3. Country-specific trends and projections

3.1 The major rice-growing nations

3.1.1 Food deficit and depth of hunger

Suggestions for FAO-supported interventions must necessarily relate to national, and often sub-national, circumstances and much less to the aggregate regional situations represented by the statistics in Tables 1, 2, 3, 5, 6 and 7. Interventions-relevant statistics (adapted from FAO 2000b, 2001b, UNICEF 2000, and IRRI 1997) are therefore presented in tables 8 to 15 for each of the seventeen Asian countries that annually harvest at least 0.6 Mha of rice. Among those seventeen countries, no fewer than eleven are classified as Low-Income Food-Deficit Countries. Those eleven comprise Bangladesh, Cambodia, China, India, Indonesia, DPRKorea, Laos, Nepal, Pakistan, Philippines and Sri Lanka. Each of them participates in the FAO-facilitated Special Programme for Food Security.

Moreover, FAO (2000h) lists most of those eleven countries as belonging to the group (also numbering eleven) falling within the two most-severe categories (Categories 4 and 5) of prevalence-plus-depth of hunger. These latter eleven are Bangladesh, Cambodia, India, DPRKorea, Laos, Nepal, Pakistan, Philippines, Sri Lanka, Thailand, and Viet Nam; (Indonesia, also, may now be Category 4).

The calculation of depth of hunger (dietary-energy deficit) is somewhat arbitrary. Nonetheless, and adapting from Singh (2001a, Table 3), we here suggest that for those persons already (arbitrarily) pre-designated as "hungry", their calorie intake as a proportion of a 2 300 kcal/person.day requirement is about 60 percent in Bangladesh and DPRKorea, and 62 - 64 percent in Cambodia, India, Laos, Nepal, Pakistan, Philippines, Sri Lanka, and Thailand. For China, FAO 1999a Annex 3 permits the conclusion that the poorest 10 percent of the populace has a calorie intake of only 70 percent of requirement.

As previewed earlier, we here record the association, at national level, of intensity of hunger with the type of rice-system water regime. Quantitative indicators for the relative extents of the four types of water regime have been tabulated by IRRI (1997 Appendix-Table 3). Using that tabulation, it may be determined that of those eleven fore-listed countries with most-severe prevalence-plus-depth of hunger, no fewer than ten, with Pakistan the exception, have at least 10 percent (and often much more) of their riceland as rainfed lowland. Three of those ten (Bangladesh, Cambodia, and Viet Nam) have at least 10 percent of their riceland in both of the categories rainfed lowland and rainfed floodprone (deep-water). Two others (DPRKorea and Laos) have 10 percent or more of their riceland in each of the categories rainfed lowland and rainfed upland. Three countries (India, Nepal, and Thailand) have more than 10 percent of their riceland in all three rainfed categories. Eight of the ten countries - Bangladesh, Cambodia, India, DPRKorea, Laos, Nepal, Philippines and Viet Nam - also have substantial rates of poverty.

[This analysis provides additional justification - if any be needed - for the emphasis accorded to irrigation and water-management support within the FAO-assisted Special Programme on Food Security.]

3.1.2 Statistical considerations

In Tables 8 to 15, some statistics necessarily pre-date the 1997-99 economic crises that affected several countries in North-East Asia and South-East Asia; (these Tables' data are correspondingly grouped as North-East Asia, South-East Asia, and South Asia). The entries in the eight Tables quantify and reinforce the existing general awareness of the magnitudes of food-security issues and of agricultural resources, productivity, and trade for the seventeen countries. They can correspondingly help identify individual countries' comparative advantages and their consequent opportunities wherewith FAO and other agencies could assist member governments to target interventions in support of their rice-region communities and enterprises. They can help identify also the national requirements to expand markets and to strengthen infrastructures, policies, and technologies.

Procedurally, and in recognition of the great disparity among the seventeen countries in terms of rice-growing area, human population, and other statistics (China: 32 Mha riceland, 1.24 B persons; Laos: 0.7 Mha, 0.005 B persons), several data and their interpretations are presented on a "per person" and a "per hectare" basis, so that inter-comparisons may realistically be effected. It is here recognized that in some countries' censuses (agricultural and demographic) the recorded annual totals for individual-crop production may be systematically inaccurate. Also, the human-population totals may fail to record all resident persons; and the non-recorded persons are most likely to be poor persons.

Thus, some population-derived statistics - such as available food energy (kcal) per person or particular-crop production per person or GNP per person may in absolute terms be over-optimistic in relation to food insecurity and poverty. However, since any shortcomings in the census compilations probably affect the whole sequence of a country's population totals, and not just those for any individual census, then the forecasts, interpretations, and recommendations made here and elsewhere in terms of trends (growth rates) and of relative (rather than of absolute) values are likely to be valid despite these shortcomings. For properties that under-standably exhibit year-to-year variations - as in crop production and international trade - the Tables present indicative values for recent triennia (usually 1997-99). Some social and human-nutritional data derive from surveys of the early 1990s.

[It is possible that in their programmes for decentralized planning and implementation of rice-region food- security/poverty interventions, national agencies may find that the format of Tables 8 - 15 can usefully be adapted to portray pertinent statistics at province or state level. It is also pertinent to indicate that FAO's Medium-Term Plan, recognizing the short-comings in some agricultural statistics, has a new activity and an ongoing Asia-Pacific Regional project to evaluate and to improve statistical-data quality.]

3.2 Demography, poverty and hunger

3.2.1 Demography and food production

Table 8 features aspects of rice production, and of human nourishment. Rice production per person (total population) ranges from as little as 40 - 45 kg/person.ann (Iran and Pakistan) to as much as 360 - 390 kg/person.ann for the rice exporters Myanmar, Thailand, and Viet Nam. Values for the proportion of rice in total kcal/person.day, which range from 5 percent for Pakistan to almost 80 percent for Cambodia and Myanmar, relate to Year 1992: values at Year 2002 are likely to be lower than the listed values; the listed values nonetheless provide reliable indicators of relativities. For five countries only (China, RoKorea, Malaysia, Pakistan and Philippines) do animal, including dairy, products exceed 15 percent of the total calories intake. Percentage growth rates (1990-2000) in rice production are seemingly highest for Cambodia, Laos, Myanmar, Pakistan and Viet Nam.

Table 9 indicates aspects of human population, poverty and undernourishment, and also rice production (repeated from Table 8). Thus, for all of the listed countries for which rice supplies 60 percent or more of the total kcal/person.day, the annual rice-production growth rate achieved during 1990-2000 (Table 8) is encouragingly higher than the annual population growth rate predicted for 1996-2010 (Table 9). However, for several of those countries for which rice supplies between 30 and 60 percent of the kcal/person.day, there is urgent need to use the methodology previewed for rice yield in chapter 1.2.3 to determine whether the rice-production growth rate during 1994-2000 exceeded the forecast 1996-2010 population-growth rate. For some of them - including the three major producers China, India and Indonesia - it perhaps did not.

Table 8: Rice production and human nourishment: Asian rice-producing countries

[Notes: Sources are FAO (2000b, 2001b), IRRI (1997), and UNICEF (2000, 2001);? indicates doubtful or not reported. Regional groups: North-East Asia, South-East Asia, and South Asia. Several statistics pre-date the 1997-99 economic crises in North-East Asia and South-East Asia. Rice production (Mt/ann and kg/person.ann) is representative of 1997-99; growth rate (%/ann) is for 1990-2000. Rice (%) in total kcal/person.day relates to Year 1992. Proportions of total dietary kcal energy supply - vegetable sources/animal sources - are at 1998-99.]

3.2.2 Undernourishment and poverty: Country-specific features

With the afore-mentioned exceptions of Bangladesh, Cambodia, Laos, Nepal and DPRKorea, and with the proviso that some countries' populations may have been underestimated, average values (1998 data) for k-calories/person.day (Table 8) are for most of the seventeen countries sufficient to satisfy national dietary kcal energy supply. Globally, at Year 2030, the average (including industrialized and developing countries) dietary energy supply is forecast to exceed 3 000 kcal/person.day. Nonetheless, at Year 2001 - because of inequity in purchase-power access to the available food, and because of seasonal shortages of food, which impact severely on growing children, seven of the seventeen listed Asian countries (Table 9) still have alarmingly high rates (> 40%) of moderate and severe infant (under-fives) undernourishment. For most of these countries, rice provides a considerable proportion of the dietary energy supply. The table 9 statistics for birth weight pre-date the East-Asian economic crises. They nonetheless indicate most encouragingly that for seven listed countries (China, Indonesia, Iran, RoKorea, Malaysia, Philippines and Thailand) the proportions of babies having birth-weight > 2.5 kg are comparable to those for industrialized countries. Conversely, for Bangladesh, India, Myanmar and Pakistan, and probably for Cambodia and Nepal, and possibly for Sri Lanka, the incidence of low birth weight is worryingly high.

Table 9: Rice production; population, poverty, and undernourishment: Asian rice-producing countries

[Notes: Sources are FAO (2000b and 2001b), IRRI (1997), and UNICEF (2000 and 2001);? indicates doubtful or not reported. Regional groups comprise North-East Asia, South-East Asia, and South Asia. Several statistics pre-date the 1997-99 economic crises that affected several countries in North-East Asia and South-East Asia. Rice production (Mt/ann and kg/person.ann) is representative of 1997-99; growth rate (%/ann) is for 1990-2000. Data for proportion of children (aged less than five years) who are moderately or severely under-weight (undernourished), and for birth weight, and for poverty derive from various pre-1999 surveys.]

Similarly, the incidence of poverty (Table 9, pre-crisis surveys for some countries) is distressingly high in Bangladesh, China, India, Nepal, Pakistan and the Philippines, and also almost certainly in DPRKorea and Viet Nam. FAO (1998b) categorizes Cambodia, Laos and Myanmar as being among "the poorest of the poor". Though not indicated in Table 9, levels of income inequity/inequality are recorded (FAO 1998b, UNICEF 2000) as being disconcertingly high (among countries for which analyses are possible) in Indonesia, Malaysia, Nepal, Philippines and Thailand, and in China and Viet Nam.

[It is here suggested that the relativities within the internationally-compiled Table 9 indications of poverty are probably valid. It is nonetheless necessary to caution that, because of their country-specific procedures of survey/analysis, national assessments of poverty and of undernourishment may differ greatly from the percentage values presented in Table 9.

Thus for India, the poverty percentage (44 percent) listed in Table 8 differs substantially and meaningfully from the Government value of 26 percent (via Singh 2001a, Table 15). Moreover, changes from time-to-time in national methodology and its threshold values for component variables introduce some variation into the time-series of poverty percentages that may distort reality. Thus seven successive all-India rural surveys (at 1983, 1988, 1993, 1994, 1995, 1996, and 1998, Kumar, personal communication, 2001) for percentage poverty among farm households generated the following sequence: 42, 31, 29, 25, 21, 20, 22; the non-monotonic nature of this sequence may be an artefact - but if it be real, it merits careful investigation.]

Within the four most-populous rice-producing countries (China, India, Indonesia and Pakistan) there resided at 1995/97 three-fourths of the global total of undernourished persons (adults and children). At 2015, of the projected global decrease (390 M) in the number of undernourished persons - a decrease consequent upon the forecast for those countries that their dietary energy supply shall increase to 2 700 kcal/person.day - these four populous countries (and their rice systems) shall in aggregate facilitate four-fifths of that global decrease in undernourishment.

3.3 Agricultural production

3.3.1 Cereals

For the seventeen rice-growing countries, and for eight field-crop-commodity groups and for livestock and for inland-fish, Tables 10 and 11 list the national-average production/person (food + feed, indicative, 1997/99) and the average annual growth rate in production (1999-2000). Rice data (listed also in Tables 8 and 9) are again listed for ease of comparison with other commodities. However, in comparing kg/person productions of cereals with those of tubers, fruits, vegetables, and milk, it will be recalled that for the latter commodities much of the mass is contributed by water. Additionally, this document cautions that for some countries the annual estimates for crop yield and production may be imprecise and/or systematically inaccurate.

Wheat - because of climatic constraints - may be profitably grown only in parts of China and in the South-Asian Indo-Gangetic Plains; annual growth rates for national wheat production are there appreciably higher than for rice. Among the listed countries, wheat production per person is highest in Iran, Pakistan and China (Table 10). Most of the listed countries annually import appreciable quantities of wheat flour and wheat products.

Maize and oil-crops, including oil-seed field crops and coconut- and palm-oil tree crops, are in most of the listed countries, and particularly in China and during 1999-2000, each increasing in production more quickly than rice (Table 10), and more quickly than human populations, and are forecast to continue to do so, as populations generally become more affluent. (Table 3 aggregates pertinent forecasts for all developing countries). These rising productions are meeting demands both for human food and, increasingly, for livestock feed.

[Global requirement for wheat-plus-coarse-grains is forecast to increase from 107 Mt/ann in 1996 to 200 Mt/ann in 2015 and 270 Mt/ann in 2030. North America, Western Europe, and Australia - each prospective donors to help decrease poverty in Asia - shall be the main suppliers of any Asian shortfalls in domestic productions of these commodities.]

Table 10: Per person production (food + feed, representative, 1997-99), and growth rate in total production (indicative, 1999-2000), for major food/feed groups: Asian rice-producing countries

[Notes: Source is FAO (2000b and 2001b);? indicates doubtful or not reported; - indicates the crop cannot profitably be grown. Regional groups comprise North-East Asia, South-East Asia, and South Asia. Some statistics pre-date the 1997-99 economic crises that affected several countries in North-East Asia and South-East Asia. Crop production (kg/person.ann) is representative of 1997-99; growth rate is for 1999-2000 - however, for some cereals and countries, production was roughly constant during 1996-98. Oil-crop production is represented by "oil equivalent".]

Within these all-cereals totals, contributions at 1996, 2015 and 2030 from rice are recorded/forecast for East Asia as 106, 102 and 98 kg/person.ann (while for East Asia excluding China as 130, 131 and 126 kg/person.ann), and for South Asia as 82, 89 and 84 kg/person.ann. Indeed, FAO forecasts that the rice/person.ann shall for many countries stabilize at the Year-2000 values. Correspondingly, much of the all-cereals food increases (at 2015 and 2030) shall be provided by wheat - of which rice-growing countries, including rice-and-wheat-growing countries, shall make substantial imports - though as IFAD (2001) emphasizes, the poor and very poor are unlikely to be consumers of these wheat imports.

3.3.2 Edible-oil crops

In Asian rice-growing countries edible-oil-crop products at 1995/97 (Table 10) contributed substantially less than the nutritionally desirable 10 to 12 percent of dietary calories (FAO 2000b). However, between 1995/97 and 2015, such products are predicted to provide one in three of all additional developing-country calories. To meet this increasing edible-oil demand, the required growth rate in oil-crop production shall be a modest 2 %/ann during 2002-2030, compared to the 4 %/ann of the 1980s/1990s. Palm-oil, of which rice-growing Indonesia, Malaysia and Thailand are dominant producers, has been and shall continue to be a major contributor to the lessening of undernourishment; conversely, Asian production of coconut oil has in recent years remained roughly constant.

For soybean, a current and prospective component of rice-system cropping - though soybean oil is less price-competitive than palm-oil and palm-kernel oil - area and production are in Asia each increasing appreciably. Table 3 forecasts the 1996-2030 growth rate in soybean production - aggregated for all developing countries, including the major producers China and India - as 2.6 %/ann. However, for oil crops in aggregate - as for meat and livestock - FAO (2001c and 2001d) describes for 1999-2000 a considerable global over-supply and an appreciable decrease in prices, such that between 1997/99 and 2000 the FAO oils/fats-price index declined from 140 to 89; however, during 2001, palm-oil prices recovered from their Year-2000 value. Notably, the USA is an increasingly price-competitive supplier of edible-oil products.

3.3.3 Fruits and vegetables

Fruits, vegetables, pulses and, to a lesser extent, roots and tubers are important for providing balanced nutrition within a cereals-dominated human diet. For pulses (Table 10), only India and Myanmar are major producers; India's rate of pulse production has during recent years increased only slightly. For roots and tubers - dominantly sweet potato, potato, and cassava, China, Indonesia and Thailand are the main Asian producers (Table 10). For root crops, the Asian area (about 7 Mha) of sweet potato is decreasing. For potato, some of which is grown in sequence with lowland or upland rice, the area, though small at 7 Mha, is increasing; yield and production also are increasing. For cassava, and in part in response to changes in European Union practices, there was a decrease in area planted (to 3.5 Mha at 1997-2000). Such decrease may provide a valuable opportunity to lessen upland-soil erosion if the cassava shall be replaced by crops that provide more-extensive and more-prolonged ground cover during the intense-rainfall season. (In Viet Nam, cassava is incorporated within erosion-control programmes.)

For vegetables, including cabbage, and for fruits, growth rates in production, though somewhat uncertainly determined, are for almost all of the seventeen countries appreciably and encouragingly higher than rates of population increase (Tables 9, 11). Among the listed countries, the substantial (per person) fruits and vegetables producers are Iran, China, DPRKorea, RoKorea, the Philippines and Thailand.

[However, FAO (1999a) cautions that many poor-family producers of fruits and vegetables perforce sell their produce, and do not themselves receive the balanced-diet benefits from their production - except perhaps through consumption of blemished items.]

3.3.4 Livestock products

For livestock and products (Table 11), Asia's major (per person) producers of meat - predominantly pig meat and poultry meat - are China, RoKorea, Malaysia, Philippines and Thailand. Prominent Asian producers of milk (including cow milk and buffalo milk) are RoKorea, India, Iran, Nepal and Pakistan. Production per person of eggs, predominantly hen eggs, increased very substantially during recent years - notably in China, and is forecast to continue to increase. For the periods 2000-2015-2030, FAO (2000d) and Steinfeld et al (1997) forecast that the annual growth rates for livestock products shall be about twice the growth rates for crops.

Growth shall derive initially from an increase in animal numbers - particularly of poultry and pigs, as compared to ruminants, and subsequently from increased weight per animal - facilitated by food supplements and concentrates (FAO 1999c). Growth shall result also from shorter growth cycles, from increased littering (pigs), and more-efficient feed-conversion (poultry).

Table 11: Per person production (representative, 1997-99) and growth rate in total production (indicative, 1999-2000) for major foods and food groups: Asian rice-producing countries

[Notes: Source is FAO (2000b and 2001b);? indicates doubtful or not reported. Regional groups comprise North-East Asia, South-East Asia, and South Asia. Some statistics pre-date the 1997-99 economic crises that affected several countries in North-East Asia and South-East Asia. Crop production (kg/person.ann) is representative of 1997-99; growth rate is for 1999-2000 - however, for some cereals and countries, production was roughly constant during 1996-98.]

However, the rate of increase in animal populations and in demand and production of meat is expected to slacken during 2002-2030: of the major consumers, China is already relatively well-nourished - though FAO 1998b Annex 3 Table 27 forecasts a doubling during 1996-2020 in the per person consumption of red meat, and India is not likely to become a meat-eating country. For mixed smallholder farms, it is highly pertinent (IFAD 2001) that small livestock grow and breed more quickly than large livestock. Moreover, smallholder populations of large ruminants may be constrained by decrease in average (mixed) farm size (LEAD 1999).

Table 11 includes values also for the per person production from inland (lake and river) capture fisheries. Such values probably do not include any fish production from rice fields; they are nonetheless here presented as an indication of prospective market opportunities for rice-fish enterprises, and possibly for rice-system crabs and shrimps. However, less than 1 percent of the Asian riceland area is currently used for rice-fish culture, whether concurrent or sequential (Funge-Smith, personal communication 2002). This notwith-standing that more than 100 fish species are suited to Asia's fresh-water and brackish-water ecozones, and that moderate-intensity rice-fish systems can typically produce 1 t fish/ha.year. Correspondingly, the number of rice holdings that harvest fish is a small fraction of the number that raise livestock. Nonetheless, crabs and shrimps - in addition to finfish - are cultivated in the Chinese rice systems; their production is forecast to increase throughout 2002-2030 (Dixon et al 2001).

In most rice-growing countries, however, the cultivar-induced shortening of the rice season, the lessening of the submergence-water depth, and the increased use of ricefield agro-chemicals, are severely constraining the living aquatic resource. That resource had heretofore included many specifically-adapted non-introduced species, including catfish, snakehead, gourami, indigenous cyprinids, eel, frog, snail, shrimp, and crab. The protein from the "by-catch" from that resource previously helped diversify cereals-dominated rural diets, or substituted own-farm-produced livestock protein which could thereby be sold for cash income - particularly among the poorest rural dwellers. That lost by-catch protein shall need to be replaced - either from fish or from livestock species.

The Asia-Pacific production of shellfish and finfish in aquaculture systems (predominantly in China) is some six times higher than the production from inland capture fisheries. During 1989-99 it grew at the extremely high rate of 13 %/ann, and is forecast (ADB 2001a) to continue that rate of growth. Globally, aquaculture currently produces about one-third of human-consumed fish; that proportion shall rise to one-half by 2030. Some of the aquaculture systems are constructed on converted ricelands. Regrettably, in some coastal ecozones those systems are non-sustainable, and are soon abandoned; but their soils are then unsuited to rehabilitation for rice

[Rice-fish systems, and aquaculture systems, feature in the ongoing activities in the Asia-Pacific Group of the FAO (2000a) Fisheries Department.]

3.4 Human and physical resources

3.4.1 Human resources

Tables 12 and 13 quantify the human and physical resources that facilitate rice-system production and rice-community livelihood in the seventeen specified countries. The quality of human resource is in Table 12 represented by secondary-school-enrolment percentages for males and for females. These data (UNICEF 2000, quoting UNESCO) pre-date the 1997-99 crises. They nonetheless indicate valuable strengthenings of human resources compared to the "Green Revolution Era" circa 1980. These higher educational attainments - notably in literacy and (less-often appreciated) in numeracy - shall facilitate the adop-tion of more-complex farming concepts and techno-logies and technology packages. They shall facilitate also the management, including financial manage-ment, of rural enterprises, and shall strengthen the ability to recognize pertinent opportunities. Women's education shall provide a crucial resource wherewith to combat rural poverty and undernourishment.

Table 12 indicates also that in East Asia (excepting Cambodia and Laos) female secondary-school-enrolment is comparable to that for males. In South Asia - and particularly in Bangladesh, Nepal, and Pakistan - female enrolment is substantially less than male enrolment, and more so in rural than in urban areas (IFAD 2001). However, gaps are slowly narrowing; and the female and male "net primary-school-attendance percentages" are becoming more-nearly equal. Crucially, women's representation in parliaments and in governments is being increasingly legally strengthened (ADB 2001a, IFAD 2001).

For poor rural households (whether landed or landless), the resource represented by non-farm employment and income (not quantified in Table 12) is almost always vital for the household's enterprises, food security, and wellbeing - particularly for farms that lack access to irrigation or to supplemental water. Such employment helps also the construction of social capital and assets.

Table 12: Indicators for human and land resources which facilitate rice-system production and livelihood: Asian rice-producing countries

[Notes: Sources are FAO (2000b and 2001b), IRRI (1997), and UNICEF (2000);? indicates doubtful or not reported. Regional groups comprise North-East Asia, South-East Asia, and South Asia. Some statistics pre-date the 1997-99 economic crises that affected several countries in North-East Asia and South-East Asia. School-enrolment percentages (male and female) derive from various pre-1997 surveys. Agricultural-land area (Mha) and agricultural population/agricultural-land area (person/ha, derived from FAO 2000b) are at 1998. Irrigated land (all crops) is indicative: Mha for 1997-98, and %/ann for 1988-99. Irrigated-riceland area and irrigated (total) land area may for some countries derive from different agencies; double-cropped irrigated riceland is generally counted twice, and data for fully and for partially irrigated ricelands are not distinguished; values here are estimated via IRRI (1997).]

3.4.2 Land and water resources

Data for agricultural-land area (Table 12) highlight the preponderance of China and India - which together contribute 304 Mha of the 454-Mha seventeen-country total. However, FAO (2000d) cautions that historic data for land area are for several countries unreliable. Notably, of the global total (73 Mha) of land brought newly into cultivation during 1975-95, no less than 70 percent (53 Mha) was devoted to oil crops.

Agricultural population per agricultural-land area indicates the prospective intensity of workforce that might be applied to agricultural production and enterprise. It correspondingly indicates that in countries such as Bangladesh, China, Laos, Nepal, Sri Lanka and Viet Nam the rural population per hectare of agricultural land may be so high as to raise concerns for land tenure and land fragmentation and the ability to support mixed crop-livestock systems (LEAD 1999).

The proportion of total agricultural land that is devoted to rice varies (among the seventeen countries) from 10 percent or less (Iran, Malaysia, Pakistan) to more than 70 percent (Bangladesh, Laos, Viet Nam) - though for these latter countries the proportions are inflated by the procedure whereby doubly-cropped rice area is counted twice. The countries devoting high proportions of land to rice are also those that have high agricultural population per hectare of agricultural land.

For irrigation, FAO (2000d) reports that at 1995/97 about 20 percent of developing-world arable land was irrigated. This irrigation facilitated 40 percent all developing-world crop production - expected to increase to 47 percent by Year 2030 - and 60 percent of developing-world cereals production. Irrigated-land production is more stable year-to-year than rainfed-land production; it correspondingly has higher benefit for food security.

The physical area of irrigated land (all developing countries) - which at 1995/97 was 197 Mha - is forecast to attain 242 Mha at 2030, representing a proportional increase x 1.23; the opportunities for additional multiple-cropping raise this factor to x 1.34. In East Asia, the irrigated area, for rice and non-rice crops, is projected to increase during 1996-2030 from 69 to 85 Mha; for South Asia, correspondingly, from 78 to 95 Mha; in each case, an annual compound rate of increase of 0.6 %/ann.

Table 12 indicates that for the seventeen countries, irrigated riceland comprises about one-half of the total of irrigated land. FAO (2000c: Facon) reports that in China and in India the allocation of irrigated area among rice, wheat, and other crops is roughly one-third each; while in South-East Asia rice occupies nine-tenths of the irrigated land.

Notably, of the total forecast increase (by Year 2030) in global developing-country arable-land area (double-cropped area counted twice), 38 percent shall be facilitated by expanded irrigation; and that 38 percent land-area increase shall generate 72 percent of the forecast increase in crop production. Within these figures there is accommodation that some irrigated land, notably in South Asia, shall become less productive because of irrigation-induced salinization, ground-water depletion, and water-logging, and that agriculture (including rice smallholdings) shall increasingly encounter responsible demands to divert scarce water to other uses (ADB 2001c, and World Bank in FAO 1997a: Plusquellec). In Asia, agricul-ture currently uses about four-fifths of freshwater withdrawals (Guerra et al 1998, FAO 2000c.)

FAO (2000d) suggests that during 2002-2030 the technical efficiency of irrigation-water distribution/allocation - the proportion of water released from the irrigation-system head-works that reaches the farmers' fields - may realistically be expected to increase from 38 percent to 42 percent in East Asia, and from 49 percent to 58 percent in South Asia. Somewhat contrarily, Guerra et al (1998, citing various sources) report that at 1990-91 such efficiency was in East Asia (Indonesia, Malaysia, Thailand) 35 - 65 percent, but in South Asia (India) only 30 - 40 percent. The concept of economic efficiency of water - and of its poverty and social components - is well presented by IFAD (2001 Box 4.10).

For such water as does reach the farmers' fields, the aspect of water productivity - which is the amount of produce generated per unit of water applied, is addressed in chapter 4.2.1 of this document. We here note that in relation to prospective irrigation-water savings at field level, IWMI (2000c) documents the substantial irrigation-water savings achieved in Zhanghe, China - during 1966-98, with progressively increasing rice yields - through the procedure of regular within-season field draining.

Such procedure may by 2030 be supplemented by the option to grow water-efficient "aerobic-rice" cultivars if such cultivars have by then been developed successfully for the upland ecozones. Such cultivars might be particularly appropriate as the second rice in an irrigated or rainfed rice-rice sequence. However, there is need to be mindful of the appreciable benefit to the productivity of current lowland-rice cultivars that derives from the rice-favourable soil-chemical regimes that result from prolonged soil submergence.

Relatedly, Woodhead et al (1994) reported that for the Haryana (India) rice-wheat sequences an appropriate combination of rice and wheat cultivars, of land preparation, and of irrigation regime, perhaps including conjunctive use of sweet and brackish waters in appropriately-drained lands, can achieve a 35 - 40 percent field-level irrigation-water saving (rice and wheat), with relatively small yield penalty. And for wheat in the rice-wheat area of the Pakistan Punjab, CIMMYT (2000a) reported that water-conserving procedures, involving preparatory laser-guided land levelling with subsequent minimum-tillage post-rice seeding of wheat into raised beds, have in farmers' fields demonstrated the capability to save 25 - 40 percent of field-level water requirement (compared to current practice) while attaining a 17 percent increase in wheat yield and in fertilizer-use efficiency.

These findings and forecasts suggest that it is reasonable to expect that there should during 2002 - 2030 be no need for any appreciable increase in water withdrawals for irrigation. However, IWMI (2000a, and 1998 cited by FAO 2000c) estimates that even if Asia-Pacific-Region agriculture does achieve considerable savings of irrigation water by Year 2030, the forecast increase by that time in domestic and industrial water requirements shall exceed the irrigation-water saving to such extent that the net requirement for diverted water shall increase by at least one-fifth.

Table 12 indicates that in the seventeen listed rice-producing countries irrigated-land area (all crops) and irrigated-riceland area (physical area - this author's estimates) are expectedly dominated by China and India - but also, for all irrigated crop-land by Pakistan. Irrigated-riceland area is substantial also in Indonesia and Viet Nam. Annual rates of irrigated-area increase (1988-99) were of order 1 - 2 %/ann for a majority of the seventeen countries, but appreciably exceeded 2 %/ann in Bangladesh, India, Laos, Myanmar and Nepal; however, these increases may not refer to riceland. For rice and particularly wheat, FAO (2000d) suggests that future increases in production shall derive more from cultivar improvement than from irrigation expansion.

3.4.3 Mechanization

Mechanization, as represented in Table 13 by the numbers of two-wheeled and four-wheeled tractors - totals aggregated - per 000 ha of agricul-tural land, is furthest developed in the extreme east (the two Koreas) and in the extreme west (Iran and Pakistan) and in Thailand and Viet Nam. Mechanized harvesting is increasingly practised in China, Malaysia, Thailand and the Indo-Gangetic Plains - often on lands that have been consolidated to benefit from contract-servicing of land-levelling, tillage, seeding, and harvesting. For many smallholder rice-system farmers in India, Kumar (personal communication, 2001) concludes that the lack of tractive power, whether gasoline- or animal-derived, delays field operations and lessens productivity.

The cost in Nepal (Rice-Wheat Consortium 2000b) of a Chinese manufactured two-wheel tractor with basic cultivation attachments is US$1 300, compared to an average Nepal income of US$200/person.ann. In Pakistan (Rice-Wheat Consortium 2000c), the cost of a locally-manufactured wheat-seed drill, which requires four-wheel-tractor power, is US$600. Other aspects of mechanization for post-rice tillage operations were introduced earlier (via Rice-Wheat Consortium 2000a).

[Mechanized operations necessarily have implication for the creation and continuance of remunerative and drudgery-free rural employment.]

3.4.4 Nutrient management

Appropriate integrated nutrient management in rice-based sequences and systems shall expect to provide or replenish the nutrients removed by the crops, and to increase the soil biomass and thereby improve "soil health", and to facilitate adoption of high-yielding crop cultivars (FAO 2000a). Table 13 therefore lists in two adjacent columns the annual rate of mineral fertilizer applications: N, P, and K nutrients in aggregate, for all crops - food and industrial - and representative for 1996-99 (FAO 2000b and 2001b), together with this author's estimate (adapting FAO 2000d, WAICENT) for the application of nitrogen to rice, as kg N/ha.ann, and regardless of whether to one, two, or three rice crops per year. These totals relate to manufactured/processed fertilizers only; nutrients from composts and manures would increase these numbers.

The N+P+K-aggregate (mineral) application rate exceeds 200 kg nutrients/ha per annum (not per crop) in China, RoKorea and Viet Nam; such rates are sufficiently high that there is risk of adverse environmental impacts. Moreover, previous national-scale analyses (Woodhead, Huke, and Huke 1994 Table 5) documented for rice in Bangladesh, India, Myanmar and Pakistan the strongly diminishing response to fertilizer (as kg rice grain/kg fertilizer nutrients) as national-average applications increased from 5 to 80 kg nutrients/ha.ann through 1965-75-80-87. However, for individual fields, long-term experiments indicated that the incremental yield response to increase of fertilizer declined only slightly - if at all - during fifteen years.

Current fertilizer-application rate (Table 13, all nutrients) is 40 kg nutrients/ha.ann or less in Cambodia, Laos, Myanmar, and Nepal. For all-India, (Singh, Kumar, and Woodhead 2002 refer) determined that at 1991/92 and for all crops in aggregate, irrigated land received about 110 kg nutrients/ha.ann irrespective of farm size (whether < 1.0 ha, intermediate, or > 4.0 ha), whereas non-irrigated land received about 40 kg nutrients/ha.ann on the smallest farms, but only 25 kg nutrients/ha.ann on the largest ones. In all countries, fertilizer applications in the floodprone/swampland, upland, and less-favourable rainfed lowlands are, expectedly, less than these national averages, but may be increasing in some ecozones - as in eastern India (IRRI-IFAD 2000).

Annual nitrogen-fertilizer applications per hectare of rice - with some imprecision because of uncertainty of rice area in multiple-rice-cropping regions - are for most of the listed countries comparable to the totals of N+P+K nutrients/ha.ann. This N-rate exceeds 160 kg N/ha.annum (not per crop-season) in China, Indonesia, Iran, RoKorea, Malaysia and Viet Nam. It is 30 kg N/ha.ann or less in Cambodia, Laos, Myanmar and Nepal: this low rate of N-fertilization is to some extent reflected (Table 13) in these countries' low rice yields (indicative for 1997-2000, and aggregated for rainfed and irrigated rice).

Table 13: Indicators for on-farm resources which facilitate rice-system production and livelihood: Asian rice-producing countries

[Notes: Sources are FAO (2000b and 2001b);? indicates doubtful or not reported. Regional groups: North-East Asia, South-East Asia, and South Asia. Some statistics pre-date the 1997-99 economic crises in North-East Asia and South-East Asia. Tractor numbers (expressed on a per 000 ha basis) include 2-wheeled and 4-wheeled models, and relate to 1996-98. Fertilizer per ha of agricultural land relates to all crops (food and industrial) and includes N, P2O5, and K2O, and is indicative for 1996-99; fertilizer per ha of rice (whether 1, 2, or 3 crops/year) is adapted from FAO 2000d. Rice yield (aggregate for irrigated and rainfed crops) is indicative for 1997-2000. Chicken-numbers growth rate (1989-2000) is here presented as an indicator of very high growth rates in the livestock sector.]

For rice-rice systems in southern and north-eastern India, Siddiq (2000) reports the incidence of P, K, and zinc deficiency, and of iron toxicity in some K-deficient areas. For rice-wheat cropping, Siddiq describes the progression through the Lower, the Middle, the Upper, and the Trans-Gangetic Plains in the applications of N and P - and also of K, Zn, and farm-yard manure. In aggregate, and for both rice and wheat, applications of N: P2O5: K2O are in the approximate proportions 6: 1: 1 - suggesting that applications of P and especially K are sub-optimal. This suggestion is supported by details - in Siddiq's Table 34 - of the farmer-field benefits, in production and in nitrogen-use-efficiency, of site-specific nutrient management.

Appropriate applications of K and especially P encourage the production of crop roots, and hence of soil biomass, and of ground cover - the latter particularly important for non-rice crops on non-submerged soil. Moreover, farmer-field research results (IRRI 2001a) indicate that for rice in rice-rice, rice-wheat, and rice-maize sequences, application of K within appropriate procedures of site-specific nutrient management can increase economic returns by 30 - 50 US$/ha.crop, and that management of N using those procedures can raise yield by about 10 percent, and - crucially - can increase N-recovery and hence N-use efficiency by 40 percent.

For Year 2020, IFPRI (1996) projects that actual applications of mineral nutrients (all developing countries in aggregate) shall be 122 Mt/ann, compared to a total of 185 Mt/ann that shall be required to ensure food security, and of 250 Mt/ann for sustainable resource use. More recently, FAO's (2000b 2000d) listings and forecasts suggest for global applications (all nutrients) to rice (predominantly Asia) at 1995/97, at 2015, and at 2030 as 22.2, 26.3, 27.6 Mt/ann, to wheat (globally) as 24.6, 27.6, 30.1 Mt/ann, and to maize (globally) as 22.3, 27.0, 30.3 Mt/ann. The totals for rice comprise N, P2O5 and K2O applications as 15.5, 4.6, 2.1 Mt/ann at 1995/97, as 18.3, 5.5, 2.5 Mt/ann at 2015, and as 19.4, 5.6, 2.6 Mt/ann at 2030.

The average compound rate of increase of N+P+K fertilizer applications, to all crops, during 2002-2030 is forecast FAO (2000d) for East Asia as 0.9 %/ann (or 0.6 %/ann on a per hectare basis) and for South Asia as 1.3 %/ann (or 0.8 %/ann on a per hectare basis). IFPRI (1996) estimated that of the total of fertilizers applied in Asia (excluding Japan) at 1995 about 70 percent was applied to food crops, as distinct from feed crops and industrial crops. For India, Kumar (personal communication, 2001) calculates that at 1991/92 the proportions of (total) N+P+K mineral nutrients applied to rice, to wheat, and to maize were respectively about 35 percent, 20 percent, and 3 percent; the proportionate applications of farm-yard manure were somewhat similar. Globally, and for nitrogen fertilizers: rice, wheat, and maize respectively received at 1995/97 about 19 percent, 20 percent, and 17 percent of the global total of N, and are each projected (FAO 2000d) to receive about 18 percent of the (increased) global total at 2030. Myanmar, and Pakistan, and probably for Cambodia and Nepal, and possibly for Sri Lanka, the incidence of low birth weight is worryingly high.

Because land in rice-producing zones is already used very intensively, and the challenges to achieve food security shall continue to be substantial, it is probable that in rice-based cropping land scarcity shall constrain local opportunities to produce plant- or large-animal-derived ("organic") fertilizers. Thus even in India, with its substantial livestock-population totals and densities, the application of farmyard manure on marginal-size holdings (< 1.0 ha) at 1991/92 (Singh, Kumar, and Woodhead 2002) was only 3.8 t/ha.ann on irrigated land and 1.6 t/ha.ann on non-irrigated land, and that on farms larger than 4.0 ha these application rates decreased to 2.0 t/ha.ann and 0.6 t/ha.ann.

However, poultry manure, composted cattle manure, and piggery slurry and pig manure are known to be effective sources of nutrients - particularly of nitrogen, phosphorus, potassium, and calcium (Dobermann and Fairhurst 2000 Table 4). These materials shall expect to become more plentiful as livestock populations increase in mixed-farming systems and in industrial livestock-production systems in many rice-growing countries.

Overall, efficient nutrient management in rice systems shall require well-designed and well-implemented policies and programmes for fertilizer supply/distribution/regulation, for prices, incentives, training, credit, and infrastructures, and for environmental monitoring.

[On all these aspects, FAO (2000a) through its Land and Water Development Division has the mandate, expertise, and ongoing activities, including activities within the Special Programme for Food Security, to assist member countries.]

3.4.5 Livestock resources

The importance of the livestock resource is featured in Table 13 by the growth rates (1989-2000) in the numbers of chickens, whether for meat or for eggs. Only for DPRKorea and Sri Lanka, and possibly India and Indonesia, did these growth rates fail to exceed 3.0 %/ann. For several countries they were substantially higher: approaching or exceeding 6.0 %/ann in China, Malaysia, Myanmar, Pakistan, Philippines and Viet Nam. For all developing-country Asia, annual growth rate for chicken population was 5.0 %/ann, and for duck population 5.5 %/ann. However, Dixon et al (2001) caution that most of these population growths derived from large-scale enterprises and not from mixed-system smallholdings. Moreover, poultry population growth rates are forecast to be appreciably lower during 2002-2030. For cattle, buffalo, pig, sheep, and goat, the Asian-developing-country populations increased respectively by 1.4, 1.0, 1.9, 1.4, and 3.1 %/ann during 1989-2000 (FAO 2000b, 2001b).

For India, IFPRI (1999b) suggests that the high growth rates for poultry and ruminant populations bring corresponding requirements and opportunities for the small-scale and large-scale production of feedstocks, and corresponding concerns for disposal of excreta. However, in all rice-growing countries, if the livestock excreta are produced in proximity to ricelands, then such excreta - whether processed on-farm, or by specialist enterprise, or not at all - may be used as a component in the nutritional management of rice-system crops. Indeed, FAO (1999c) reports that on Bangladesh smallholdings the value of the manure constituted two-fifths of the value of large-ruminant production. Correspondingly, by-products and waste and pest products from rice-system crops may be converted, either on-farm or in village-scale enterprises, into feed supplements for rice-system livestock.

[Such crop-livestock synergy features as one of seven thrusts to foster FAO's Medium-Term Plan (2000a) Production-Systems Priority-Area Inter-disciplinary Action.]

3.5 Economics and trade

3.5.1 Rural and national economies

Economic and trade indicators for the seventeen countries are presented in Tables 14 and 15. Values for Gross National Product (GNP) per person, at 1998, and hence mid-crises, exhibit strong variation (Table 14): from below US$ 400/person.ann (barely US$ 1.00/person.day) in Bangladesh, Cambodia, Laos, Nepal, and Viet Nam to almost US$ 4 000/person.ann in Malaysia and to more than US$ 8 000/person.ann in RoKorea.

Table 14: Economic and agro-economic indicators: Asian rice-producing countries

[Notes: Sources are FAO (2000b, 2001b);? indicates doubtful or not reported. Regional groups are North-East Asia, South-East Asia, and South Asia. Some statistics pre-date the 1997-99 economic crises. GNP per person is at 1998, and GNP annual growth rate for 1990-98. Percentage change in agricultural-population total is for 1989-99: increases indicate countries where pressures for land fragmentation may become severe. Agricultural population/total population is at 1999; agricultural GDP/total GDP at 1998-99.]

GNP growth rate (1990-98, pre-crises) also ranges widely: from 2 %/ann or less in Iran, DPR Korea, Pakistan and Philippines, to 4 %/ann (or more) in India, Indonesia, RoKorea, Malaysia and Thailand, and to 6.5 %/ann in China and Viet Nam. Worryingly, for many listed countries - notably South-Asian countries - growth rate achieved during 1990-98 was below the 5 %/ann value that shall in future be needed if poverty is to be eradicated.

It is thus reassuring that at April 2002 the Asian Development Bank forecasts for Years 2002 and 2003 average annual GDP growth rates for India of 6.4 percent and 5.1 percent for Bangladesh; but for Sri Lanka and Pakistan only 4.5 and 4.0 percent respectively. Corresponding South-East-Asia fore-casts include 5.0 %/ann for Malaysia and 4.3 %/ann for the Philippines, but only 3.3 %/ann for Indonesia and 2.8 %/ann for Thailand (Table 15 refers)

For the rice-system farm families, and for the associated landless labourers, it is noteworthy that the real prices of both rice and wheat have declined during recent decades. The profitability of rice-sequence cropping probably declined correspondingly in some countries and ecozones - notwithstanding decreases in production costs and increases in yields. Such declines are greater where support prices are so low as to constitute a disincentive to rice-systems farming and a consequent incentive to the adoption of higher-profit, often horticultural, systems - as in peri-urban China following agrarian liberalization. Pro-hungry, pro-poor interventions must ensure that rural small- holders and landless families share in the benefits of these higher-profit systems while retaining the food-security insurance of their staple-food supplies.

Indeed, profitability and production have usually increased where there have been supportive government interventions - as through appropriate pricings and by expansion of irrigation and mechanization and market and credit facilities and of seed- and agrochemical-supply systems and by moderation of tenancy law (Woodhead, Huke, and Huke 1994; Singh and Paroda 1994; Singh 1997). However, such interventions have generally brought greater benefit in irrigated than in rainfed ecozones - and it is the latter ecozones that may be home (interpreting Table 2) to the larger proportions of poor and hungry.

Values for the ratio (at 1999) of agricultural population to total population are in Table 14 presented as a proxy for the percentage of population that resides rurally. This proportion is below 20 percent for the countries with higher GNP per person (Malaysia and RoKorea), and approaches or exceeds 70 percent in six of the listed countries: Cambodia, China, Laos, Myanmar, Nepal and Viet Nam.

The percentage change in agricultural population (1989 to 1999) is also presented in Table 14. For six countries this change exceeds + 10 percent. Comparison with Table 12 indicates that three of them (Laos, Nepal, and Viet Nam) already (at 1999) had high ratios for agricultural population to agricultural-land area. It is probable that in these countries, and in India and Pakistan also, many additional persons shall seek to become owners and/or tenants of part of the finite quantity of farmland.

Three data columns in Table 14 indicate the status of the agricultural (rural) economy in relation to the total (national) economy. The ratio of agricultural GDP to total GDP at 1998-99, immediately following the East-Asian economic crises, is notably high (³ 40 %) in four predominantly rice-producing countries (Cambodia, Laos, Myanmar, and Nepal) and is substantial (» 25 %) in India, Bangladesh, and Pakistan.

Crucially - in relation to interventions to relieve rural poverty and undernourishment and to strengthen rural enterprise and employment - the ratio of agricultural employment: total employment is for all countries seen to exceed, and most-often very substantially, the ratio of agricultural GDP: total GDP - implying a need for redirection of resources, investments, and policies if rural impoverishment and under-employment are to be lessened. It is thus pertinent to reiterate that the cost of a rural workplace is substantially less than the cost of an urban workplace.

IFPRI (1999b) expresses strongly - in relation to India's cereals farmers - this need for a policy redirection, and for a review of current subsidies. The more-general need for a redirection of resources and policies is emphasized, and the attendant difficulties highlighted, by the relativities (Table 15, first column) of the growth rates (1990-98) in agricultural GDP and in total GDP: the latter growth rate is in all countries much higher than the former.

[To assist rice-growing member countries to make such redirections, the FAO (2000a) Medium-Term Plan indicates that the Divisions of Commodities and Trade and of Policy Assistance (headquarters and regional) have the required expertise and ongoing programmes.]

Table 15: Economic and trade indicators: Asian rice-producing countries

[Notes: Sources are FAO (2000b, 2000i, 2001b) and ADB (2002);? indicates doubtful or not reported. Regional groups are North-East, South-East, and South Asia. Some statistics pre-date the 1997-99 economic crises. Historic growth rates for agricultural and total GDP at 1990-98. Forecasts growth rates for total GDP are for 2002 and 2003. Imports and exports as US$/person.ann, indicative for 1996-98. International agricultural trade/total international trade (per country, as %) is for 1991-93 and 1994-96 (pre- and post WTO inception); country-specific shares of global international agricultural trade for 1994-96 and for 1991-93 are here expressed as a ratio.]

In the first data column of Table 15, the second number within each pairing quantifies the 1990-98 historic growth rate in Total Gross Domestic Product (GDP). The second data column presents very recent informal forecasts (ADB 2002) for the GDP growth rate at Years 2002 and 2003 for some of the listed rice-growing countries. For 2003, for some countries these estimates approach or exceed 5.0 %/annum; such growth rate is sufficiently high as to permit cautious optimism that there can be additional national resources wherewith to combat rural hunger and poverty. Thus, among the several rice-growing countries that have high incidence of hunger and/or poverty (Table 9), nine are forecast to have at 2003 a GDP that shall be 5.0 percent higher than in 2002 (Table 15). Those nine are: Bangladesh, Cambodia, China, India, Nepal, Pakistan, Philippines, Sri Lanka and Viet Nam.

3.5.2 Rural taxation

Accusations of anti-rural bias in regimes of taxation and fiscal policy are regularly made (World Bank 2001a, IFAD 2001). FAO (1998b) queries whether in some instances this bias is actually anti-smallholder rather than anti-rural (anti-agriculture). Conversely, FAO (1998b Annex 3) marshals incisive evidence of anti-agriculture/pro-industry bias in China: there is there substantive quanti-tative evidence of a strong government-expenditure bias against agriculture throughout 1965 to 1996 - and continuing (despite policy changes) at 1998.

Furthermore, the Chinese rural sector has since 1985 been subject not only to agricultural tax and agricultural fee but also to tax on Township and Village Enterprises. The rural sector thereby makes a very substantial contribution - which at 1996 was still increasing - to the urban sector and its industry. Additionally, input- and output-price policies - including high taxation on rice, and fertilizer-price control - imposed a further "40 percent disprotection against agriculture" at 1993-94 (World Bank via FAO 1998b). Thus, despite good intentions of the Chinese government, "there is disprotection to farmers resulting from procurement/quota/pricing policies". Similar situations may prevail in other rice-producing countries.

Thus FAO (2000i) opines that "in several countries, taxes on agriculture are a substitute for more-imaginative taxes - such as income tax". And suggests (FAO 1998b Annex 1) that even within the agricultural sector, "subsidies - on fertilizer, water, tubewells, electricity, credit - are crowding out necessary investments". Correspondingly, for East Asia, the World Bank (2001a) Regional Strategy Objectives give specific priority to the removal of anti-rural biases.

There are, however, measures with which countries have sought to counter these biases: FAO (1998b Annex 1) reports that macro-economic policies in some South-Asian countries have been successfully directed to removing some pro-industry tariffs, quotas, and licences that had dis-advantaged their rural/agricultural sectors. National policies of low inflation, of realistic currency-exchange rates, and of low human-population growth, similarly have operated (ADB 2000a) to the benefit of rural (and smallholder) interests. Thus, post-crises depreciations of currencies in some North-East-Asian and South-East-Asian countries may have lessened much of that part of the pre-crises taxation that resulted from over-valued exchange rates (ADB 2001a).

[It is thus appropriate to indicate that FAO (2000a) has the technical and legal experience, expertise, mandate, and programmes to assist rice-growing member countries to assess, and if appropriate to strengthen, amend, and/or repeal, their several policies that promote or hinder food security, rural poverty/human nutrition, and agricultural production/processing/pricing and enterprise.]

3.5.3 Agricultural imports and exports

Table 15 provides, in three contiguous data columns, an insight concerning the rice-growing countries' international agricultural trade and the implications for that trade of the World Trade Organization's procedures and operations. Agricultural imports and exports - expressed in US $/person.ann, and representative for 1996-98, and for food crops and industrial crops in aggregate - indicate for each of the seventeen listed countries the extent to which the country is a net importer or exporter of agricultural products.

Major net agricultural-product importers are Bangladesh, Cambodia, Iran, DPRKorea, RoKorea, Nepal, and Pakistan. Major net exporters are Malaysia (industrial crops) and Thailand and Viet Nam (food crops). Agricultural imports and exports are more-nearly in balance for China, India, Indonesia, Laos, Myanmar, Philippines and Sri Lanka.

For the net agricultural-produce importers, interventions to strengthen within-country comparative advantage to produce, and thence to substitute, high-cost imported food items could be worthwhile. It is thus noteworthy (Singh 2001b Table 15, adapting FAO 2000d) that at 2015 and at 2030 East Asia is forecast to make net imports of 60 and of 79 Mt/ann of (all) cereals, and South Asia similarly 16 and 26 Mt/ann.

For rice, Asian rice-growing countries' imports and exports are of similar magnitude. Major (consistent 1995-99) rice exporters are India, Pakistan, Thailand, and Viet Nam. Rice importations into rice-growing countries are episodic - responding to individual years' climatic and other circumstances; Indonesia, Iran, Malaysia, and the Philippines each imported substantially during 1995-2001.

For wheat and wheat products, all seventeen (Table 15) countries, including wheat-growing countries, are consistent and substantial net importers - though imports were less during the years of economic crisis.

For all cereals in aggregate, only India, Thailand, and Viet Nam are net exporters. During 2002-2015, rice-growing countries' requirements for imports of wheat, of coarse grains, and of livestock products are all forecast to increase appreciably. There shall of course be need that the importing countries shall have the resources wherewith to purchase the forecast imports. The major suppliers of the imports are expected to be the Americas (Argentina, Canada, USA), Australia, and various countries of Western Europe.

3.5.4 Global trade implications

The World Trade Organization (WTO) became effectively operational at 1994/95. Two Table-15 columns of data (derived from FAO 2000i, quoting Asian Development Bank) permit preliminary country-specific comparisons of patterns of international agricultural trade before (1991-93 triennium) and after (1994-96 triennium) WTO inception. [IFAD (2001 Table 5.1) presents similar sample data from continents additional to Asia.]

A first comparison (from the Table-15 data) assesses for each specified country the percentage ratio of agricultural trade to total trade at 1991-93 and at 1994-96. So far as these preliminary data permit any interpretations, it is suggested that for those countries for which any change between 1991-93 and 1994-96 was discernible, and excepting Cambodia and DPRKorea, such change indicates a decrease in agriculture's proportion of the country's international trade. Such decrease implies that agricultural trade increased less rapidly than did non-agricultural trade.

Within global agriculture, however, a more definite pattern emerges: the ratio for 1994-96 compared to 1991-93 shows that for most Table 15-countries (and excepting only DPRKorea, Sri Lanka and Thailand) the share of international agricultural trade - imports plus exports - increased appreciably after WTO inception. For Asia as a whole, agricultural trade increased x 1.37 between 1991-93 and 1994-96, compared to a global agricultural-trade increase of only x 1.25. However, for most Asian countries during 1991-96, agricultural imports increased proportionately more than agricultural exports (FAO 2000i).

Correspondingly, an Association of World Council of Churches survey (1999) concluded that national structural-adjustment programmes, which possibly included trade liberalization components, worsened the food security of poor rural families. Similarly, IFAD (2001, citing various sources) reports that in China, India, and Philippines, increased food prices following liberalization impacted adversely the rural poor. Dixon et al (2001) suggest that in developing countries the recent trade liberalization has favoured the urban consumers at the expense of the rural producers, and that such impressions are sustaining a "A profound unease among developing-country governments concerning the wisdom of trade liberalization".

Contrarily, the World Bank (2001b - drafted after 11 September 2001) reports that 24 developing countries, with aggregate population of 3 billion persons, that during 1970-90 increased their integration into the world economy have achieved higher growth in incomes, longer life expectancy, and better schooling. Conversely, other countries, of aggregate population 2 billion persons, that did not increase their integration into the global economy have experienced contracting economies, meagre improvements in education, and increased poverty.

The geographic patterns of the rice-growing countries' international trade - and of trade within the regional groupings of ASEAN (Association of South-East Asian Nations), SAARC (South Asian Association for Regional Cooperation), and North-East Asia - exhibit several features that may give insight in relation to prospective interventions.

Thus, adapting FAO 2000i, representative proportional statistics for 1994-1996 for agricultural food commodities, products, and feeds indicate that of ASEAN's exports about 30 percent went to Asian countries, and less than one-third of that 30 percent to other ASEAN countries, and 70 percent went outside Asia. Of SAARC's exports, about 50 percent went to Asian countries, and about one-half of that 50 percent to other SAARC countries, and 50 percent went outside Asia. For North-East Asia, including Japan, about 20 percent of exports went to Asian countries, and about two-thirds of that 20 percent to other North-East-Asian countries, and 80 percent went outside Asia.

For agricultural imports: for ASEAN, about 30 percent came from Asian countries, with about one-half of that 30 percent from other ASEAN countries, plus rather less than 20 percent from Pacific countries, and 50 percent from outside Asia - notably from developed countries. Of SAARC's imports, somewhat less than 20 percent came from Asian countries, and less than one-half of that 20 percent from SAARC countries, plus 10 percent from Pacific countries, and rather more than 70 percent from outside Asia - predominantly from developed countries. For North-East Asia, about 20 percent of imports came from Asian countries, and about one-half of that 20 percent from North-East-Asian countries, plus 10 percent from Pacific countries, and 70 percent from outside Asia - notably from developed countries. The dominance of agricultural trade beyond Asia and with developed countries is apparent.

Statistics for absolute trade, in monetary value, are even more revealing: UK/DFID (2000) and FAO (2000b, 2001b) indicate that during 1997-99 the agricultural exports from the whole of South Asia, with population 1.3 billion persons, were marginally less in value those from Thailand, of population only 60 millions. Thailand's export success reflects in part a commitment to high-quality added-value products.

One interpretation of this foregoing contrast is that in Asia there are substantive constraints to the export of agricultural commodities and products - but that such constraints can be overcome. The several ADB-initiated sub-regional growth triangles - as in the contiguous areas of Indonesia, Malaysia, and Thailand - may here have a dynamic role complementary to that of the much-larger regional groupings. There might similarly be commonalities of interest between Bangladesh and India's West Bengal, or between Nepal's Terai and India's Bihar and Eastern Uttar Pradesh. The extent to which global and regional agricultural trade can - and cannot - help lessen hunger and poverty in rice-growing ecozones within the medium term and within the near term is here addressed in the following paragraphs.

3.5.5 Global trade: "The level playing field"

Thus, there is now growing awareness - as exemplified by the intentions in UK/DFID 2000 and in World Bank (2001b) and by the global commitment to International Development Targets - that globalization of trade can and must be made to work for the poor and hungry, and that elimination of poverty is in the interests both of developed and of developing nations. This awareness extends among various well-motivated governments, agencies, and civil-society groups - including the "faith groups" in whom poor people have highest confidence.

Moreover, some developing-country governments, and their agricultural and commercial sectors, consider that their countries' food-security interests and their comparative advantages are best served by policies of food self-reliance (ability to produce or purchase all required food) rather than of food self-sufficiency (ability to produce all required food). Other countries, however, including China, the populous countries of South Asia, and non-petroleum-producing countries especially, consider that "a large degree of food self-sufficiency is desirable to ensure food security" (FAO 1998b).

But all developing-country - and others' - opinions are unanimous and justified in insisting that global trade must be conducted on a level playing field - implying that global trade must be open, equitable, accountable, and rules-based, while allowing country-specific choices within those rules.

The benefits of agricultural trade in lessening poverty and food insecurity are that international trade promotes the general economic and income growths within a participating nation that are necessary to enable pro-poor and pro-equity and pro-employment policies to achieve their intended objectives. FAO (2000i) suggests that international agricultural trade can augment national domestic food supplies and lessen their seasonal and their year-to-year variabilities, and can foster economic growth and profitable product specialisation.

Conversely, there are attendant risks, well formulated in ADB 2001a, of uncertain supplies and of variable prices and the lack of resources wherewith to pay those prices. And particularly there is risk (FAO 1998b, Annex 1) that international-market prices for cereals are likely to increase - though the experience of 1999-2000 may in the near term gainsay this forecast.

Moreover, within countries, trade liberalization shall create both "gainers" and "losers" among regions, among sectors, among producers, and among consumers. Explicitly, IFAD (2001) cautions that in countries with high levels of inequality the benefits of trade liberalization are garnered by the rich - through abuse of special-access privileges - with minimal benefit to the poor.

The level-playing-field concept requires that there shall be social and economic "safety nets" wherewith to protect and assist the "losers". Such WTO-compatible safety nets - which shall require resources - include state procurement and private-sector "futures" contracts for commodities at guaranteed prices, subsidies (not crop-specific) for credit and inputs and training, and "Green-Box" general services. Such services include food aid and buffer stocks, irrigation and market (including land-market) and other rural-infrastructural supports, phytosanitary and food-quality services and communal pest management, agricultural and veterinarian research/extension, decoupled income and insurance support, and export prohibition.

That the present WTO "playing field" is not level, particularly for agriculture, is manifest in several instances - instances which feature both developing and developed nations. For developing rice-growing countries, and despite the fact that the ASEAN and SAARC regional groupings have been operational for many years, the tariffs and disincentives to within-Asia international agricultural trade are considerable. Thus the South Asia Free Trade Area and the South Asia Preferential Trading Agreement (each within SAARC) and the Asian Free Trade Agreement and the Common Effective Preferential Tariffs (each within ASEAN) generally exclude agricultural commodities and many agricultural products. Strong barriers to agricultural trade thus persist within those trade groupings.

Additionally, there is minimal synergy between regional groupings. Moreover, Asian countries are often competing with each other to export similar ranges of commodities to the industrialized countries. Also, within some Asian countries, vested interests, of producers or of labour, are often successful in opposing the liberalization of agricul-tural trade; and as Singh (2001b) cautions: "trade-liberalization and food-security policy platforms are in many developing countries not well-defined".

However, it is developed-world trading con-straints that most impact on the rice-growing countries' agricultural exports and on their farm families. Thus, notwithstanding that the General System of Preferences (GSP) intends that developed countries shall provide non-reciprocal preferences to developing countries, many developed countries do confront developing-world agriculture with formi-dable tariffs and own-farmer subsidies, and with unjust phytosanitary and quarantine restrictions. FAO (2000i) reports that OECD countries operated tariffs (in 1995) on developing-countries' rice, wheat, and maize at the rates of 89, 94 and 90 percent. However, the EU has announced its abolition by 2009 of its tariff on rice from developing countries.

Moreover, the world-average tariff for non-agricultural products is 4 percent: for agricultural products it is 40 percent. UK/DFID (2000, quoting EU sources) estimates that if tariffs, both of developed and developing countries, were decreased by one-half, then developing countries would in aggregate gain US$ 150 billion (B) - about three times greater than the total OECD development-aid flows.

In terms of subsidies, the developed-world farmers and the developing-world farmers are clearly on different playing fields: FAO (2000i citing OECD 1996) reports that at 1995 the transfers, via export subsidies and otherwise, to each full-time-farmer-equivalent amounted to US$ 19 600 in the European Union, to US$ 29 200 in USA, and to US$ 32 800 in Japan. These transfers per farmer vastly exceed the total of income and subsidies received by any deve-loping-country rice-farm family (GNP/per person is listed in Table 14). OECD farm-export subsidies in aggregate at 1995 were about US$ 360 B - seven times the total of OECD development assistance.

Developing-world farmers have concern also, as expressed in a late-2000 Forum of the ASEAN Poultry-Producers' Association, that in addition to export subsidies to compete against local producers in developing-country markets, developed-world exporters may be resorting to unfair labelling to win market share. More encouragingly, FAO (2001c) reports that WTO procedures have facilitated increased use of anti-dumping and of countervailing duties on meat (including poultry meat) - though not specifically in Asia.

3.5.6 Global trade: Industrialized countries and aid for developing-world agriculture

It is particularly noteworthy that several developed-world (industrialized) countries are substantial exporters of agricultural and of non- agricultural products to developing-world countries - including Asian countries. In aggregate, about 20 percent of all EU exports, and 40 percent of all USA exports, and 50 percent of US agricultural exports, are bought by developing countries.

The importance of these exports to the industrialized-country economies was highlighted by the 1997-98 East-Asian economic crises. Thus, IFPRI (1998c) estimates that, at 1998, US exports, including agricultural exports, to Asia decreased from US$19 B to US$5 B, and that there were similar impacts on the EU and on the Australian economies. EU and USA cereals exports each decreased by 7 percent (IFPRI 1998b). Similar effects may be expected as a consequence of the terrorism of 11 September 2001.

IFPRI (1998c) also calculates that each US$ 1.0 increase in developing-world farm output results in an additional US$ 0.73 of imports - of which US$ 0.17 are agricultural imports - pre-dominantly from developed countries. Such increase in developing-world farm output concomitantly increases rural employment - particularly women employment - and helps lessen poverty.

IFPRI (1998c) correspondingly rationalizes that - for all concerned - aid for agriculture is much preferable both to food aid and to crisis-relief aid, and that developed-world governments therein have a crucial role and responsibility. It is thus encouraging to recognize the mutuality of interest between developing-world agriculture and a major developed-world agricultural agency - the Foreign Agricultural Service (FAS) of the US Department of Agriculture. Within its many activities, FAS includes the management of international agricultural financial assistance programmes. Such programmes (FAS 2000) are undertaken towards the FAS Mission of "Serving US agriculture's international interests by expanding export opportunities for US agricultural, fish, and forest products, and by promoting world food security".

Thus, overall, and despite concerns for tariffs and subsidies, UK/DFID (2000) and ADB (2001a) counsel that the General System of Preferences (GSP) affords the most appropriate route whereby the developing-world farmers can increase their share of global agricultural markets and maintain or increase their share of their domestic markets. Correspondingly, the World Bank (2001b) proposes a seven-point "holistic" strategy wherewith developing countries can be assisted to access the benefits of globalization and to manage the attendant risks.

Relatedly, UK/DFID (2000) recognize that the GSP is difficult to comprehend and operate, and that some countries have cultural anxieties concerning the WTO. It recognizes also that many smaller countries lack the human resources wherewith to participate effectively in the GSP and WTO, but notes that the developing countries command a substantial majority in the WTO and its committees, and that there is an Advisory Centre on WTO Law that assists poor countries to present cases under the WTO Dispute-Settlement Procedures. Committees of particular importance to developing countries include the Committee on the Provisions of the Technical Barriers and Trade (TBT) and the Committee on the Provisions on Application of Sanitary and Phytosanitary Measures (PSP). UK/DFID therefore recommends programmes of technical assistance and training to enable the developing world to use its collective strength to achieve desirable modifications of the GSP and WTO procedures - and to "level the playing field".

Within the Asian rice-growing countries, agricultural trade is variously constrained on both the supply side and on the demand side. On the supply side - and excepting the consistently effective exporters India, Malaysia, Pakistan, Thailand, and Viet Nam - the constraints of increasing population and of domestic food demand, and the historic low level of investments for commodities other than rice and wheat, have determined that despite some comparative advantages few of the seventeen rice-growing countries can consistently produce marketable surpluses of export quality - either of food-crop or of livestock products. Moreover, prospective exporters lack supports of export credits, of rural infrastructures, and of working/investment capital - against which there are more-attractive investment options. On the demand side, and for exports to developed or developing countries, the major constraints comprise tariffs and subsidies, some importers' discriminatory preference for particular suppliers, and the lack of low-cost distribution channels for small-volume producers.

[FAO (2000a) through its Divisions for Commodity and Trade, Policy Assistance, and Research, Extension, and Training has expertise, mandate, and ongoing activities - with training courses and manuals for TBT and PSP - to assist member countries to identify and address their various agricultural-trade concerns and opportunities. One of the five FAO Asia-Pacific-Region integrated programme thrusts features "World Trade Organization: Capacity building, multi-lateral trade, and an enabling policy environment".]