3. The International Rice Market

Eric Monke and Scott Pearson

This chapter explains why the level and instability of the world price of rice are both critical in shaping Indonesian rice strategy and difficult to forecast. For Indonesia, the level of the world price directly affects the desirability of expanding rice output, the budgetary consequences of rice trading when policy decisions cause domestic and world prices to differ, and the income distributional effects of price policy for rice. The instability of the world price influences the need for stabilization measures and the costs and benefits of BULOG's buffer stock program to stabilize domestic prices. The first section of the chapter briefly develops these arguments.

Subsequent sections review recent patterns of international rice trade and prices, focusing on the forces that determine world price level and variability. The small size of the world rice market contributes to a high degree of price variability because minor changes in the balance between production and consumption in individual countries translate into relatively large changes in world market availabilities or demands. The tendency toward instability is accentuated by government policies that influence the rice market. In a policy-dominated trading environment, variation in international trade is used to maintain domestic price targets. Most governments place great importance on holding domestic rice prices stable, and their changing positions in the world market tend to aggravate international price instability.

Indonesian planners and policymakers make many crucial decisions that require judgments about future patterns of world rice prices. But prediction of the world price is a hazardous exercise. The thinness of the world rice market means that world rice prices are difficult to forecast in the short run; small errors in forecasting production or consumption for individual countries become large mistakes in world market forecasts. To forecast long-run rice prices, market analysts have to contemplate future shifts in policies as well as changes in the behavior of consumers and producers. The final section of the chapter reviews plausible scenarios for the future.1

World Prices and Indonesian Food Policy

The insulation of domestic prices from highly unstable world prices is one of the principal goals of Indonesia's food policy. Information about world price instability aids policymakers in selecting the appropriate size of BULOG buffer stocks, necessary to achieve varying degrees of food security. The costs of maintaining these stocks need to be compared with the probability of high world prices during years of low domestic production because an alternative to buffer stocks is to increase imports and sell at subsidized prices. Variabilities in Indonesian production and world prices have been closely linked; Indonesia's rice production is more than twice as large as the amount of trading in the world market. Unexpected production shortfalls sometimes cause Indonesia to import, whereas unusually good harvests can result in occasional exports.

Price instability also has a large impact on the government budget. Budgetary consequences occur if the rice price in Indonesia departs from the world price. For example, Indonesia's domestic prices for rice were well above comparable world prices between 1982 and 1986; as a result, in 1985 and 1986 Indonesia was forced to pay very large export subsidies (about $150 per metric ton) to move surplus rice onto the world market. When domestic prices are below world prices, as in much of the 1970s, policymakers must worry about the costs of subsidizing imports. Given the desire for domestic price stability, the only way to avoid budget outlays is to remain self-sufficient-choosing rice and input prices that exactly balance the growth in production with that in consumption. The vagaries of climate, biology, and economic incentives outside of rice price policy make such precision impossible in most years.

The long-run world price also plays a critical role in investment and price policy. Long-run average world prices of rice are efficiency measures in evaluating investment projects (e.g., in irrigation, varietal research, and extension) or in policy analysis of input subsidies (e.g., on fertilizer, insecticides, and pesticides) that lead to increased rice production. The long-run average world price measures the opportunity cost of the planned increase in production because this rice will be either exported or substituted for imports of rice. If, for example, the world rice price were to approach the temporary peak it reached in 1981 (over $400 per metric ton for Indonesian qualities), Indonesia could expand rice production in most parts of the country and export the new output efficiently.

Indonesia's potential influence on world prices complicates the choice of an appropriate efficiency measure. The c.i.f. price acts as the efficiency cost of imports, while the f.o.b. price measures the efficiency value of exports. Normally, these prices differ only by the costs of international transportation and handling and are relatively small. But for Indonesia, c. i. f. and f. o. b. prices also are influenced by the magnitude of Indonesia's participation in the world market. If Indonesia is a large importer, world market prices are forced upward; if Indonesia exports, world market prices move downward. As a result, efficiency prices for the evaluation of import substitution (the c. i. f. price) and exportation (the f. o. b. price) will differ by more than just transportation and handling costs.

These differences will be larger in the short run than in the long run because of unused production capacity in several of the surplus-producing countries. Indonesia's effect on world prices thus will be more important for price variability than for price level. For example, if Indonesia were to import large quantities of rice, prices in the world market would increase immediately. After a short adjustment period, however, world exports would rise, bringing prices back down. During the last fifteen years, for instance, the internationally traded volume of rice has increased by about 50 percent, while world price levels have steadily fallen. The United States, Thailand, and Japan are examples of countries that are capable of increasing exports in response to increased world market demand in the future.

The long-run average world price also serves as the arbiter of the income distributional effects of domestic rice price policy. When domestic prices depart on average from world rice prices, income transfers occur among producers, consumers, and the government treasury. During the 1970s, the domestic price in Indonesia was usually less than the world price; as a result, Indonesian consumers gained, producers lost, and the treasury (through the food agency) had to subsidize rice imports. In the first half of the 1g8os, the domestic price exceeded the world price; consequently, policy caused consumers to pay more, while producers received higher incomes. The government collected import taxes through 1984 but paid export subsidies in 1985 and 1986 after self-sufficiency was achieved. Following the rapid rise in world prices during 1987-89, domestic and world prices were at similar levels. But if world prices return to their low levels of 1985 and 1986, expansion of production cannot occur without subsidies from consumers, the government budget, or both.

Recent Changes in World Trade and Prices

Rice import and export statistics for the 1969-86 period are reported in Table 3.1. During the 1g8os, world trade averaged about 12 million metric

ric tons annually, fluctuating between 11 and 13 million metric tons, a substantial increase over trade volumes of a decade earlier, when quantities traded varied between only 8 and 9 million metric tons. But the international market remains tiny in comparison to total production. Trade represents only 3.5 to 4.0 percent of output (in contrast, about 2o percent of wheat production is traded), and the ratio for rice has not varied much in recent years. Between 1974-76 and 1984-86, annual average world production of rice increased by 38 percent (from 343 million [paddy] to 472 million metric tons), a rate of growth comparable to that of the international rice trade.

Five countries-Burma, China, Pakistan, Thailand, and the United States-provided between 7o and 8o percent of total exports of rice during this period. Thailand had the most rapid growth and by 1979-81 had surpassed the United States as the world's largest exporter. In 1984-86 Thai exports exceeded 4 million metric tons, accounting for more than one-third of total export trade. United States exports have been as large as 3.1 million metric tons (in ig8i), but magnitudes have depended heavily on government subsidy programs. During the past decade, subsidized exports accounted for about one-third of total sales, primarily through the food aid (Public Law 48o) program. Recent Chinese exports are only about half of their level of the early 1970s, while Pakistan's exports increased nearly as much in relative terms. Burmese exports remain stable (except for 1988) in spite of high recent growth in production (a 2o percent increase between 1979-81 and 1984-86).

Asia accounts for over go percent of total production and consumption. Asian countries also provide most of the exports; the growth in Thailand's market share is primarily responsible for the increase from 54 to 66 percent in the Asian countries' export share between ig6g and 1988. On the import side of the market, however, Asian countries have become much less important over time. Asian importers have reduced their reliance on international markets, and equally dramatic increases in imports have occurred in countries of the Middle East, Africa, and the developed countries of the Northern Hemisphere.

In the bottom half of Table 3.1 regional aggregations are used to illustrate the structural change in import demand. With the disappearance of Indonesia from the import side of the market, no single country accounts for more than a few percent of total trade. The quantity data show the decline in Asian imports of nearly 3 million metric tons (from 5.6 to 2.8 million). At the same time, total world trade volumes increased by 3 million metric tons; imports by Middle Eastern and African countries each increased by about 2 million, and imports by the developed market economies (primarily Canada and Europe) grew by more than i million. By 1985, the Asian import share had declined from 63 to 24 percent, and the Middle East, Africa, and the developed countries (including the developed centrally planned economies) had each become as large.

In Asia, the declines in rice imports mirror production growth in a number of low-income economies, particularly during the last decade. Although Indonesia's growth of rice output is the most impressive in relative terms (a 47 percent increase between 1979 and 1985), large production growth was also achieved in a number of other Asian importing countries, including India (44 percent), Vietnam (43 percent), Sri Lanka (23 percent), and Bangladesh (21 percent). In the Middle East, import growth stems from population increases and, especially, income growth. Iran, Iraq, and Saudi Arabia alone accounted for 1.5 million metric tons of increased imports during 197o-85 (their total imports rose from o.2 to 1.7 million metric tons). Increased income also spurred import growth in the developed countries. In Africa, poor production performance in both rice and alternative grains and the availability of food aid are major reasons for the increases in imports.

Figure 3.1 provides a graphic description of price movements for the major grains during 197o-89. The international markets for wheat and corn can be considered in concert, primarily because of the wide use of both corn and wheat as animal feed. In most years, one-fifth or more of wheat production is fed to farm animals, and prices for these two grains are tightly linked (Pearson 19go). But price relationships with the rice market are much more volatile. In 1973_-75, rice prices reached record highs, primarily as a consequence of consecutive poor harvests in Asia during 1972 and 1973. Although prices of other grains increased sharply as well, rice prices increased more in relative terms. By 1974, the prices of both high-quality rice (represented by the prices of Thai 5 percent brokens) and low-quality rice (represented by the average unit values of Burmese exports) were much higher than those of wheat and corn. These margins remained large for nearly a decade. Siamwalla and Haykin (1983) identify three reasons for the increases in rice prices relative to wheat prices before 1981-slower growth rates of rice production, faster population growth in rice-consuming areas, and higher income elasticities of demand among rice consumers.

After ig8i, the production growth of rice was faster than that of wheat and corn, and this reversal provides part of the explanation for the post-198o decline in relative rice prices. Between 1979 and 1985, world rice production grew by 22 percent (258 to 316 million metric tons of milled rice), world wheat production increased by ig percent (424 to Sot million metric tons), and coarse grain production grew by only 13 percent (744 to 843 million metric tons) (World Food Institute, 1986). By 1987, Burmese rice prices were only 6 percent higher than wheat prices, the smallest relative premium since 1972.

In late 1987, rice prices again increased relative to wheat and corn prices, reflecting the poor weather in Asia. The U.N. Food and Agriculture Organization (FAO) estimates that 1987 production of rice was nearly 4 percent below 1986 production, the first absolute decline since 1972_-73. During the spring of 1988, the price of Thai 5 percent brokens rose above $300 per metric ton and fluctuated around that level throughout 1989.

This review of patterns in the main world cereals markets shows that world rice prices are not closely linked to world prices of wheat and corn. Some constraint on the upward movement of rice prices is provided by importing countries substituting wheat and rice. Further, the suitability of rice for animal feed means that rice prices will not fall below corn and wheat prices. The much larger corn and wheat markets, therefore, provide floor prices for the relatively small world rice market. But, in general, the magnitude of price fluctuations in the rice market has been much larger than those in the wheat and corn markets. The world rice price has remained highly sensitive to small changes in the balance between domestic rice production and consumption.

Trade Policies and World Rice Prices

The continued thinness and price variability in the rice market are caused in part by the domestic rice policies. Table _3.2 summarizes information on trade policies used by some of the main participants in the world rice market. Among the exporters, only Australia allows unrestricted trade. The United States and the European Community (EC) countries (primarily Italy) use export subsidy programs; other exporting countries employ either government monopolies or export licenses to set trade limits. Among importers, government monopolies, usually enforced by a national food agency, are widespread. Although the list given in Table _{3.2 is} not exhaustive, a survey of eighty-six countries, carried out by the FAO in _1983, confirms the prevalence of government monopolies and quantitative controls. These instruments were used by ₇₀ percent of importing countries and ₅₆ percent of exporting countries. A comparison with policies of ₁₉₇₃ shows very little change; although some countries reduced tariff barriers, none eliminated their nontariff controls.

Such widespread and persistent use of quantitative controls and variable taxes or subsidies reflects the effectiveness of trade policy in allowing governments to achieve a wide range of goals-domestic price stability, high prices for consumers or producers, or low prices for consumers or producers. Principal attractions of trade policy are low budgetary costs and ease of administration relative to alternative policies of domestic taxes or subsidies.

Price stability figures prominently in the trade policy decisions of most developing countries. Importing countries increase or decrease international purchases as needed to maintain the desired internal price level; exporters behave similarly with respect to the amounts offered to the international market. Because neither domestic consumers nor producers observe any changes in price, all of the instability must be absorbed in other countries that do respond to world prices. Hence domestic price stability, which is desired in many countries, is achieved only by creating more international price instability and by transferring the needed adjustment to other countries (Falcon and Monke _1979-8o; Siamwalla and Haykin _1983; Monke and Taylor _1985). The world price would be more stable if all countries dropped their trade barriers, but under current conditions no country dares to change its policies unilaterally. A stalemate results, and world prices remain highly variable (Monke and Salam _1986).

Low domestic rice prices relative to world prices have been desired in some developing countries such as Indonesia and Thailand in the _1970s. Importing countries can use import subsidies and exporting countries can use export taxes to keep domestic prices below world prices. For many developing country governments, trade policies have been preferred to

direct consumer subsidies to reduce budgetary expenditures. Import subsidies lower market prices and force domestic producers to provide some of the desired subsidy to domestic consumers; export taxes contribute to budget revenues. Because the desire for price stability usually exists concomitantly with the goal of low price levels, variable levies, quantitative controls, and government monopolies on trade often have replaced trade taxes and subsidies as the instruments of intervention.

A high producer price has been prominent in the rice policies of many developed countries-the United States, the EC, and Japan-but also has been evident in developing countries desiring to promote rice production. Indonesian policy in the mid-ig8os illustrates this policy in a developing country. The attraction of trade policy relative to direct subsidies to producers is again linked to budgetary implications. In importing countries, tariffs cause prices on the domestic market to rise. Domestic consumers thus provide the desired subsidy to domestic producers, and the government gains revenues from import taxes. In exporting countries, the provision of export subsidies permits the government to maintain a favorable price for farmers without adding to government-owned stocks. Domestic consumers again provide some of the subsidy to producers because they are forced to pay prices that are above those in the world market. The simultaneous desire for domestic price stability mandates the use of quantitative trade controls or variable levies in place of trade taxes.

Domestic policy objectives have almost always taken precedence over international market incentives. Table 3.3 reproduces historical information gathered by Barker, Herdt, and Rose. This table shows the consistency of internal price differences among three groups of Asian rice economies between 1961 and ig8o. The highest domestic prices were in high- or middle-income countries-Japan, South Korea, Malaysia, and Taiwan. Domestic prices were near the world price in China and in the low-income, traditional importing countries (some of which are now self-sufficient)-Indonesia, Bangladesh, Philippines, India, and Sri Lanka. Very low domestic prices arose from taxing policies in three exporting countries-Thailand, Pakistan, and Burma. In some of the countries, the pattern has changed over time. Indonesian producer prices have risen substantially in the last decade relative to world prices; Sri Lankan prices have fallen even more dramatically. These changes correspond to fundamental shifts in domestic policies: in Indonesia, the achievement of self-sufficiency and agricultural income growth; in Sri Lanka, the emergence of government and consumer budget constraints on high producer prices.

Policies and Future Changes in the World Rice Market

World price effects can arise from any of a large number of possible shocks to the international demand-supply balance in major producing countries. A number of stochastic events, if transmitted to world markets, could cause international prices to rise: disease infestation or pest attacks on high-yielding varieties; drought in any of the major producing countries; a reduction in producer support in developed countries (such as Japan); political instability in major producing or consuming countries causing reduced exports or increased imports; and emergence of a new source of import demand. Although the probability of any one of these events is small, the joint probability is high. When that probability is accompanied by trade policy controls in many of the market participants, world prices can be expected to fluctuate frequently.

The world price of rice peaked four times in the past two decades-in

1974, ₁97₈, 1981, and 1988-and fluctuations are likely to continue at similar frequencies in the future. Problems from weather, diseases, and pests will continue to cause relatively large variations in market demand and supply. The increased emphasis given to accumulations of domestic stock in some countries, such as India and Indonesia, may moderate the magnitude of future fluctuations relative to those of the 1970s. But policymakers' demand for food reserve stocks has proven nearly as inelastic as the demand for imports to meet domestic consumers' needs, and stock policy seems unlikely to reduce the frequency of fluctuations.

The prominence of policy interventions in the world rice market means also that forecasts of long-run world prices are hazardous. Econometric models can confirm the importance of domestic policy objectives in explaining world market performance and can show that world prices are unstable because government institutions transmit variabilities in domestic production to the international market. But such models are not useful for making price forecasts because predictions of future changes in government rice policies are so precarious. Policy changes depend on alterations in political environments and political responses to varying economic conditions. Statistical models are of little assistance in predicting these changes. For example, economists in the mid-1970s were unable to foresee the major changes in Indonesian or Chinese agricultural policies. Indonesia emphasized positive production incentives and reduced rice imports by z million metric tons, whereas China encouraged crop diversification and reduced exports of rice by ₁ million metric tons. Such changes had direct implications for world prices.

Predictions of long-run rice prices thus depend on judgmental analysis of the interactions between likely economic and policy changes. The marked shifts in import shares in recent years provide some reason to think that international demand may be more stable than in the past. Between 197o and 1985, the import share of the Middle East and the developed countries doubled-from _Zz to 44 percent. The addition of import shares of high-income countries from other regions (such as Hong Kong, Malaysia, and Singapore) means that well over half of world import demand in the mid-1g8os arose in countries with relatively high per capita incomes. Most of these countries have little or no local production of rice. Demand for imports tends to be highly price-inelastic because high-income consumers are insensitive to prices, and stable, because income elasticities are near zero and domestic production is unimportant. But increased stability of total demand depends largely on the ability of former importing countries-particularly in Asia-to sustain production growth in the future.

Among rice-producing countries, key policies involve investment and prices. Investments in new technology, including expenditures on research to develop modern varieties and improved farming and processing practices, subsidies on inputs (especially fertilizer), and investments in irrigation and other farm and postfarm infrastructure can have enormous impacts on supply. If such investments result in global expansion of rice output at rates exceeding those of rice consumption, the world rice price level will continue to decline. Technological change in rice (and in other cereals) has been sufficiently rapid in the past to permit the trend of real prices of rice (and of wheat, corn, and other coarse grains) to fall slightly during the past several decades. The interesting question is whether this gradually declining trend is likely to continue.

Barker, Herdt, and Rose (1985) project likely changes in both supply and demand for eight Asian countries (China, India, Indonesia, Bangladesh, Thailand, Burma, Philippines, and Sri Lanka) that together account for 85 percent of Asia's rice production (and thus over three-fourths of global rice output). Table 3.4 shows the results of their base-run projection of production for the year gooo. This projection assumes that irrigated area grows at historic rates, modern varieties spread wherever possible, fertilizer availability increases at a rate of 5 percent per year, and productivity of inputs does not change. The result of this projection, 409 million metric tons of paddy, is much less than their projection of demand by these eight countries for gooo, 482 million metric tons. This demand estimate assumes lower population growth rates than those that occurred in the 1970s but similar rates of growth of income to those of the 1970s and declining income elasticities of demand.

The authors next change a critical policy assumption and project the results of a doubling in the historic rates of growth of irrigated land in seven of the eight countries (excluding China); irrigated area in 2000 then climbs from 54 percent to 62 percent of total area, and projected production expands to 466 million metric tons, approximately the level of the demand projection. Finally, they relax the antihistorical assumption of constant productivity and make projections of the effects of productivity increases for fertilizer, irrigation, and both fertilizer and irrigation combined; the resulting projections of rice output are 460, 499, and 543 million metric tons, respectively.

This exercise shows the difficulty of making price projections for a market in which price and investment policies dominate. Whether growth of the supply of rice in Asia is likely to fall short of, just keep up with, or exceed growth of demand depends in large part on what is assumed for investment policy, especially in irrigation, and for price policy and productivity growth. Projections that run far into the future are particularly vulnerable to errors of being too optimistic or pessimistic about the speed and effects of future technical change.

This central point can be illustrated by Indonesia's recent experience with investment and price policies for rice production, detailed in Chapter 2. Indonesia's recent success in expanding output to achieve trend self-sufficiency in rice was not predicted. Appropriate technology policy combined with high incentive price policy to create conditions for rapid growth of rice production, especially between 1977 and 1984. But these policies were complemented and made possible by a stable political situation, macroeconomic policy that did not regularly tax agriculture, petroleum revenues that permitted investments in agricultural and transportation infrastructure, the development of capable individuals and institutions to carry out analysis and to implement policy, and freedom from sustained periods of losses caused by poor weather, disease, or pests. Sensible projections of the future impacts of policies on rice production, consumption, and trade must consider these complementary forces as well.

The future direction of price trends will depend primarily on the conduct of policy in the Asian countries. Other importing regions are small relative to the size of the international market, and production capacity in the major exporters is probably adequate to satisfy increased demands from these regions. Barker, Herdt, and Rose (1985) provide evidence that in the larger Asian countries production capacity is adequate to maintain pace with demand, if policies become more supportive of domestic producers. Such changes probably will be forthcoming. The agricultural sector has become increasingly important in economic development plans as better ways are sought to address problems of employment and income distribution; the recent experiences of Indonesia and India provide successful examples. Although the timing of policy changes is difficult to predict-and such changes could occur partially in response to a rising world rice price-their introduction should be adequate to prevent any substantial increases above the 1989 price levels of $300 to $350 for Thai 5 percent brokens. Prices could be even lower if investment activity greatly accelerates or new technologies are developed. In this event, prices can be expected to continue their long-run decline.

Conclusion

The level of the world price of rice influences Indonesian food policy because it serves as an indicator of the value of additional rice output and sets bounds on the income distributional effects of price policy. The instability of the world price of rice has encouraged the creation of a buffer stock policy, carried out by BULOG, to insulate domestic rice prices. Both the level and instability of world prices have implications for the government budget. Consequently, any Indonesian strategy for rice production, consumption, and trade needs to be grounded in a careful analysis of past and future world rice prices.

World rice prices are influenced heavily by government price policies that insulate producers and consumers from international price instability and by technology policies that increase output and productivity through public investments. Because of the large impact of policies, efforts to project future world prices are hazardous. Econometric models have difficulty incorporating these policy effects and are of little help in predicting them. Less sophisticated exercises to contrast projected growth in demand with that in supply indicate that future growth in rice production has the potential to match growth in consumption.

Indonesian policymakers can expect the world price of rice to continue to be highly variable and difficult to predict. The best guess for the expected future world price trend is that there will be little or no change in the trend of the real rice price. But this outlook-and especially year-to-year world price fluctuations-are influenced heavily by government policies and possible changes in them. In addition, world rice prices in any given year are buffeted about by supply shifts caused by good or bad weather and political disturbances. Perhaps the central lesson for rice policymakers in Indonesia is to measure the impact of their policies relative to the world price trend while implementing flexible trade policies that minimize the costs of domestic price stabilization.

1

We gained valuable insights for this chapter from discussions at a two-day seminar on the world rice market, held in Jakarta, Indonesia, during January 1987. Participants from outside of Indonesia included Donald Mitchell (World Bank), Robert Schwartz (Harvard Business School), Ammar Siamwalla (Thailand Development Research Institute), C. Peter Timmer (Harvard University), and the Stanford Project team.