prices of most basic food commodities observed in the mid-1990s were essentially due to unfavourable weather conditions that curtailed production in large parts of the globe and led to substantial declines in carryover stocks of the commodities concerned. Subsequent “spill-over” changes were triggered by these spikes, in part due to the strong linkages among the various commodity markets, with, for example, the increase in the prices of feed grains affecting both the meals and livestock sectors. Such widespread climatic influences on production though infrequent, are not exceptional. With declining global inventories of some major food commodities, especially in major exporting countries, moreover, their impact on food security of vulnerable countries may give rise to serious concerns as financial instruments to cope with such sharp shortfalls may not be the best way of coping with physical unavailability of food supplies.

Changes in economic performance in countries in different parts of the world and policy responses at the national level have been superimposed on these to complicate the assessment of the developments over the latter half of the decade. Output increasing effects of price increases combined with relatively slow economic growth in many parts of the world have weakened international prices of food commodities since the mid-1990s, a phenomenon that has been facilitated by increasing integration into the global markets of even the remotest countries. With increasing pressure to reduce accumulated domestic inventories in some countries and policy responses in others to counter the effects of lower returns, the downward pressure on international prices have continued well into the beginning of the new millennium.

It should also be noted that this period has also seen the beginning of the implementation of the Uruguay Round of Agreement on Agriculture. Although the members party to the Agreement have generally kept to their commitments, which however did not lead to substantial changes in policies, the developments described above in the market fundamentals have made it difficult to isolate the contribution that the Agreement has made to those developments.

Oil prices and agricultural commodity prices

Introduction

The issue of whether oil price movements have a significant influence on the markets and prices of internationally-traded agricultural commodities has been discussed in a number of international fora, most recently in the sessions of the FAO intergovernmental groups dealing with natural fibres. However, no clear conclusions have emerged concerning whether oil price changes do have an impact on the prices of other commodities and if so, what form - positive or negative - they might take. It seems likely that if there are linkages, these will vary from commodity to commodity and from country to country. The dynamics of such relationships are also of interest: given the secular downward trend in the prices of many commodities, it may be that any oil price effects are confined to the short-run. These are essentially empirical questions, but there has been little economic analysis of possible linkages to provide empirical evidence either way. This article discusses the nature of possible relationships between oil prices and agricultural commodity prices, and presents the results of some preliminary econometric analysis which tries to test for the existence and significance of any such effects in relation to fibres.

Oil prices are currently (October 2001) averaging just over $20/barrel, some 30 percent lower than one year ago. The 2000 oil price shock appears to have been only temporary: after reaching a peak of more than $30/barrel in September and November of that year, prices have fallen steadily, averaging $28 in 2000 and $25 so far in 2001. While the current international situation makes forecasting particularly problematic, prices had been expected to average $18-19/barrel in the longer term. It should be remembered, however, that the final price of oil is not simply dependent on oil supply and demand, but also includes national taxes and subsidies. For example, in the EC taxes account for two thirds of the retail petrol price. Compensatory changes in national taxes might reinforce international oil price movements by maintaining demand in the face of rising prices.

The macroeconomic consequences of oil price shocks have been widely discussed. While price hikes may be beneficial to the oil-exporting countries, it is generally believed that the effect on overall world output and aggregate demand is adverse. Industrialised countries have succeeded in raising the efficiency of their energy use, partly in response to higher oil prices, but also as a result of environmental concerns, such that oil price shocks are less severe than in the past. The consequences for oil-importing developing countries are more severe, since developing countries tend to use more energy per unit of output - up to twice as much - and this has increased further with industrialisation. The oil import bill therefore generally accounts for a greater share of the current account balance. Meanwhile earnings from exports of many agricultural commodities have trended downwards as prices have fallen, exacerbating the negative impact of increases in oil import bills. However, if there are linkages between the price of oil and the prices of these agricultural commodities, the impact of oil price changes on the balance of payments of oil-importing developing countries may be reinforced or cushioned by the resulting relative price changes. The nature of any such linkages between oil prices and the prices of selected agricultural commodities is explored in this article.

Economic links between oil prices and agricultural commodity prices

This section considers the economic mechanisms through which variations in oil prices might be expected to impact upon agricultural commodity prices. These are multiple and complex, and as noted above, are likely to differ from commodity to commodity and from country to country. It is difficult therefore, a priori, to generalise concerning the nature of these linkages and impacts.

As with any other products, the costs of production and hence profitability of agricultural commodities are influenced by the prices to be paid for oil. The strength of these effects would depend on the oil intensity of production of the particular commodity concerned, but other things being equal, it might be expected that increases in oil prices would eventually be reflected in higher agricultural commodity prices.

Perhaps the most discussed impact of oil price changes is on the economic growth performance of the oil-importing countries, especially the industrialised countries which also provide major markets for traded agricultural commodities. An increase in oil prices is expected to lead to a slowdown in the rate of economic growth and hence in the demand for all commodities. As a result, commodity stocks rise and prices fall. The impact on economic growth is likely to be more severe for oil-importing developing countries, for the reasons given above, but any knock-on effect on commodity prices would be less pronounced since these countries are generally less significant as commodity importers. It is however possible that the reduction in domestic demand for tradeable commodities might lead to an increase in exportable surpluses, reinforcing downward pressure on prices.

For the oil-exporting countries, the effect of an oil price rise on demand is obviously the reverse of that for oil-importers. Increasing oil revenues stimulate an expansion of import demand for agricultural commodities, which for certain products could conceivably lead to an increase in their price. Oil price variations may at least have an impact on the pattern of trade. Whether a commodity price effect would actually materialise would depend on the importance of the oil-exporter as an importer of a particular commodity, and the magnitude of the income elasticity of import demand. Changes in the volume of imports by certain oil-exporting countries, notably in the Near East and the Area of the former USSR, can be a major source of price variations for certain commodities. In the case of tea, for example, the Russian Federation is the world’s largest importer and Near Eastern markets are also significant. The volume of imports into these markets appears to be highly income elastic, and to have a significant influence on international tea prices in the short-run.

A potential direct link between oil prices and agricultural commodity prices arises where oil-based synthetic products compete with natural products. Agricultural raw materials markets are the most obvious example, where natural fibres and rubber face generalized competition from synthetic alternatives. Of course, substitution is based on the overall competitiveness of the alternative products in a given use, and this involves not only simple price comparisons but also consideration of product characteristics relevant to the intended use. A rather more complex substitution-based linkage with oil prices arises in the case of sugar in Brazil through use of sugar in ethanol production. As oil prices have risen, increasing amounts of sugar have been diverted for ethanol, and so there might be upward pressure on domestic sugar prices. However, whether this would spill over into an impact on world sugar prices would depend on the supply response of Brazilian sugar. It is possible that this might be asymmetric, with rising oil and sugar prices leading to an expansion of productive capacity which would not be completely reversed at least in the short-run when prices fall.

The overall impact, if any, of an oil price change on agricultural commodity prices would depend on the balance between these various channels of influence. Effects arising through possibilities of technical substitution of natural for synthetic products, or through the costs of commodity production would be expected to be positive. The influence on commodity demand via income changes is less clear. In general terms, changes in the level of economic activity as a result of oil price changes would lead to agricultural commodity prices moving in the opposite direction to oil prices. However, in specific markets, namely the oil-exporting countries, certain agricultural commodity prices might move in the same direction. Whether this latter effect would outweigh the former in influencing international prices would depend upon the importance of the oil-exporting countries in the world market for the commodity concerned.

Quantifying the impact of oil prices: some preliminary empirical evidence

The previous section outlined possible channels through which changes in oil prices might impact upon agricultural commodity prices. However, it is an empirical question whether these linkages exist in practice, and if they do how significant they might be. Given the diversity of the possible linkages it is likely to be difficult to obtain unequivocal results in most cases. This section discusses some results from a preliminary econometric analysis. Further technical detail and a full set of results are available on request from the Raw Materials, Tropical and Horticultural Products Service in the Commodities and Trade Division of FAO. The analysis focuses on fibres where oil price changes might be expected a priori to have the most direct effect through changing the relative prices of the natural and synthetic products.

The analysis seeks to establish whether there are statistically significant relationships between oil prices and fibre prices[9]. Two econometric approaches are used: the first is an application of cointegration analysis to test for the existence of any stable long-run equilibrium relationships between oil and fibre prices; the second is a test of whether oil prices add explanatory power to structural econometric models of the determination of international fibre prices.

Cointegration analysis focuses on possible long-run relationships between the price series for oil and the various fibres. Specifically, it tests whether there is some stable long-run or equilibrium relationship between the prices which will tend to be restored following any deviations from it, and analyses the adjustment process towards restoration of the long-run relationship. In statistical terms, the existence of such a relationship would be revealed by the satisfaction of two conditions. The first is that the price series concerned should have similar properties, such as similar trend behaviour. The second is that the price series should move together through time with no tendency to depart in any systematic way from their long-run relationship. Various statistical testing procedures are available in relation to each of these conditions. However, in many instances these tests have low power and can give inconclusive or even conflicting results.

The results of the cointegration analysis are mixed. All the individual price series appear to share the same basic statistical properties in terms of trend behaviour so the first condition appears to be satisfied. However, results of tests for the existence of long-run relationships between prices are conflicting, depending upon the tests applied and whether quarterly or monthly data are used. The Johansen test procedure, which focuses on estimates of the long-run relationships between oil and fibre prices, indicates no such long-run relationships, and hence no price transmission existing between oil prices and each of the fibre prices, including polypropylene. However, an alternative test procedure which focuses on analysing how fibre prices adjust to oil price changes so as to restore a long-run relationship between them indicates that fibre prices do adjust to oil prices in this way. Using quarterly data, this result holds for each of the fibres considered. The adjustment coefficient, which measures how fibre prices adjust to restore the long-run relationship with oil prices, are highly statistically significant, although they are small in magnitude indicating that adjustment is relatively slow, and that oil price changes are not fully propagated to fibre prices. However, repeating this analysis using monthly data, the results indicate that only cotton prices adjust to oil prices in this way. Again, while a long-run relationship appears to hold, there appears to be only partial transmission from oil to cotton prices and the adjustment of cotton prices to oil price changes is slow.

The alternative approach to analysing whether oil prices have a significant impact on fibre prices uses simple dynamic econometric models of the determination of fibre prices. Such models are widely-used in commodity market analysis. They relate each fibre price to its recent history, and the levels of demand and stocks. These models were estimated for cotton, jute and sisal, since quantity data for polypropylene could not be obtained. The explanatory power of these models is acceptable and the estimated coefficients are statistically significant. A test of whether oil prices also have a role in determining fibre prices can be made by adding oil prices to the models, and then testing whether the estimated oil price coefficient is statistically significantly different from zero. If so, this implies that oil prices do make a significant net contribution to explaining movements in fibre prices. In the case of cotton and jute, it appears that oil prices do not provide any significant improvement in the explanatory power of the models. However, according to these results, they do appear to make a significant contribution to the determination of sisal prices.

Do oil prices affect agricultural commodity prices?

While in theory there are a number of economic relationships through which oil price changes might impact upon agricultural commodity prices, the empirical analysis gives only limited support to the view that any such impacts might be significant. The econometric analysis undertaken is, of course, preliminary, and confined to a small number and range of commodities. Nevertheless, it is in relation to natural fibres that some of the most direct and perhaps consequently some of the strongest effects might be expected. Some significant effects are indicated, but the evidence is not consistent. Overall, the results do not support the view that oil prices are a regular and major influence on fibre prices. A number of reasons might be put forward to account for this.

Key to the hypothesised link between oil prices and fibre prices is the assumption that oil price changes lead to changes in synthetic fibre prices, and hence in their relative prices vis-à-vis natural fibres. However, there does not appear to be a strong stable relationship between oil prices and polypropylene prices, for example. This may reflect competitive conditions in the industry which mean that oil price changes are not fully reflected in polypropylene prices. While substitution in use provides a channel through which changes in oil prices may impact on agricultural commodity prices, the significance of any such impacts which change the relative prices of the natural and synthetic products is not clear a priori, since the wider aspects of competitiveness need to be taken into consideration. The latter may have more of an impact on the longer-term trends in substitution. In the case of jute, for example, the introduction of polypropylene sacks and bags made substantial inroads into the market share of jute sacks and bags not only because the polypropylene product was cheaper, but also because in many uses it was seen as offering superior product performance. Polypropylene is also challenging jute for market share in carpet backing, although it is not necessarily a cheaper alternative. Furthermore, the extent of any substitution, whether in response to relative price changes or a re-evaluation of relative technical performance may be limited by the technical and economic possibilities for switching between alternatives, and the associated adjustment costs.

It may also be that the various possible effects of oil price changes may be offsetting so that any positive effects arising from substitution may be offset by negative effects arising from the influence of oil prices on overall commodity demand. However, as noted above, downturns in economic activity in the industrialised countries as a consequence of oil price hikes are nowadays less severe than in the past. Such generalized demand-side effects might therefore be limited, especially against the background of longer-term commodity price trends resulting from growth in commodity production and stocks, substitution by synthetic alternatives, and simple consumer preference shifts. The more generalized potential impacts of oil prices on agricultural commodity prices operating through the demand for those commodities would presumably be more diffuse. For certain commodities where oil-exporting countries are significant importers in terms of market share the impact on demand and prices of changes in oil revenues might well be significant, and certainly this would be worth quantitative investigation. However, the preliminary analysis of fibre markets presented here suggests that a statistically significant link between oil prices and fibre prices cannot be unequivocably established, and that if such links do exist they are weak.

Tariff peaks in agricultural markets and tariff cutting formulae

Introduction

Under the market access provisions of the Agreement on Agriculture (AoA), WTO members agreed to convert their non-tariff import barriers into equivalent tariffs. This was done through a process known as “tariffication”[10]. In addition, developing countries were given the option of adopting ceiling bindings for tariffs that were not previously bound. The Uruguay Round (UR) reduction commitments established that developed countries would reduce their tariffs, including those resulting from tariffication, on a simple average basis of 36 percent (24 percent for developing countries) with minimum cuts of 15 percent (10 percent for developing countries) per tariff line implemented in equal annual instalments up to year 2001 (year 2006 for developing countries). No reduction requirements were established for the least developed countries.

This paper analyses the current market access situation in agricultural markets. Specifically, it identifies those commodity groups in the food and agricultural sector for which tariff peaks prevail, and examines how the implementation of alternative tariff cutting formulae could further affect those tariffs. The study uses data available from the Agricultural Market Access Database (AMAD)[11], and covers the 1995-98 period. For some countries, where available, 1999 data are also taken into account. The countries included in the AMAD database are the developed and developing members of the World Trade Organization (WTO) that scheduled tariff rate quotas (TRQs) and those additional developing members that listed tariff reduction commitments on a tariff line basis in their Uruguay Round schedules.

The incidence of tariff peaks

Tariff peaks are defined here as rates above 20 percent ad valorem[12]. Table 3.1 shows that high bound and applied tariffs are still common in global agricultural markets, even after the implementation of the UR AoA commitments. Moreover, the frequency of heavily protected tariff lines, bound and applied, is still significant for many commodity groupings. In the developed economies the highest tariff peaks are in the oilseeds, dairy, meat and wheat sectors. In the case of developing countries, the most prominent peak levels are in the meat, oilseeds and coarse grains sectors. It should be recalled, however, that this analysis on the basis of most favoured nation (MFN) tariffs does not take account of preferential market access that is frequently extended to developing or least developed countries.

Table 3.1. Tariff Peaks in selected Developed and Developing Countries, 1995-98 (percentages)

Commodity Group	Average Peak Tariff				Average Number of Tariff Lines above the 20% Peak		Percentage of Countries with Tariff Peaks
	Bound Rate		Applied Rate
	Developed Countries	Developing Countries	Developed Countries	Developing Countries	Developed Countries	Developing Countries	Developed Countries	Developing Countries
Bovine	192	83	123	49	18	13	65	60
Ovine	134	81	111	69	20	13	45	40
Pork	168	73	100	44	31	21	60	50
Poultry	140	73	129	50	34	30	65	55
Other meat	90	53	49	37	9	8	45	45
Dairy	153	79	119	35	69	35	75	70
Coarse Grains	124	81	93	44	18	22	60	55
Rice	123	61	71	35	16	7	50	60
Wheat	139	75	127	41	11	11	60	60
Oilmeals	31	68	23	41	4	5	15	25
Oilseeds	208	77	179	52	19	9	50	30
Vegetable Oils	107	57	90	39	15	32	70	50
Sugar	83	70	75	36	14	11	70	70
Fruits & Vegetables	120	51	110	33	161	176	70	55
Cocoa	117	43	86	26	15	7	60	60
Coffee	70	54	20	32	1	5	15	40
Tea	95	77	23	50	2	3	15	40
Tobacco	70	84	61	56	8	10	50	55
Cotton	30	62	29	45	3	2	10	5
Hard Fibres	55	100	53	32	3	3	10	5
Hides & Skins	48	58	32	36	15	7	10	15

Source: AMAD and FAO. Includes only countries in the AMAD database

The tariffication process under the UR, which disciplined the use of non-tariff measures, often resulted in higher bound tariffs than those prevailing before the Round. This fact has allowed for the persistence and even the increase of peak applied tariffs. In the developed economies, the application of the special agricultural safeguard clause has also led to additional duties in the beef, sugar, poultry and dairy sectors (UNCTAD, 2000).

Tariff cutting formulae and tariff profiles

A variety of tariff-cutting formulae have been proposed over successive rounds of multilateral trade negotiations. This section examines the potential effects of applying alternative tariff-cutting formulae to the identified bound peak levels. The objective is to assess how application of alternative formulae would affect the tariff profile for agricultural products, including the prevalence of peak tariffs. The tariff-cutting formulae utilized in this analysis were selected on the basis of tariff reduction proposals that were presented in the previous rounds of GATT negotiations, and that have already been compiled and analysed in a systematic way (Wainio et al, 2000; Laird, 1999[13]). Although not all of these formulae have been put into practice, some of them, such as the Swiss formula, have been subject to intensive debate in academic and professional circles during the post UR period.

The various alternative tariff-cutting formulae which have been discussed or applied in the context of previous trade rounds are defined in Table 3.2. These include: average linear tariff cuts as agreed in the AoA (36 percent for developed countries and 24 percent for developing countries)[14]; 50 percent linear cuts as applied to industrial goods in the Kennedy Round; four formulae proposed during the Tokyo Round, including 40 percent linear cuts; and the compromise Swiss formula finally adopted in that Round. Harmonising formulae, such as the Swiss formula, impose proportionately higher cuts on higher tariffs. However, their effect depends critically on the value of the formula coefficient - the maximum tariff which is to be permitted, and this would obviously be subject to negotiation in any actual trade round. In the Tokyo Round this was set at 16 percent, but only for industrial goods.

Table 3.2. Tariff Cutting Formulae (t₀ is the original tariff and t₁ the revised)

Cutting Formulae	GATT Round	Mathematical Expression
UR linear cut	Uruguay	t1 = t0 * (1 - 0.36) or t1 = t0 * (1 - 0.24)
50% linear cut	Kennedy	t1 = t0 * (0.5)
40% linear cut	Tokyo	t1 = t0 * (1- 0.4)
Linear cuts plus selective one step reduction	Tokyo	If t0 > 40%, t1 = 20%; otherwise, t1 = t0 * (0.5)
Iterative cuts	Tokyo	If t0 > 50%, t1 = 0.5 (1 - 0.5); applied three times Otherwise, t1 = t0 (1 - t0); applied three times
Linear cuts with harmonization	Tokyo	t1 = t0 * (0.3) + 3.5%
Swiss formula	Tokyo	t1 = (a* t0)/(a + t0); a is a parameter = 0.16

Source: USDA and S. Laird.

Figure 3.6. Examples of tariff reductions resulting from two tariff cutting formulae

Although in the on-going negotiation process no formal proposals have been adopted so far, current negotiating proposals are similar to the ones simulated in this study. Current tariff cutting proposals even include requests-and-offer tariff cuts on product by product and case-by-case basis (e.g. Switzerland, Japan, Norway) and zero-for-zero sectoral negotiations for selected products (e.g. Canada, US). However, most reduction proposals rely on formulae that are applied over a broad range of products. These tariff cutting formulae vary from a replication of the UR tariff cut modalities (e.g. EC, Poland) to the application of harmonising formulae to eliminate tariff peaks and escalation (e.g. Cairns Group, CARICOM, African Group).

Figures are given in Table 3.3 for the tariff rates that would result after the application of each of the formulae listed in Table 3.2. The first three columns show the effects of linear tariff cuts of 36 percent or 24 percent (Uruguay Round), 50 percent (Kennedy Round), and 40 percent (Tokyo Round). The final column of Table 3.3 shows the effects of application of the Swiss formula with a coefficient (maximum permitted tariff) of 16 percent as agreed in the Tokyo Round.

Perhaps the most notable finding of this analysis is that a repetition of the cuts implemented under the UR would leave significant tariff peaks remaining in all commodity groups, both for the developed and for the developing countries. The application of more aggressive linear cuts, such as those proposed for industrial goods in the Kennedy (50 percent) and the Tokyo (40 percent) Rounds, would lead to lower protection levels, although significant bound tariff peaks could still persist in most agricultural sectors. This suggests that linear cuts, even at higher levels than those implemented in the UR, would not eliminate tariff peaks. Only the application of harmonizing formulae, such as the iterative cuts formula or the Swiss formula, would effectively reduce tariff peaks, both for the developed and the developing countries.

However, it should be noted that the results of the application of the Swiss formula depend on the formula coefficient, which would need to be defined. As noted above, the results in Table 3.3 are based on the Tokyo Round application of the Swiss formula, with the coefficient set at 0.16, so all tariffs above 16 percent are driven below this level. This is perhaps an extreme example in the case of agricultural products. However, this example illustrates that such a formula is well suited for eliminating tariff peaks, as they are defined in here, and that it compresses tariffs within a narrow range without necessarily imposing deep cuts in tariffs below the formula coefficient. Therefore the results presented here reinforce those proposals that aim to reduce tariff peaks, showing that the application of harmonising formulae is the most effective way to achieve that objective.

Table 3.3. Resulting tariffs after the Application of Tariff Cutting Formulae to Selected Commodity Groups with Peak Tariffs in the Developed and Developing Countries (average percentage)

Commodity Group	Tariff Cutting Formulae
	UR formula		Kennedy Round formula		40% linear cuts		Linear cuts plus one step reduction		Iterative cuts		Linear cuts with harmonization		Swiss formula
	Developed Countries	Developing Countries	Developed Countries	Developing Countries	Developed Countries	Developing Countries	Developed Countries	Developing Countries	Developed Countries	Developing Countries	Developed Countries	Developing Countries	Developed Countries	Developing Countries
Bovine	123	63	96	42	103	50	19	18	15	15	61	29	13	12
Pork	107	55	84	36	86	44	19	18	15	15	54	25	13	12
Poultry	89	55	70	36	73	44	18	18	15	15	45	25	13	12
Dairy	100	60	78	40	86	48	20	19	15	15	50	27	14	13
Ovine	86	61	67	40	80	49	18	16	15	14	44	28	13	12
Other meat	58	41	45	27	54	32	18	17	14	14	30	20	12	12
Coarse Grains	79	61	62	40	74	48	19	19	15	15	41	28	13	13
Rice	79	47	61	31	74	37	19	18	15	15	40	22	13	12
Wheat	89	57	69	37	83	45	19	18	15	15	45	26	14	12
Oilmeals	20	52	15	34	18	41	15	19	14	15	13	24	10	13
Oilseeds	133	59	104	39	125	46	20	19	15	15	66	27	14	12
Vegetable Oils	69	43	54	28	64	34	18	17	15	14	36	20	12	12
Sugar	53	53	41	35	50	42	19	18	15	15	28	24	13	12
Fruits & Vegetables	77	39	60	25	72	30	19	18	15	14	39	19	13	12
Cocoa	75	33	58	22	70	26	18	16	15	14	39	16	13	11
Coffee	44	41	35	27	42	32	15	16	13	14	24	20	11	11
Tea	61	58	48	38	57	46	15	17	13	14	32	27	12	12
Tobacco	45	61	35	40	42	48	18	16	15	14	25	28	12	12
Cotton	19	47	15	31	17	37	15	16	14	14	12	22	10	12
Hard Fibres	35	76	27	50	32	60	20	20	15	15	20	34	12	14
Hides & Skins	31	44	24	29	29	35	20	18	15	14	28	21	12	12

Includes only countries in the AMAD database of which 30 are developing and 16 are developed.
All tariffs in ad-valorem equivalent

Source: AMAD and FAO.

Laird, S. (1999). “Multilateral Approaches to Market Access Negotiations”, in Rodriguez, M., P. Low and B. Kotschwar, eds. Trade Rules in the Making Challenges in Regional and Multilateral Negotiation, Organization of American States, Brookings Institution Press, Washington DC, 1999.

UNCTAD (2000). The Post-Uruguay Round Tariff Environment for Developing Country Exports: Tariff Peaks and Tariff Escalation, (TD/B/COM.1/14/Rev.1), January 2000, UNCTAD, Geneva.

Wainio, J., P. Gibson, and D. Whitley (2000). Evaluating Alternative Formulas for Reducing Agricultural Tariffs, Economic Research Service, USDA, Washington DC, 2000.

Recent trends in deficits and surpluses in basic food commodities in Africa

Trade can be an important contributor to economic growth and food security in developing countries. In Africa, most countries rely on exports of primary commodities such as coffee, cocoa, cotton, tobacco, tea, copper, etc, to earn foreign exchange with which to finance development programmes, as well as to import foods they do not produce or to cover shortfalls in domestic food production. However, experience has shown that relying on one or two primary commodities for export, as most do, is highly risky because world commodity markets are prone to sharp price swings. One way of minimizing the impact of price instability is to diversify the range of exports and also to find new markets. Intra-African trade is relatively very small compared to the region’s trade with the rest of the world. In 2000, intra-African trade grossed $11 billion compared to $145 billion from trade with the rest of the world, or a share of only 7.6 percent. Intra-regional trade can reinforce the economic complementarities among neighbouring countries and stimulate additional growth both in agriculture and related sectors. One potential area where intra-African trade could be promoted is trade in food commodities such as cereals, edible oils, livestock and livestock products.

The purpose of this note is to examine recent trends in surpluses and deficits in major, readily tradable food commodities in Africa, sub-region by sub-region, in order to explore potentials for intra-African trade. Trends in imports of rice and wheat, for which Africa is a net importer are examined. Trends in imports of other food commodities such as vegetable oils, animal products and sugar are also briefly summarized as well.

Deficits/surpluses in coarse grains[15] in 1999/2000 and 2000/2001

Maize and to a lesser extent sorghum and millet are the most important food crops traded in Africa. Root and tuber crops and plantains/bananas, though important staples in some countries, are hardly traded internationally (including within Africa) due to their high transport cost relative to their unit value. Table 3.4 presents estimates of cereal surpluses and deficits in 1999/00 and 2000/01, the latest two marketing years for which full data are available. In 1999/00, all the sub-regions were in deficit, whilst in 2000/01 southern Africa showed a surplus of 651 000 tonnes, accounted for by only three countries, namely South Africa, Malawi and Zambia. However, in marketing year 2001/02[16], this sub-region is likely to be in deficit, as grain production has declined significantly in most countries, including the sub-region’s leading grain producer and exporter, South Africa. North Africa (a non-producer) has the largest deficit, while Central Africa (a minor producer, dependent mainly on roots and tubers) has the smallest. Whereas a number of countries posted a surplus in 1999/00 (Ethiopia, Sudan, Uganda in eastern Africa; Benin, Chad, Côte d’Ivoire, Ghana, Mali, Mauritania, Nigeria, Togo in western Africa; Madagascar, Malawi, Mozambique in southern Africa), only three countries in the whole African Region, namely Malawi, South Africa and Zambia, had a surplus in 2000/01.

In eastern Africa, the deficit increased by 70 percent from around 1.4 million tonnes in 1999/00 to 2.4 million tonnes in 2000/01, reflecting mainly the impact of a severe drought. Kenya, among the hardest hit, accounts for the largest share, 56 percent, of the total deficit in 2000/01.

Table 3.4. Africa: Coarse grains[17] surpluses/deficits in 1999/00-2000/01 (‘000 tonnes)

		1999/00	2000/01			1999/00	2000/01
Northern Africa		-7 661	-9 689	Western Africa		-275	-593
	Algeria	-1 501	-1 806		Coastal countries	-5	-166
	Egypt	-3 310	-4 220		Benin	25	-3
	Libya	-650	-710		Ghana	15	-45
	Morocco	-1 600	-2 100		Nigeria	10	-70
	Tunisia	-600	-853		Togo	10	-
					Côte d’Ivoire
Eastern Africa		-1 425	-2 425		Guinea	5	-8
	Ethiopia	153	-70		Liberia	0	-
	Sudan	93	-90		Sierra Leone	-10	-10
	Uganda	80	-50	Sahelian countries		-60	-30
	Burundi	-40	-81		Mali	-135	-427
	Comoros	0	-		Chad	50	-
	Djibouti	0	-3		Mauritania	18	-15
	Eritrea	-140	-90		Gambia	2	-15
	Kenya	-800	-1 364		Guinea-Bissau	0	-
	Rwanda	-185	-150		Burkina Faso	-5	-15
	Somalia	-60	-71		Cape Verde	-15	-80
	Seychelles	-6	-6		Niger	-30	-35
	Tanzania	-520	-450		Senegal	-70	-230
Southern Africa		-1 003	651	Central Africa		-85	-37
	Malawi	400	300		Cameroon	-78	-82
	Mozambique	150	-50		Equatorial Guinea	0	-10
	Madagascar	5	-30		Gabon	0	-
	Angola	-215	-394		Sao Tome	-2	-2
	Botswana	-185	-185		Central Afr. Rep.	-2	-2
	Lesotho	-155	-190		Congo, Rep. of	-4	-3
	Namibia	-84	-60		Congo, Dem. Rep.	-10	-5
	South Africa	-77	1 250			-60	-60
	Swaziland	-22	-40
	Zambia	-370	50
	Zimbabwe	-450	-

Trends in imports and exports of maize

In most of Africa, trade in coarse grains is almost synonymous with trade in maize. Although some trade in sorghum and millet takes place among the Sahelian countries, it is insignificant on Africa-wide scale. In eastern Africa, Sudan is the only major producer of sorghum and exports significant amounts in good years, mainly to the Near East. Thus, only trade in maize is reviewed in detail. The graphs below show trends in maize imports and exports during the period 1994/95-2000/01.

Imports and exports 1994/95-2000/01

Fig.1 - North Africa Maize Imports & Exports, 1994/95-2000/01

Fig.2 - North Africa Wheat & Rice Imports, 1994/95-2000/01

Fig.3 - East Africa Maize Imports & Exports, 1994/95-2000/01

Fig.4 - East Africa Wheat & Rice Imports, 1994/95-2000/01

Fig.5 - Western Africa Maize Imports & Exports, 1994/95-2000/01

Fig.6 - Western Africa Wheat & Rice Imports, 1994/95-2000/01

Fig.7 - Central Africa Maize Imports & Exports, 1994/95-2000/01

Fig.8 - Central Africa Wheat & Rice Imports, 1994/95-2000/01

Fig.9 - Southern Africa Maize Imports & Exports, 1994/95-2000/01

Fig.10 - Southern Africa Wheat & Rice Imports, 1994/95-2000/01

Maize Imports

North Africa is the largest importer of maize and consignments show a strong upward trend (Fig.1). Imports averaged over 5 million tonnes annually over the 7-year period (1994/95-2000/01). It would seem that North Africa could be a major market for sub-Saharan African maize, if large exportable surpluses were available and were competitive on the world market.

In eastern Africa, a rising trend in maize imports is evident (Fig. 3). These imports are largely driven by Kenya, which has imported an average of 745 000 tonnes per year during the 7-year period, out of a sub-regional average of 1.33 million tonnes. The dip in imports in 1998/99 reflected Kenya’s bumper harvest that year. Tanzania is a distant second (average 224 000 tonnes), followed by Rwanda (average: 169 000 tonnes). In western Africa, maize imports appear to have stabilized at around 300 000 tonnes annually since 1998/99 (Fig. 5). Sahelian countries account for then bulk of the imports (7-year average: 240 000 tonnes compared to 150 000 tonnes for coastal countries). This is rather surprising as sorghum and millet production in the Sahel has been at record to above-average levels since 1996. In Central Africa (Fig. 7), maize imports are on the increase, but the quantities involved are relatively small. In southern Africa (Fig. 9), the pattern is less clear, but largely reflects seasonal rainfall behaviour. For example, the high maize imports in 1995/96 reflected the impact of the severe drought in 1994/95. As already indicated, the sharp falls in maize production in various countries in 2000/01 are likely to result in a significant increase in the sub-region’s maize imports in the 2001/02 marketing year.

Maize Exports

In eastern Africa, the trend is downward (Fig. 3), exports having fallen from a peak of nearly 1 million tonnes in 1995/96 to an annual average of around 280 000 tonnes since 1997/98. Only Uganda has consistently exported maize over the period, although the quantities have been declining. Maize exports are mainly within the sub-region, but marketing problems arise for surplus producers when their neighbours become self-sufficient in good years, as the tendency is to ban imports. In western Africa, a downward trend is also apparent (Fig. 5), although over the last three years the average has hovered around 125 000 tonnes. The main exporters (albeit with relatively small quantities) are Nigeria, Benin and Mali. In central Africa, only Cameroon exports small quantities of maize, which have averaged about 10 000 tonnes per annum over the 7-year period. In southern Africa, the high level of maize exports (nearly 5 million tonnes) in 1994/95 was followed by a sharp drop (to about 400 000 tonnes) in 1995/96 due to drought. A modest recovery was posted in 1996/97, after which a downward trend was evident up to 1999/00. This downward trend is set to resume in 2001/02 after a relatively small increase in 2000/01, mainly due to mid-season dry spells that affected crops in several countries. The main exporters have been South Africa and Zimbabwe (the latter up to 1999/00).

Trends in wheat and rice imports

Figs. 2, 4, 6, 8 and 10 show trends in wheat and rice imports. All African countries are net importers of wheat and, with the exception of Egypt, also of rice. Wheat imports appear to be growing fastest in eastern Africa, having almost tripled between 1995/96 and 1999/00. In western Africa, rice imports have overtaken wheat imports since 1998/99 (Fig. 6).

Trends in deficits/surpluses in other traded food commodities

The other major food commodities traded extensively in Africa include edible oils, livestock products (meat and dairy) and sugar.

Edible Oils

Africa as a whole is a net importer of edible oils. The deficits have increased continuously over the 1990s, nearly doubling from 2.2 million tonnes in 1991 to about 3.8 million tonnes in 2000. The main deficit countries are located in northern Africa which, as a region, accounts for nearly 60 percent of the total deficit. However, Ethiopia, Kenya, Tanzania, South Africa, Nigeria and Senegal from the sub-Saharan region are also important importers, despite the fact that some of them are significant producers. The deficits vary considerably from year to year depending on domestic production and other factors. The main oils imported by Africa are (in order of importance) palm oil, soybean oil and sunflower oil.

The only potential net exporting sub-regions are western and central Africa, although west Africa seems to have turned into a net importing region since the mid-1990s, while central Africa's net exporting position depends on a single exporter, Cameroon (in palm oil).

Western Africa's main net exporters are Côte d'Ivoire (coconut oil, palm oil, palm kernel oil) and Ghana (palm oil). Côte d'Ivoire's surplus fell steadily during the mid-1990s but has stabilised in recent years. Minor net exporters include Mali (groundnut oil) and Benin (palm oil, palm kernel oil). Senegal used to be a net exporter, but became a net importer in the early 1990s.

Dairy Products

Africa as a whole, its sub-regions as well as individual countries, has been consiste