CHANGING FISH TRADE AND DEMAND PATTERNS IN DEVELOPING COUNTRIES AND THEIR SIGNIFICANCE FOR POLICY RESEARCH by Christopher L. Delgado and Claude Courbois MSSD DISCUSSION PAPER NO. 18 Markets and Structural Studies Division International Food Policy Research Institute 1200 17 St. N.W.th Washington, D.C. 20036 U.S.A. Tel. (202)-862-5600 and Fax (202)-467-4439 October 1997 Contact: Carolyn Roper Phone: 202/862-5600 or Fax: 202/467-4439 Paper prepared for the International Consultation on Fisheries Policy Research in Developing Countries: Issues, Priorities and Needs, North Sea Centre, Hirtshals, Denmark, June 2-5, 1997. MSSD Discussion Papers contain preliminary material and research results, and are circulated prior to a full peer review in order to stimulate discussion and critical comment. It is expected that most Discussion Papers will eventually be published in some other form, and that their content may also be revised. TABLE OF CONTENTS Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Trade Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Export Earnings and Prices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Fish Consumption Trends, Past and Present . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Projections of Fish Consumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Further Evidence on Demand Projections . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 LIST OF TABLES Table 1: Net imports and import unit value of fresh demersal fish, developed countries, 1973-1993 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Table 2: Net imports and import unit value of fresh pelagic fish, developing countries, 1973-1993 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Table 3: Net exports and export unit value for fresh demersal fish and crustaceans, developing countries, 1973-1993 . . . . . . . . . . . . . . . . . . . . . 7 Table 4: Real value of exports of fishery products from developing countries in perspective, 1983-1993 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Table 5: Shrimp and fishmeal export prices in context, 1970-1996 . . . . . . . . . 10 Table 6: Average landing prices of various seafood groups, 1984-1994 . . . . . 11 Table 7: Per capita annual consumption of demersal and pelagic fish in selected countries, 1973-1993 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Appendix Table 1: Income effects on fish, beef, and grain expenditures . . . . . 24 Appendix Table 2: Price effects on fish consumption . . . . . . . . . . . . . . . . . . . . 25 LIST OF FIGURES Figure 1: Effects of a ten percent increase in income on consumption of fish, beef, and grain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Figure 2: Effects of a ten percent price increase on consumption of fish . . . . . 18 i Abstract Trends for major fisheries products are evaluated for the past two decades, using aggregate annual data. Major changes have been propelled by income growth, changes in preferences and health concerns about meat in developed countries, leading to increased consumption of high-valued fisheries items such as shell and filet fish. Developing countries, especially East Asia, are rapidly increasing consumption of lower valued fishery items, and fish-culture is becoming an increasingly important source of food and exports. Developed countries accounted for 85 percent of net world fish imports in 1994, mostly at the high end of the value spectrum, from about twenty countries. In the ten years preceding 1993, the net value of fisheries exports from developing countries went from less than a third of net developing country exports of sugar, beverage crops and tropical specialty products combined, to a level exceeding that total. While real fish prices have remained relatively stable since 1970, real beef prices have declined by 300 percent, suggesting that a rally in meat prices would further accentuate the shift to fish. Current evidence suggests a 15 percent relative strengthening of fish prices to beef through 2020. 1 The analysis of aggregate fisheries trends draws heavily on the seminal1 State of Fisheries and World Agriculture: 1996 (FAO, 1997) and the underlying FAO technical papers prepared for the Kyoto Conference on the Sustainable Contribution of Fisheries to Food Security, December 1995. For non-fisheries oriented readers, pelagic fish refers to surface feeding2 marine fish. Tuna and Bonitos are in fact pelagic fish, but are relatively high valued species that also tend to be highly migratory. Most pelagics are low- valued species found close to continental shelves (shoaling). They are important as food to artisinal fisherman in developing countries, and as feedstock to capital-intensive industrial fisheries seeking material for fishmeal. (continued...) Introduction Consumption of fish and fishery products increases rapidly with income. Thirty percent of world fish production is traded, and an increasingly large share of fish exports to developed countries will be produced in developing countries in the future (FAO, 1997). Policy research in developing countries therefore needs to address the factors that influence demand for fish and competing products in both the developed and developing countries, and how these changing patterns are likely to influence relative fish prices and trade flows. The objective of the present paper is to marshal and analyze salient evidence on aggregate fisheries trade and consumption trends, almost all of it produced by the Fisheries Department of FAO, and to put this in the context of other work on present and future demand for both meat and fish. Conclusions1 will be drawn where possible, and priorities for further research highlighted. Other papers, dealing with supply issues, household food security considerations, and interactions with the environment have been prepared by others for the present conference; no effort will be made to deal with those important topics here. Trade Patterns Even aggregate trade patterns must be put in the context of broader production changes occurring in the world. Global production of fish and shellfish rose nearly 15 percent from 1990 to 1995. Of the preliminary figure of 112.3 million metric tons of fish and shellfish produced world-wide in 1995, roughly 19 percent came from culture, as opposed to capture, compared to 13 percent in 1990. Growth in marine capture fisheries is mainly limited to lower value, highly variable, stocks of pelagic fish, whereas the generally higher valued production of demersal species has been fairly stable since 1970 (FAO, 1997) . About 28 percent of global production in 1995--largely shoaling pelagics2 2 (...continued)2 Demersal fish are mid-water marine fish; they tend to be the higher-valued species common in Western fisheries (such as flatfish). This in fact creates a "price" of imports that is a weighted average import3 price, in constant dollars, where the weights are the relative contribution of each transaction to the total. such as anchoveta--was used for reduction into fishmeal. The dynamic factor in world fisheries is China, where fisheries production as a whole grew at 14 percent per annum between 1990 and 1994, with most growth coming from aquaculture production of herbivorous carps. This increase is the main factor behind the increase in the fish production share of developing countries, from 26 percent of world total fish production in 1988 to 35 percent in 1994. In the latter year, China and India alone accounted for roughly 60 percent of world aquaculture, which with annual growth rates of the order of 11 percent in recent years is one of the fastest growing sectors in world food production (FAO, 1997). In value terms, world fish trade has accounted for a fairly stable 50 percent of the estimated value of fish production since 1980 (FAO, 1997). However, in quantity terms it has increased steadily from 32 percent in 1980 to 38 percent in 1990 (Alexandratos, 1995). Since 1990, this trend has continued, with trade in low value products such as fishmeal accounting for most increased trade. In 1995, 85 percent of fish imports by value were to developed countries, with Japan alone accounting for 30 percent of imports (FAO, 1997). There has been an important shift in the direction of trade. In the mid to late 1980s, the U.S. was the second largest fish exporter in the world (shown in Table 1 as negative net imports of fresh demersal fish). By the early 1990s, the U.S. had become a net importer of fresh demersal fish. Japan greatly increased its reliance on imports after the mid-1980s. Europe continued as a major player in world trade in fisheries, with increasing net imports over time. As a whole, the developed countries are net importers of fish overall, and of fresh demersal fish, at the present time, and this trend is expected to strengthen (Table 1). The "real import unit values" for fresh demersal fish shown in Table 1 are obtained by dividing the total deflated value (constant 1980 US$) of gross imports by the total quantity of gross imports. The increase in the import unit3 value for the U.S. appears to result from an upgrading in the 3 Table 1: Net imports and import unit value of fresh demersal fish, developed countries, 1973-1993 Year Net imports (metric tons) Real import unit value (1980 US$/metric ton) U.S. Japan All Developed U.S. Japan All Developed 1973/75 41,267 N/A 84,642 $1,584 N/A $1,505 1976/78 42,463 1,538 67,283 $873 $4,329 $1,593 1979/81 -43,416 2,663 -21,680 $1,023 $4,788 $1,565 1982/84 -403,287 2,557 -433,337 $1,031 $4,704 $1,264 1985/87 -1,173,911 5,449 -1,235,505 $1,089 $4,776 $1,448 1988/90 -815,224 29,837 -816,837 $1,367 $2,897 $1,786 1991/93 26,781 48,539 19,800 $1,723 $2,123 $1,978 Source: Computed from data in FAO, 1996a. Note: Net imports are imports minus exports in the same calendar year. Real import unit values are the aggregate value of imports for the country and year concerned divided by the aggregate quantities, deflated by the World Bank's G-5 Manufacturing Unit Value Index base 1980. Annual values are first calculated then averaged over three years. 4 In West Africa in the mid-1980s, long-distance fleets of ships, particularly4 from the former Soviet Union and North Korea, delivered cubic meter blocks of frozen pelagics during the high pelagic landing season, from June to September. Although much of this may have gone for human consumption, the rapid rise of the poultry industry in West Africa at that time may have been a factor also. At other times of the year, the ships would concentrate on more valuable species for sale outside Africa. types of fresh demersal fish imported, reflected in higher prices for those items. The opposite effect is observed in Japan after the mid-1980s, with substitution of cheaper demersal fish. A similar table (Table 2) is shown for net imports of fresh pelagic fish in developing countries. Africa is a net importer on the whole, although less so now than in the mid 1980s, when factory ships of the former socialist economies would unload huge quantities of cheap frozen pelagic fish in West Africa. 4 Developing Asia is once again becoming a major net exporter (negative net importer, in Table 2) of pelagic fish. Both net exports of pelagic fish and the real import unit value are very dependent on five pelagic species that are highly variable in their appearance, and whose production runs tend to alternate between boom and bust (such as the Japanese Pilchard). High value fisheries exports with a broad market in the developed countries include fresh demersal fish and crustaceans. Most developing countries are net exporters of these items; as shown in Table 3, developing Asia is the big player. Growth of shrimp exports in countries such as Thailand and India has been phenomenal in the past decade. Interestingly, the white-tailed shrimp is the only fisheries product for which there is a developed international futures market (FAO, 1997). Export Earnings and Prices Fisheries have become the major foreign exchange success story of the developing countries over the last decade. From 1985 to 1994, foreign exchange earnings from net fisheries exports from developing countries have been estimated to have gone from US$5.1 billion in 1985 to US$16 billion in 1994 (FAO, 1997). After allowing for inflation, this is a 95 percent increase in real terms over ten years, coming at a time when receipts from traditional exports of tropical agricultural products were declining due to sharp price decreases for those items in terms of purchasing power. Developed countries have also rapidly expanded their fisheries trade with each other. 5 Table 2: Net imports and import unit value of fresh pelagic fish, developing countries, 1973-1993 Year Net imports (metric tons) Real import unit value (1980 US$/metric ton) Africa Asia All Africa Asia All developing developing developing developing developing developing 1973/75 -5,174 -268,874 -277,144 $520 $546 $568 1976/78 11,502 -46,012 -35,492 $641 $416 $460 1979/81 22,732 -84,606 -56,556 $981 $342 $476 1982/84 138,464 -34,257 104,933 $615 $478 $550 1985/87 180,703 -40,118 83,795 $475 $436 $457 1988/90 118,546 -152,781 -59,450 $478 $355 $407 1991/93 86,358 -251,246 -190,398 $514 $499 $505 Source and note: See Table 1. 6 Table 3: Net exports and export unit value for fresh demersal fish and crustaceans, developing countries, 1973- 1993 Year Net exports (metric tons) Real export unit value (1980 US$/metric ton) Africa Asia All Africa Asia All developing developing developing developing developing developing 1973/75 8,164 266,055 -25,987 $3,298 $1,325 $3,251 1976/78 2,736 60,057 13,620 $948 $1,351 $1,517 1979/81 1,784 91,284 29,790 $4,786 $1,211 $1,945 1982/84 2,669 78,636 86,502 $3,965 $1,257 $1,543 1985/87 5,463 123,953 106,899 $3,937 $1,298 $1,919 1988/90 6,589 162,845 63,718 $3,519 $1,245 $2,184 1991/93 8,332 247,123 58,843 $3,566 $1,423 $2,212 Source: Computed from data in FAO, 1996a Note: Net exports are exports minus imports in the same calendar year. Real export unit values are the aggregate value of exports for the country and year concerned divided by the aggregate quantities, deflated by the World Bank's G-5 Manufacturing Unit Value Index base 1980. Annual values are first calculated then averaged over three years. 7 Global net foreign exchange receipts increased 70 percent in real terms between 1985 and 1994, ending at US$47 billion in the latter year (Ibid.). Table 4 shows the rise of real foreign exchange receipts of developing countries from fisheries from 1983 to 1993, compared to events in other agricultural and livestock export markets. While other traditional exports stagnate or decline in real terms, fisheries receipts are growing rapidly. By 1993, they surpassed those from sugar, traditional beverage crops (coffee, tea, cocoa) and tropical agricultural exports combined, for developing countries as a whole, although fisheries exports tend to originate from a small number of countries. Over the long run, prices for developing country fisheries exports have stagnated, even if in the last half-dozen years there has been some improvement. While the increase in nominal shrimp prices since 1970 in Table 5 may seem tremendous, the 1996 actual price was only 2 percent above the 1970 price expressed in 1996 dollars! Seen the same way, fishmeal prices in 1996 were only 55 percent of their real 1970 levels. In recent years, the market prices of higher valued demersal fish (such as flatfish) have tended to increase because of exhaustion of supply (Alexandratos, 1995). Events affecting prices of substitutes for fisheries products may be more significant for developing countries in the long-term. The price of beef in 1996 was only 29 percent of the 1970 price when the latter is converted to 1996 dollars. Using the figures from Table 5, the ratio of the wholesale prices of shrimp to beef was 2.1 / 1 in 1970, 3.8 / 1 in 1980, and 7.1 / 1 in 1996. The ratio of fishmeal to soymeal wholesale prices was 3 / 1 in 1970, 1.9 / 1 in 1980, and 1.5/1 in 1996. Shrimp is one of the more expensive fishery items by weight, at 2 to 4 times the value by weight of demersal fish, and 2 to 3 times the value of tuna, as shown in Table 6. Yet shrimp prices do not appear to have moved in radically different manner than demersal fish prices over the long haul. Therefore, the cheapening of beef relative to shrimp also applies to white fish filets, beef's closest competitor on the fisheries side. Fish Consumption Trends, Past and Present Average human consumption of fish as food is estimated by FAO as being 13.4 kg/capita in 1994; more surprising, the global average for fishmeal utilization in the same year was 5.6 kg./capita (FAO, 1997). Fish as food has recently been estimated to account for 19 percent of animal protein supply to humans, and 4 percent of total protein (Alexandratos, 1995). Globally, fish consumption per capita has grown only modestly over 8 Table 4: Real value of exports of fishery products from developing countries in perspective, 1983-1993 (in billions of constant 1988 US dollars)a 1983 1988 1993 Fishery products 9.3 15.0 17.4 Sugar, beverages, and 31.0 24.1 17.1 tropical products Oilseeds and products 12.7 13.3 14.2 Cereals and products 8.2 4.8 5.9 Meat and livestock 6.3 5.3 5.0 Milk and products 0.3 0.4 0.6 Sources: Computed from data in FAO Commodity Review and Outlook (various) and World Bank Commodity Markets and the Developing Countries (various). Note: Value of gross exports from developing countries including re-a exports, deflated by G-5 Manufacturing Unit Value index, base 1988. 9 Table 5: Shrimp and fishmeal export prices in context, 1970-1996 (current US$/metric ton) Year Shrimp Fishmeal Soymeal Beef MUVa b c d e Past 1970 2,780 309 103 1,304 25.1 1980 10,230 504 262 2,760 72.0 1985 10,490 280 157 2,154 68.6 Recent 1990 10,790 413 200 2,563 100.0 1991 11,542 478 197 2,663 102.2 1992 10,950 482 204 2,455 106.6 1993 11,390 365 208 2,618 106.3 1994 13,080 377 192 2,331 110.2 1995 13,540 496 197 1,907 119.4 1996 13,119 585 257 1,781 116.4 Projected 2000 14,036 n.a. 250 2,590 127.6 2010 15,859 n.a. 306 3,220 161.0 Sources: All price data and projections are in current US$ and from various issues of World Bank, Commodity Markets and the Developing Countries. Notes: U.S. frozen, Gulf brown, shell-on, headless, 26/30 count per lb., a c.i.f. New York. Any origin 64-65%, c.i.f. Hamburg.b Any origin, 44-46% extraction, c.i.f. Rotterdam.c Australia/New Zealand, frozen boneless cow forequarters, 85% d lean, c.i.f. U.S. East Coast ports. G-5 Manufacturing Unit Value index, base 1990 (to measure e inflation). 10 Table 6: Average landing prices of various seafood groups, 1984-1994 (current U.S. dollars/metric ton) Year Fish for Cods, Flounders, Tilapias Tunas, Salmons, Shrimps, Lobsters reduction hakes, halibuts, and other Bonitos, trouts, prawns spiny-rock haddocks soles cichlids billfish smelts lobsters 1989 $90 $700 $1,198 $1,250 $1,700 $3,500 $4,000 $11,270 1990 $95 $900 $1,217 $1,300 $1,830 $3,430 $3,650 $11,400 1991 $90 $1,250 $1,095 $1,300 $1,560 $3,190 $3,280 $11,700 1992 $93 $1,350 $1,163 $1,350 $1,650 $3,470 $3,750 $12,100 1993 $74 $990 $1,103 $1,250 $1,525 $2,780 $3,900 $11,500 1994 $85 $1,060 $969 $1,255 $1,500 $2,750 $4,000 $11,800 Source: Annual average values from FAO Yearbook, Fisheries Statistics (1994). 11 the past quarter century, from about 10.5 kg in 1970 (Westlund, 1995) to 13.4 kg in 1994 (FAO, 1997). While the developed countries consumed nearly 27 kg/capita of fish in 1991, developing countries as a whole consumed under 9 kg/capita (Phelan and Henriksen, 1995). The latter figures may be compared to 36.6 kg of milk and 17.7 kg of meat. Excluding China and Oceania, per capita growth rates of fish consumption as food in developing countries since 1970 have been low. Rates of the order of 0.5 to 1 percent p.a. (Westlund, 1995) may be compared to aggregate growth rates for beef consumption per capita in developing countries of 0.6 percent p.a. from 1967 to 1982, and 1.1 percent p.a. from 1982 to 1993 (Rosegrant et al., 1997). It is however noteworthy that growth rates for poultry per capita consumed as food in developing countries were over 5 percent p.a. in the earlier period and 3 percent p.a. in the latter period (Ibid.). The big growth in fish consumption in the developing world in recent years has been in China, starting from a very low base. In China in 1970, per capita consumption of fish was estimated at 3.6 kg; meat was 10.3 kg. By 1990, fish was 9.8 kg and meat was 27 kg, exhibiting very similar growth rates (Westlund, 1995). In southern coastal areas such as Guangdong, Westlund estimates per capita fish consumption at 35 kg. capita. Fishery items consumed as food cover a large variety of species; pelagic and demersal fish combined account for about half on average in most areas. Table 7 shows trends for important areas of per capita consumption of these items. Much of the growth in fish consumption in China has come from freshwater aquaculture production. The increase in per capita marine fish consumption has come mostly from pelagics. The same can be seen to be true for the rest of Asia and Africa, but not for the USA and Japan. Japanese consumption trends also reflect the beginnings of diversification out of a very highly seafood-intensive diet (71.5 kg/capita in 1994) into meat (FAO, 1997). In the U.S. and Western Europe, higher priced calories from higher quality fish is the likely route of the future, reflected in declining consumption of pelagics in the U.S. (mostly canned fish) and rapidly rising consumption per capita of demersals. Projections of Fish Consumption Policy research tends to be effective when it reduces the risks of uncertainty faced by investors and governments, both of whom need to take 12 Table 7: Per capita annual consumption of demersal and pelagic fish in selected countries, 1973-1993 (kilograms) Year China Africa developing Asia developing USA Japan except China Pelagic Demersal Pelagic Demersal Pelagic Demersal Pelagic Demersal Pelagic Demersal 73/75 .19 .85 1.81 1.20 2.22 1.45 3.80 4.12 19.39 29.11 76/78 .31 .68 2.44 1.47 2.36 1.57 3.44 4.38 18.50 22.68 79/81 .28 .81 2.81 1.63 2.40 1.46 3.48 4.28 19.55 17.64 82/84 .31 .75 2.60 1.58 2.66 1.43 3.16 5.00 21.18 16.89 85/87 .39 .71 2.43 1.05 2.55 1.50 3.65 6.13 20.57 18.19 88/90 .38 .72 2.45 1.37 2.64 1.43 3.43 8.68 18.83 17.55 91/93 .59 1.16 2.40 1.22 2.79 1.46 3.18 9.15 18.00 13.58 Source: Computed from FAO 1996a. 13 Other areas where the degree of detail in the existing quantitative fish5 projections is not clear include urban/rural differences in consumption patterns, detailed regional disaggregation of income elasticities, and formal incorporation of supply constraints, although all of these are addressed in detail outside the projections per se. a long-term perspective about the world in which they interact. Therefore, projections of future fish consumption in different parts of the world, and of future fish prices relative to other items, are potentially of great interest for policy researchers and analysts. Projections of fish consumption into the future tend to be scarce due to data limitations and methodological complexity. Because of data problems and the chosen focus on cereals, the detailed IFPRI global food projections to 2020 arising out of the IMPACT model do not include fish (Rosegrant et al., 1995). The IMPACT model does include five sub-categories of livestock products, along with many other agricultural categories, and this work, updated in 1997, arguably employs the most theoretically satisfactory approach to projections in the current global modeling literature, based on heavy data input. The model includes substantial regional disaggregation, allows for endogenous formation of major prices, and generally reconciles supply and demand constraints. FAO has done the only large scale modeling of long-term future fish demand on a global scale known to the authors. Although reported in different places, it is based on work with FAO's Food Demand Model, and appears to be most extensively documented in an appendix to Westlund, 1995. This work makes major simplifications in order to deal both with data limitations and the need for fast results. The cost is that it assumes constant relative prices for fish products into the indefinite future, which we have seen not to be the case in the past, and which FAO experts have clearly argued is not likely to be the case in the future (Westlund, 1995; Alexandratos, 1995; FAO, 1997). The absence of5 real price variation therefore does not allow for substitution effects between fish and meat (and fishmeal and soymeal), nor does it allow for induced investment in fisheries from increased relative profitability. On the positive side, it is all we have and it addresses a critical set of issues. The FAO projections predict an aggregate food utilization of fish of 110- 120 million MT in 2010, compared to 75-80 million MT in 1994/95. Per capita consumption is expected to grow from 13 to 16 kg. Regionally, significant per capita growth occurs mainly in China and the high-income countries of East Asia. China's per capita consumption is projected to go from 9.8 kg in 1990 to 20 kg in 2010 (15 percent of total world consumption), an enormous increase when population growth is considered. The increase in the high-income countries of East Asia is projected to be an additional 14 kg over the 1990 total 14 of 23 kg. Other growth is in North America (+5 kg in quantity, plus higher priced items) and Australasia (+3 kg). Otherwise, growth in per capita consumption is projected to be relatively stagnant (Westlund, 1995). Generally, the projections confirm the view that increasing aquaculture production will be associated with higher fish consumption in East Asia, and that quality upgrades will be important for growth in North American markets. Further Evidence on Demand Projections Although we do not have better global projections for fish consumption, we do have three kinds of evidence that can be used to improve our view of the future of fisheries demand. These are: (1) detailed and methodologically sophisticated fish consumption projections for China by Huang et al. (1997), (2) detailed estimates of future annual growth in consumption for beef and poultry at the global scale, using the IMPACT model, by Rosegrant et al. 1997, and (3) a growing literature of systems demand studies for individual countries that include disaggregated information on fish. The first set gives information on the part of the world market for fish that is growing the fastest. The second set shows what is happening to competing commodities. The third set can be used to elucidate income and substitution effects. With regard to fish in China, Huang et al. (1997) distinguish rural from urban and a high income growth scenario from a low income growth scenario. Fish prices are not endogenous, but price substitution and income effects are allowed for. Under low income growth, per capita fish consumption in rural areas is projected to grow from 4 kg in 1991 to 8 kg in 2010, and to 10 kg in 2020. In urban areas under low growth, per capita fish consumption growth is from 12 kg in 1991 to 22 kg in 2010, and to 30 kg in 2020. Under high income growth, fish consumption in rural areas goes from 4 kg in 1991 to 12 kg in 2010, to 19 kg in 2020. In urban areas, it is projected to go from the current 12 kg to 35 kg in 2010 (like Scandinavia currently) to 61 kg in 2020 (like Japan currently). Thus income growth and urbanization are critical to fish demand in China. Huang et al. (1997) do back of the envelope calculations to show that urbanization alone will add 10 percent to China's fish consumption by 2020. With regard to the second set of evidence, using the IMPACT model, Rosegrant et al. (1997) project for developing countries as a whole over the 1993-2020 period annual growth in total food demand for beef of 2.78 percent p.a., and 3.14 percent for poultry. The comparable figures for developed countries are 0.28 percent p.a. for beef and 0.94 percent p.a. for poultry. Combined with the fact that population growth will occur primarily in the developing countries, it is clear where the significant growth in demand for animal protein will occur. Furthermore, projected total demand growth is higher in Southeast Asia (3.60 percent p.a. over the period for all meats) and in Sub- Saharan Africa (3.38 percent p.a. for all meats). 15 However, they are based on compensated elasticities, that are adjusted6 to allow for the fact that price changes also affect consumption through changes in overall purchasing power, and this "income effect" of prices is compensated for. Turning to the third type of evidence, properly executed systems demand studies that include the significant complements and substitutes for fish, as well as fish itself, give us considerable perspective on the effects of income and different prices on fish demand. A series of relatively recent studies of this type was surveyed, and four "polar" country studies are reported in Appendix Tables 1 and 2. These were chosen by separation of the studies between developed and developing countries, and choosing "high" and "low response" countries for each group. The elasticities in the appendix are converted for presentation into more user-friendly bar charts in Figures 1 and 2. Figure 1 shows the importance of income response assumptions to projections of demand. Among developed countries, total expenditure on fish in the U.S. is seen to be highly sensitive to income. Under income growth, U.S. consumers substitute higher priced calories for lower priced ones, rather than increase their caloric intake. It is likely that choice seafood items are being substituted for meat. Grain intake does not grow with income. Norway, on the other hand, represents the case of a developed country where fish is a staple. With income growth, consumers are substituting beef for fish, although total expenditure on fish continues to rise. Among poorer countries, China has high fish response, although it is doubtful that fish is being substituted for beef. Instead, consumers are increasing their overall animal protein intake as income rises, and fish is the commodity most in vogue at this time, perhaps due to increased availability. Egypt, on the other hand, already has significant exposure to fish because of the close proximity of virtually the entire population to fish-bearing bodies of water. Incremental demand for animal protein appears to be relatively low, as income increases are associated more with attempts to diversify the diet. Clearly very different mechanisms are at work in the four cases, which illustrates some of the difficulties of global modeling, even with regional disaggregation. The detailed country demand studies do allow a view of price-substitution relationships among commodities, and the likely reaction of consumers to changes in relative prices. These are partial equilibrium relationships, and therefore are only indicative of the overall impact of price changes.6 Fish Beef Grain 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Percent increase Fish Beef Grain 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Percent increase Fish Beef Grain 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Percent increase Fish Beef Grain 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Percent increase 16 Figure 1: Effects of a ten percent increase in income on consumption of fish, beef, and grain More USA CHINA responsive Less NORWAY EGYPT responsive Sources: See Appendix Table 1. Fish Beef Grain 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 -12 -13 -14 -15 Percent change in fish expenditure Fish Beef Grain 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 -12 -13 -14 -15 Percent change in fish expenditure Fish Beef Grain 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 -12 -13 -14 -15 Percent change in fish expenditure Fish Beef Grain 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 -1 -2 -3 -4 -5 -6 -7 -8 -9 -10 -11 -12 -13 -14 -15 Percent change in fish expenditure 17 Figure 2: Effects of a ten percent price increase on consumption of fish More NORWAY CHINA responsive Prices Prices Less USA EGYPT responsive Prices Prices Source: See Appendix Table 2. 18 Figure 2 shows the impact on demand for fish of a 10 percent change in the price of each of three commodities. Fish consumption falls when its own price rises in all four countries, since it is more expensive. Fish price increases affect demand most in China. A rising fish price is bad news for fish producers in China (besides consumers), since a 10 percent price increase for fish is forecast to provoke a 13 percent cutback in expenditures on fish. The effect of higher fish prices in China is lower fish producer revenues. More revenue is lost through decreased sales than is gained through higher prices. In Norway, higher fish prices also have significant effect on the amount of fish consumed, and it seems unlikely that producers would gain much revenue from higher fish prices. In the U.S., higher fish prices increase the revenues of fish sellers (a 10 percent rise in prices is associated with only a 4 percent fall in sales). In Egypt, higher fish prices hardly affect fish consumption, suggesting that producers and sellers can potentially capture most of the increased revenue from higher prices. Beef sellers should not be indifferent to these results. In countries where fish consumption is an important part of the diet, such as Norway and Egypt, beef prices have significant opposite effects on fish consumption; higher beef prices increase fish demand on almost a one percent for one percent basis. Since cross price effects such as these tend to work in both directions, higher fish prices are likely to affect beef consumption significantly also. In Norway, consumers appear to be ready to shift from fish to beef, and back, rather responsively. In relatively poor China, the effect of higher fish prices may be to provoke a cut-back in fish, with relatively minor substitution into beef; possibly some of this decreased demand for fish shifts to other proteins. In the U.S., where fish is hardly a staple food, fish consumption patterns tend to be rather unresponsive to prices of either fish or beef. In all four countries, fish consumption is not much affected by grain prices. If we accept the view in Rosegrant et al. (1997) that world beef prices are likely to decline about 5 percent in real terms by 2020, and the emerging consensus that real fish prices are likely to rise on the order of magnitude of 10 percent, we would look for a long run change in relative fish / beef prices of 15 percent. As seen above, this is actually a rather conservative estimate, given the three-fold decline in the price of beef relative to fisheries products since 1970 (Table 5). Coupled with the cross-price elasticity estimates between fish demand and beef price in Figure 2, other things equal, this implies major adjustments in world markets for both fish and beef. In China, for example, thought to account for roughly 15 percent of world fish consumption in 2010 under constant prices (Westlund, 1995), such a relative price increase for fish would imply that fish demand would be 13-16 percent lower than it would have 19 These partial equilibrium estimates should not be over-worked, but they7 are indicative of orders of magnitude: 10% fish price change times 1.3 price elasticity of fish demand plus 1-2% (0.3 X 5%) further decrease in demand for fish because of lower beef prices. been if world relative prices had stayed at 1990s levels. In Egypt, beef demand7 would be at least 10 percent higher. Conclusions This paper has attempted to raise empirical and analytical issues from the recent literature pertaining to fish exports and consumption in developing countries, recognizing the increasing interrelatedness of events affecting fisheries around the world. It has highlighted major changes in the world over the past 15 years that imply that bodies interested in poverty alleviation and growth in the developing world cannot ignore the potential impact of fisheries on real incomes of producers and consumers. Fisheries are increasingly becoming one of the means that the developing world can use to pay for what by common agreement will be burgeoning cereals imports from the developed countries over the next 20 years. The developed countries are clearly going to become even more significant net importers of fish than they are presently (85 percent of net imports in 1994). Markets for high- value products, and for inputs to produce these items (such as fishmeal) will continue to expand rapidly. Fisheries have already surpassed the traditional export crops as a foreign exchange earner for developing countries (US$19.7 billion in 1993 in current dollars). Prospects for aquaculture-based high value exports appear good, at least from a market-outlet perspective. It seems likely that policy attention to fisheries issues will increase as was the case previously for cash crops in developing countries, once policymakers in those countries recognize that fisheries issues are no longer an obscure sectoral concern, or solely a welfare consideration for coastal people. As this policy attention increases, the glaring knowledge gaps will become more obvious. The most obvious one in the context of this paper is that we really have little idea of the forces driving relative fish prices, at least in a quantitative sense that would permit consistent long-term prediction. Data availability is improving, and this area needs to become better integrated within mainstream food modeling. Second, sectoral marketing policies, as well as trade liberalization, will do much to permit the expansion of high-valued fisheries production in the developing countries. However, fisheries products, like milk, meat and other perishables, tend to be associated with high transactions costs in trade, since 20 processing is an important part of value-added, and consistency and trust are so important in quality control along the market chain. The high importance of transactions costs in the growth of fisheries exports for developing countries suggests an important role for governments in fisheries, not as a substitute for producers or traders, but as builders of infrastructure, enforcers of rules, and facilitators of information flows. Third, many of the high value opportunities in fisheries are capital and information intensive, which will mean that artisanal fisherfolk will have trouble participating in the fisheries boom without participatory institutions of collective action to allow them an equal footing. Governments faced with a need to involve such populations need urgently to explore how they can facilitate the growth of such institutions without stifling them. Many issues appear, at least on the surface, to be similar to the construction of smallholder dairy systems in tropical climates. There may be symbiosis to explore here. Finally, this paper has intentionally ignored supply side issues such as technology development and environmental questions, to avoid duplication. Yet there is an urgent need for more policy research on the full costs and returns to technology development in fisheries, especially for aquaculture adapted to the needs of rural and coastal people in developing countries. This may end up being one of the most important policy areas to explore. 21 References Alderman, H. and J. von Braun. 1984. The Effects of the Egyptian Food Ration and Subsidy System on Income Distribution and Consumption. Research Report 45. Washington, D.C.: International Food Policy Research Institute. Alexandratos, N., ed. 1995. World Agriculture Towards 2010: An FAO Study. Rome and Chichester, U.K.: Food and Agricultural Organization of the United Nations and John Wiley. Brandow, G. 1961. Interrelations Among Demands for Farm Products and Implications for Control of Market Supply. Interregional Publication for the State Agricultural Experiment Stations Bulletin 680. University Park, PA: The Pennsylvania State University. Delgado, C. and A. McKenna. 1997. Demand for Fish in Sub-Saharan Africa: The Past and the Future. Naga, The ICLARM Quarterly XX (July- December): 8-11. Edgerton, D., B. Assarsson, A. Hummalmose, I. Lautila, K. Rickertsen, and P. Vale. 1996. The Econometrics of Demand Systems: With Applications to Food Demand in the Nordic Countries. Boston: Kluwer Academic Press. Food and Agriculture Organization of the United Nations. Annual. FAO Yearbook: Fishery Statistics. Rome. ________. Annual. Commodity Review and Outlook. Rome. ________. 1996a. FAOStat database. . Accessed May 1997. ________. 1997. The State of World Fisheries and Aquaculture 1996. Rome. Gao, X., and T. Spreen. 1994. A Microeconometric Analysis of the U.S. Meat Demand. Canadian Journal of Agricultural Economics 42: 397-412. Huang, J., and H. Bouis. 1996. Structural Changes in the Demand for Food in Asia. Food, Agriculture, and the Environment Discussion Paper 11. Washington, D.C.: International Food Policy Research Institute. Huang, J., S. Rozelle, and M. Rosegrant. 1997. China's Food Economy to the Twenty-First Century: Supply, Demand, and Trade. Food, Agriculture and the Environment Discussion Paper 19. Washington, D.C.: International Food Policy Research Institute. 22 Rosegrant, M., M. Sombilla, N. Perez. 1995. Global Food Projections to 2020: Implications for Investment. Food, Agriculture and the Environment Discussion Paper 5. Washington, D.C.: International Food Policy Research Institute. Rosegrant, M., M. Sombilla, R. Gerpacio, and C. Ringler. 1997. Global Food Markets and US Exports in the Twenty-First Century. Paper presented at the Illinois World Food and Sustainable Agriculture Program Conference "Meeting the Demand for Food in the 21 Century: Challenges andst Opportunities for Illinois Agriculture", Urbana-Champaign, Illinois, 28 May, 1997. Phelan, J. and J. Henriksen. 1995. Global Issues in the Supply of Livestock Food Products to Urban Populations. In Supply of Livestock Products to Rapidly Expanding Urban Populations, Food and Agriculture Organization of the United Nations. Rome. Westlund, L. 1995. Apparent Historical Consumption and Future Demand for Fish and Fishery Products--Exploratory Calculations. Paper prepared for the International Conference on Sustainable Contribution of Fisheries to Food Security, Kyoto, Japan, 4-9 December, 1995. Rome: FAO/KC/FI/95/TECH/8 World Bank. Bi-annually to 1991. Commodity Trade and Price Trends. Washington, D.C. World Bank. Quarterly since 1993. Commodity Markets and the Developing Countries. Washington, D.C. 23 Appendix Table 1: Income effects on fish, beef, and grain expendituresa (percent) Effect on Developed Developing More Less More Less responsive responsive responsive responsive (USA) (Norway) (China) (Egypt)b c d e Fish 1.7 .3 1.5 .5 expenditure Beef 1.1 1 .7 .7 expenditure Grain 0 .4 .1 .1 expenditure Sources and notes: Total expenditure elasticities, rounded to one decimal place.a The fish and beef elasticities are from Gao and Spreen (1994). The beefb elasticity is the average of steak, roast, and ground beef expenditure elasticities. The 'cereals and baking products' estimate is from Brandow (1961). Fish, meat, and 'bread and cereals' elasticities from Edgerton et al. (1996).c Fish, 'beef and mutton,' and grain expenditure elasticities from Huang andd Bouis (1996). The values are averages of urban and rural estimates. Fresh fish, fresh meat, and balady flour expenditure elasticities from Aldermane and von Braun (1984). The values are averages of urban and rural estimates. 24 Appendix Table 2: Price effects on fish consumption (percent)a Effect of a Developed Developing More Less More Less responsive responsive responsive responsive (Norway) (USA) (China) (Egypt)b c d e One percent -.8 -.4 -1.3 -.1 increase in the price of fish One percent .9 .1 .3 .8 increase in the price of beef One percent .1 0 0 0 increase in the price of grain Sources and notes: Compensated price elasticities from expenditure systems estimators, roundeda to the first decimal place. Own-price elasticity and cross-price elasticities of fish demand with respect tob meat and 'bread and cereals' prices from Edgerton et al. (1996). Own price elasticity and the average of the cross-price elasticities of fishc demand with respect to retail steak, roast, and ground beef prices from Gao and Spreen (1994). The cross-price elasticity of fish demand with respect to cereals is from Brandon (1961). Compensated own-price and 'beef and mutton' and grain cross-priced elasticities of fish demand are calculated from data reported in Huang and Bouis (1996). The values are averages of urban and rural estimates. Own-price elasticity, and cross-price elasticities of demand for fresh fish withe respect to prices of fresh meat and balady flour are from Alderman and von Braun (1984). The values are averages of urban and rural estimates.