Journal of Internationat Devetopment: Vol. 6, No. 6, 665-688 (1994)

POLICY AND TECHNOLOGY FOR RICE PRODUCTIVITY GROWTH IN ASIA MARK W. ROSEGRANT International Food Policy Research Institute, Washington, DC, USA

and PRABHU L. PINGALI International Rice Research Institute, Los Bafios, Phillippines.

Abstract: In the past decade, declining rice prices, a slowdown in research expenditures and output growth, reduced irrigation investment and degradation of irrigated land, declining marginal returns to input use, and a stagnant technological yield frontier caused declining growth in rice yields per hectare in much of Asia. Future growth in rice productivity will increasingly come from improved management and efficiency of use of the resources utilized in rice production, in contrast to the rapid dissemination of modern technology which has been dominant in the past. The foundations for increased efficiency in rice production are greater investment in research, extension, and education to upgrade human capital, combined with establishment of economic incentives that reflect the social opportunity costs of scarce resources.

1 INTRODUCTION The introduction and rapid spread of high-yielding rice varieties throughout Asia in the late 1960s and the early 1970s resulted in strong output growth (Herdt and Capule, 1983; Dalrymple, 1986). Aggregate rice output growth rate for Asia increased from 2.6 per cent per annum during 1958-66 to 3.3 per cent per annum during 1966-82. Rapid yield growth from 1966 to 1982 was the primary contributor to rice output growth. Area expansion contributed about one-third of Asian rice output growth in the 1960s and one-fifth in the 1970s. However, after 1982 the growth in aggregate rice output declined to 2.2 per cent per annum. Area expansion, which was already slow in the 1970s, virtually halted in the 1980s. Rice yield growth in Asia also declined sharply in the 1980s, from an annual growth rate of 2.9 per cent in the 1970s to 1.9 per cent during the period beginning in 1982 (Table 1). Yield gains declined by more than half in South-East Asia and China, which were the earliest adopters of modern rice varieties.^ 0954-1748/94/060665-24 by John Wiley & Sons, Ltd

666

M. W. Rosegrant and P. L. Pingali

Table 1. Rice: annual growth rates of area, production, and yield in Asia from 1957-59 to 1988-90. Countries/ regions

1957-59 to 1988-90

1957-59 to 1965-67

1965-67 to 1973-75

1973-75 to 1981-83

1981-83 to 1988-90

Total Asia Area Production Yield

0.73 3.08 2.36

0.85 2.60 1.74

1.09 3.37 2.27

0.24 3.09 2.86

0.25 2.16 1.91

South-East Asia" Area Production Yield

0.93 3.24 2.32

1.73 3.17 1.46

0.35 3.29 2.94

1.51 4.28 3.22

0.72 2.29 1.57

South Asia^ Area Production Yield

0.89 2.33 1.45

1.26 3.13 1.89

0.61 1.63 1.02

0.88 2.57 1.71

0.25 2.31 2.03

China Area Production Yield

0.52 3.55 3.03

-0.58 2.62 3.21

2.25 3.92 1.68

-1.07 2.98 4.06

-0.38 1.25 1.63

India Area Production Yield

0.67 2.49 1.81

1.21 1.95 0.74

0.74 2.90 2.15

0.46 2.22 1.57

0.34 3.62 3.23

•"South-East Asia includes Burma, Indonesia, Kampuchea, Laos, Malaysia, Philippines, Thailand and Vietnam. ^South Asia includes Bangladesh, Nepal, Pakistan and Sri Lanka but excludes India. Source: IRRI World Rice Statistics, 1990.

Comparative information is lacking for long-term trends in total factor productivity in rice production because time series data are not available for input use on specific crops. The limited available evidence on recent trends in total factor productivity growth for rice is also indicative of a slowdown in growth. In Indonesia, the growth rate in total factor productivity for rice was 4.3 per cent from 1974 to 1982, but declined to 1.4 per cent from 1982 to 1990 (Pardey et al, 1992). Rice yields per hectare in Indonesia grew at about 4.5 and 1.9 per cent per year respectively over these two periods. In India, where rice yields did grow more rapidly in the 1980s, the increase in total factor productivity growth was much ' Rice yield growth actually increased in India during the 1980s. However, the strong growth in India in the 1980s may not be a cause for optimism for continued rapid growth in the 1990s. The northern and western regions were the earliest adopters of modern seed-fertilizer technology in India, and these regions showed the characteristic decline in yield growth in the 1980s. The relatively slow-adopting southern rice regions entered a more rapid growth period in the 1980s (Sarma and Gandhi, 1990). This strong regional growth more than maintained the national average yield growth. However, there are some signs that growth is now slowing in these regions. This pattern of strong regional growth followed by a slowdown as new technology is adopted and then matures is a common pattern in Asia (Hayami, 1975; Bouis, 1990a).

Rice Poticy and Technotogy 667 smaller. Total factor productivity on rice in India grew at 0.75 per cent per year, 1971-80, and at 0.77 per cent per year, 1981-88, compared with rice yield growth of 1.61 per cent and 3.17 per cent in the respective periods (Kumar and Rosegrant, 1993). Demand for rice in Asia is estimated to grow at the rate of 2.1-2.6 per cent per year until 2005 (IRRI, 1989; IFPRI, 1992). This demand growth is anticipated despite the slowing contribution of the income-induced component in demand growth. The slowdown in income-induced demand for rice is caused by declining income elasticities of demand with rising per capita income and the shift of population to urban areas in much of Asia (Bouis, 1989, 1990b). Even accounting for this slowdown, projected demand increases are equal to or higher than the growth in rice production in the 1980s. Given the shifts in the structure of production and demand growth, are there economically efficient technologies and policies to sustain rice productivity growth in the coming decades? This paper reviews trends in rice prices, research expenditures, irrigation investment, the yield frontier and fertilizer use; examines policies and technologies for efficient rice productivity growth; and concludes with a brief summary of recommended policy directions.

2 WORLD RICE PRICE-INDUCED DECLINE IN PRODUCTIVITY GROWTH Virtually all of future growth in rice production must come from increased rice yield per unit of land since, as indicated by the cessation of area growth shown in Table 1, the opportunities for further area expansion are minimal. Both aggregate and field-level data indicate that the growth in rice yields has slowed and there is a danger of continued declines in yield growth, especially in the irrigated lowlands of Asia (Pingali etal., 1990). The slowdown in rice productivity growth in Asia since the 1980s has been caused by declining world rice prices and by factors related to the process of intensification of rice production. The long-term decline in the world rice price has resulted in reduced investments for irrigation infrastructure and rice research. At the same time, increased intensity of irrigated land use has led to increasing input requirements in order to sustain current yield gains. Figure 1 shows the long-term decline in real world rice prices, a decline which sharpened in the 1980s. The declining price of rice has caused a direct shift of land out of rice production and into more profitable cropping alternatives, and has slowed the growth in input use, and therefore yields. This shift into more diversified cropping, while an appropriate farmer response to changing incentives, puts greater pressure on productivity growth in existing rice areas. Probably more important in the long run, the declining world price has caused a slowdown in investment in rice research and irrigation infrastructure. Declining Growth in Research Expenditure and Output Time series data on crop-specific rice research expenditure by country are not available. There was, however, a sudden decline in growth in total agricultural

668 M. W. Rosegrant and P. L. Pingali

Price (US$/MT) 1.000

800 -

600 -

400 -

200 -

1950

1955

1960

1965

1970

1975

1980

1985

1990

Year

Eigure 1. Real world rice price, 5% brokens, free on board (EOB) Bangkok, 1950-91 (1985 prices).

research expenditure and personnel in Asia in the late 1970s and 1980s. The annual growth in agricultural research personnel dropped from 6.9 per cent during 196878 to 4.0 per cent during 1978-83. Agricultural research expenditure grew at a rate of 6.9 per cent during 1968-73, but declined to about 4.6 per cent during 1978-83 (Pardey et ai, 1991). Although it is likely that Asian rice research expenditure has declined at least in proportion with agricultural research expenditure, given the increasing interest of governments in crop diversification, clear evidence is available only for Indonesia. Pardey et at. (1992) showed that rice research expenditure in Indonesia has declined more rapidly than total agricultural research expenditure, dropping from an average share of total agricultural research expenditure of 13.8

Rice Poticy and Technology 669 per cent during 1978-80 to 9.4 per cent during 1988-90. From 1982 to 1988, rice research expenditure in Indonesia declined at a rate of 1.3 per cent per annum. Rice research output growth has declined in the 1980s, as measured by the number of rice research publications (Hayami and Morooka, 1987). The growth in number of rice research publications increased from 2.8 per cent for 1963-70 to 4.8 per cent for 1970-79, and then declined sharply to 1.6 per cent for 1979-85. Hayami and Morooka showed that the decline in rice research output over this last period was directly related to the decline in world rice prices. Declining Irrigation Investment

Since the mid-1960s, the growth rate of irrigated area in the world has declined by about 60 per cent; in Asia, it has declined by 72 per cent. Recent sharp reductions in irrigation investment are likely to further slow the rate of growth in area irrigated. Aggregate lending and assistance for irrigation in South and South-East Asia in the 1970s and 1980s by the four main financial donors for irrigation development—the World Bank, Asian Development Bank (ADB), the US Agency for International Development (USAID) and the Japanese Overseas Economic Cooperation Fund (OECF)—is shown in Table 2 and Figures 2 and 3. Lending and assistance for South and South-East Asia reached its peak in real terms during 1977-79, and by 1986-87 it was less than 50 per cent of the 1977-79 level. Total public expenditure for irrigation for individual countries in Asia also declined significantly during the 1980s (Table 3). In the Philippines, annual expenditure on irrigation investment in the late 1980s were only one-third the level of the early 1980s. Annual expenditure in Sri Lanka was cut nearly in half between the late 1970s and the late 1980s. Declines in the late 1980s from peak annual expenditure levels in India, Indonesia and Thailand range from 10 to 30 per cent. Among the factors contributing to the reductions in investment are the large public and foreign debt loads carried by most of the countries in the region, the declining share of unexploited irrigation development and concerns about the Table 2. Average annual lending and assistance for irrigation by the World Bank, Asian Development Bank, US Agency for International Development and the Japanese Overseas Economic Cooperation Fund to South and Southeast Asia (constant 1980 prices). Lending and Assistance to Irrigation ($ million) Period 1969-70 1971-73 1974-76 1977-79 1980-82 1983-85 1986-87

World Bank — 668 981 888 680 405

ADB

OECF

53 69 84 219 253 162 144

6 7 16 33 46 69 21

USAID

Total

— — 68 71 69 38

Sources: World Bank, ADB, OECF, USAID (unpublished data)

1301 1258 980 608

670 M. W. Rosegrant and P. L. Pingali Millions US $ 1,600

n n

1.400

1.200

n

1.000

800

600

400

200

I

74

75

76

II

77

I

78

AL 79

80

AL 81

82

83

84

85

86

87

Year

Figure 2. World Bank, ADB and OECF irrigation loans to South and South-East Asia (1980 prices).

Table 3.

Index of average annual public expenditures for new irrigation construction (1976-80 = 100).

Time period 1971-75 1976-80 1981-85 1986-90

India

Indonesia^

Philippines^

Sri Lanka

Thailand

60 100 94 80

20 100 192 170

25 100 125 45

37 100 92 55

99 100 151 109

"For Indonesia, and the Philippines, the successive time periods are 1969-73, 1974-78, 1979-83 and 1984-88 (1974-78 = 100). Sources: Computed from Gulati; Rosegrant and Pasandaran; Azarcon; Aluwihare and Kikuchi; and Rosegrant and Mongkolsmai.

Rice Policy and Technology

671

NAilllons IJS$ 1,600 1 •

AID

DOEC F DADB

1,600

Dworti Bank 1,400

1,200

1,000

n

n

n

800

600

400

200

79

60

81

82

83

84

85

86

87

Year

Figure 3. World Bank, ADB, OECE, and AID irrigation loans to South and South-East Asia (1980 prices).

environmental implications of irrigation projects (Rosegrant and Svendsen, 1992). However, econometric analyses show that the most important causes of declining investment are the decline in world rice prices and the increasing real costs per hectare of new irrigation development (Aluwihare and Kikuchi, 1990; Rosegrant and Mongkolsmai, 1990; Rosegrant and Pasandaran, 1990; Svendsen and Ramirez, 1990). Degradation of Irrigated Area

Declining investment in irrigation has been accompanied by a decline in the quality and performance of existing irrigation systems. Although data are limited and definitions of damaged area vary between sources, estimates of annual global losses of agricultural land because of waterlogging and salinization range from lower estimates of 160,000-300,000 hectares (Tolba, 1978; Barrow, 1991) to higher estimates of 1.5 million hectares (Kovda, 1983; Brundtland and Khalid, 1987), with

672 M. W. Rosegrant and P. L. Pingali most of the waterlogging and salinization in irrigated croplands of high production potential. Estimates of the total area already affected by waterlogging and salinity but still under production are only slightly more consistent, at 20-46 million hectares (Barrow, 1991; El-Ashry, 1991; Rhoades, 1987; Kayasseh and Schenck, 1989). As much as one-half of the world's waterlogged and salinized area is in Asia, more particularly in India, Pakistan and China. In India, 6-10 million hectares are estimated to be waterlogged and salt affected (Joshi and Agnihotri, 1984; Joshi, 1987; Joshi and Jha, 1991). In Pakistan, it is generally estimated that waterlogging and salinity affect 20-35 per cent of the 15 million hectares of irrigated land (Bowonder and Ravi, 1984; Middleton, 1988; Lai et al., 1989; Postel, 1990; Barrow, 1991). For China, Postel (1990), citing a paper from the Ministry of Water Resources and Electric Power (1987), reported that an estimated 7 million hectares, or 15 per cent of the total irrigated area, suffer from the combined effects of salinity and alkalinity. Although even the lower estimates indicate that degradation of irrigated land is a significant and growing problem, the exact degree of the problem is poorly understood. Crosson and Anderson (1992) point out that knowledge is surprisingly sparse on the causes of, and policy solutions to, land degradation, on the impact of degradation on productivity and income, and on the reversibility of degradation.

3 INTENSIFICATION-INDUCED DECLINE IN PRODUCTIVITY GROWTH Intensification of irrigated land use, that is the movement from one to two or three crops of rice per year and increased use of inputs per hectare, has had both positive and negative consequences. The positive impacts (especially employment and income effects) of intensification and the adoption of modern seed-fertilizer technology have been examined in great detail by the 'green revolution' literature (Otsuka et al., 1990). The post-green revolution phase of declining productivity and stagnant incomes has been analysed only recently (Barker and Chapman, 1988; Herdt, 1988; Pingali et al., 1990). The current farm-level situation for much of Asian rice production is characterized by a diminished gap between the yield frontier and farmer yields, stagnant or declining yields on experiment stations and increased input requirements for sustaining current yield gains. Is there a Yield Gap? At the time IR-8, the first widely adopted modern rice variety, was released, farmers in the neighbourhood of IRRI, growing traditional rice varieties, were getting yields of 2.0-2.5 metric tons per hectare (mt/ha), implying a yield gap of around 3.5-4.0 mt/ha between the farmer and the experiment station (IRRI, 1967). Comparisons in 1989 of average farm data from samples in Laguna and Central Luzon with the technological potential for improvement show that the current yield gap is 1.2 mt/ha (Moya and Pingali, 1989). About a third of the farmers in both samples have yields that match or exceed the highest yields obtained in the experiment stations. Farmers with yields below experiment station

Rice Policy and Technology 673 levels do not have significant differences in varieties or input levels relative to farmers whose yields match the experiment station; rather these farmers have poorer access to irrigation water and may be less efficient in the utilization of modern technology. These differences are more structural and cannot be rectified through research alone. Stagnant or Declining Technological Yield Frontier

At the micro-level, experiment station data indicate that rice yield potential may actually be dechning. When IR-8 was released in 1966 it yielded as much as 10 mt/ ha in the dry season and 6 mt/ha in the wet season on IRRFs experimental farm. Following IR-8, 33 rice varieties have been released in the Philippines, many with better insect and disease resistance, shorter crop duration and better eating quality than IR-8. However, none of these newer varieties have been able to match the yield potential of IR-8. Perhaps more disturbing is the observation that the highest yields obtained from the long-term fertility trials conducted at IRRI and three other experiment stations in the Philippines are exhibiting a long-term decline (Pingali et al., 1990). Table 4 presents estimates for the annual growth in yields at the four experiment stations for the years 1966-88. Significant wet season yield declines are observed for three of the four experiment stations, the exception being Visayas, where yields increased at an annual rate of 0.18 per cent. Significant dry season yield dechnes are observed for two of the four experiment stations, the exceptions being Maligaya and Visayas. In the former, yields increased at the rate of 0.15 per cent per year, while in the latter yield growth was not significant. Similar yield declines have been observed in other experiment stations in India, Thailand and Indonesia. Long-term fertilizer trials on rice conducted in four locations in India, from 1972 to 1982, found declining long-term yields in two locations and stagnant yields in two locations. For rice-wheat cropping system trials, significant long-term rice yield declines have been found in most sites in India during the period 1973-88 (Byerlee, 1990). Long-term yield declines have been observed in 11

Table 4.

Growth in yield potential in the Philippines (1966-88). Annual growth in yields

Locations IRRI Maligaya Rice Research and Training Center Visayas Rice Experiment Station Bicol Rice and Corn Experiment Station

Wet season

Dry season

-1.29

-1.28

-1.01

+0.15

+0.18

+0.18"

-0.62

-0.38

significantly different from 0.0%. Source: Growth rates were estimated using data from the long-term fertility trials conducted at the above experiment stations.

674

M. W. Rosegrant and P. L. Pingali

years of continuous rice cropping at Maros, Indonesia and 10 years of continuous rice cropping in Chiang Mai, Thailand (Pingali et ai, 1990). The long-term stagnation or decline in yield potential on experiment stations under intensive irrigated rice production can be attributed to degradation of the paddy environment due to intensification of production. Degradation of the paddy environment occurs for one or more of the following reasons: increased pest pressure and incidence of plant diseases such as sheath blight and stem rot; build-up of root nematode populations that reduce root function; depletion of soil nitrogen and micro-nutrients over time; and changes in soil chemistry and fioodwater biology brought about by intensive use of fertilizers and pesticides and the increased reliance on low-quality irrigation water (Elinn and De Datta, 1984; Pingali et ai, 1990; IRRI, 1992). While it is not currently possible to quantify the relative contribution of these factors on yields, research under way at IRRI is assessing their physical and economic impacts. Declining Efficiency of Fertilizer Use

After land and labour, chemical fertilizers account for the largest share of inputs for irrigated rice production. The relative importance of chemical fertilizers to nutrient supply has increased over time with the reduction in use of manure and other organic fertilizers as the real wage of labour has increased (Rosegrant and Roumasset, 1988). Eertilizer use in Asia has grown rapidly over the past three decades on both a total and per hectare basis (Table 5). However, owing to both price and intensification effects, growth in fertilizer use in Asia has slowed during the 1980s, declining on a per hectare basis from 12 per cent during 1972-81 to 7.4 per cent per year during 1981-88. With long-term high growth rates in fertilizer use and declining rates of growth in yield, fertilizer levels in relatively favourable areas of Asia are now quite high, and increasing amounts of fertilizer are being used to maintain current yield levels. In parts of Asia, including West Java, the Punjab and much of China, fertilizers are being used at or above economically optimum levels at border prices. The achievement of relatively high levels of fertilizer use on rice in Asia has shifted concern from simply increasing the levels of use to improving the efficiency of fertilizer use. Yield-based growth in rice production in Asia has rapidly increased the rate of nutrient removal from the soil, a rate that has not been matched by balanced growth in the supply of nutrients through chemical and organic fertilizers. The result of unbalanced application of fertilizers has been a decline in the efficiency of fertilizer use over time (Ahmed, 1985; Stone, 1986; Desai and Gandhi, 1989). Increasing Losses Due to Pests

The use of purchased inputs for plant protection was unimportant for rice prior to the mass introduction of modern varieties. Farmers had traditionally relied on host plant resistance, natural enemies, cultural methods and mechanical methods such as hand weeding. Relatively minor pests—leaffolder, caseworm, armyworm and cutworm—started to cause noticeable losses in farmers' fields as the area planted

Rice Policy and Technology

675

Table 5. Annual growth rates (%) of total and per hectare use of fertilizer in Asia (1964-88). Region

1964-72

1972-81

1981-88

1964-88

Asia Total use Per hectare use

14.84 13.71

12.68 11.94

7.41 7.39

11.57 10.91

South Asia^ Total use Per hectare use

17.77 15.99

12.41 11.04

7.98 8.06

11.10 10.30

South-East Asia^ Total use Per hectare use

17.77 16.80

8.83 7.27

7.16 6.50

10.62 9.69

India Total use Per hectare use

17.27 16.54

11.07 10.01

9.34 9.27

11.01 10.37

China Total use Per hectare use

12.76 10.39

14.53 15.30

6.35 7.03

12.19 11.87

^South Asia includes Pakistan, Sri Lanka, Nepal, and Bangladesh and excludes India. ''South-East Asia includes Burma, Indonesia, Thailand, Kampuchea, Laos, Malaysia, Philippines and Vietnam.

with modern varieties increased. The green leafhopper and brown planthopper (BPH) became major problems, the former as a vector of rice tungro virus and the latter as a direct result of insecticides killing its natural enemies (Teng, 1990). The emergence of BPH as a dominant insect pest problem in Asia is explained by the use of susceptible varieties of rice or varieites whose resistance has broken down, and the persistence of indiscriminate pesticide-use practices. A severe BPH outbreak was reported in Thailand during 1989 and 1990. Irrigated rice areas in the central plains were most seriously affected (Heong, 1990). The common rice variety in these areas is Supan Buri 60, a susceptible variety. The damage was not only caused by hopperburn: in many areas the ragged stunt virus contributed to yield loss. The present BPH build-up and spread of ragged stunt virus in the irrigated areas may well result in the spread of this insect and associated disease problem to the deep-water areas, as happened during 1981-84.

4 POLICIES FOR SUSTAINING PRODUCTIVITY GROWTH

The evidence thus points to increasing difficulty and higher cost in maintaining yield growth in rice and the need for innovative policies to maintain cost-effective productivity growth. The following sections of the paper will discuss policy options for growth in rice productivity in the area of research, fertilizer, crop protection, extension and education, irrigation and pricing policy.

676 M. W. Rosegrant and P. L. Pingali Research Analyses of rates of return to rice research in Asia consistently show very high rates of return. Azam, et al. (1991) summarizes estimates of marginal rates of return to rice research in Asia ranging from 75 to 155 per cent. These high marginal rates of return have been sustained in the 1980s, despite the emphasis of research on yield maintenance and stability research as compared with generation of yield breakthroughs. In the future, rates of return will be affected by the structure and organization of research systems, the allocation of resources across agroclimatic environments and specific breeding strategies to improve the yield potential, stability and quality of rice. Structure and organization of agricultural research With the slowdown in growth of total resources for research it is necessary to be more efficient in utilizing resources. Greater emphasis should be placed on improving the quality of scientists and research management and on providing adequate operating funds and technical support. Oram (1990) suggests that the division of responsibilities and working relationships between the international and national research centres need to be re-examined in the light of seeking increased efficiency, and that decentralization of research regionally and to the farm level, based on agroecological characterization, may be the most effective approach, because it provides better farmer input and feedback to upstream researchers and policymakers. Research priorities for rice Setting research priorities across crops or across environments for a given crop is a complex and dynamic process. Barker (1988) identifies the major approaches used: congruence or parity methods, which allocate resources in proportion to the value of the commodity produced; weighted criteria models, which rank priorities according to a range of criteria, such as equity, efficiency and productivity; and expected economic benefit models, which attempt to rank research priorities based on expected future economic benefits of the research. Barker et al. (1985) use the last method to assess research priorities for rice, and show that the net expected returns are highest for research on irrigated rice, with substantial expected benefits also for the more favourable rainfed areas. Expected yield increments and probabilities of success are highest for research on the irrigated rice environment. Expected benefits from research on upland and deep-water rice are quite small, because the areas devoted to these crops are relatively small, the probable gains in yield are small and the potential for crop intensification is limited. These conclusions are supported by the actual distribution of total rice area by environment and the proportion of area under modern varieties (Table 6). Nearly 60 per cent of rice production in South and South-East Asia is accounted for by irrigated lowlands, and 95 per cent of irrigated area is under modern varieties. Another 17 per cent of production comes from relatively favourable rain-fed lowlands and 13 per cent from relatively unfavourable lowlands. The major production impact of yield improvement research is therefore likely to come from the irrigated and favourable rain-fed environments. Only 12 per cent of production is accounted for by deep-water and upland rice, so even if research breakthroughs are

Rice Poticy and Technology 677 Table 6. Rice area and production by environment for South and South-East Asia (1985).

Irrigated lowlands Rain-fed lowlands Favourable Unfavourable Deep water Upland Total

Rice area (000 ha) (% of total rice area)

Rice production (000 tons) (% of total production)

31,181 (34%) 36,850 (40%) 16,580 (18%) 20,270 (22%) 12,416 (14%) 11,151 (12%)

146,551 (58%) 77,385 (30%) 44,750^ (17%) 32,635" (13%) 18,624 (7%) 12,266 (5%)

91,598

254,826

% of area under modern varieties Yield (t/ha) 95%

4.7

40%

2.1

n.a."

2.7^

n.a.

1.6"

0

1.5

0

1.1 2.8

48%

"Preliminary estimates, ''n.a. means not available. Sources: IRRI World Rice Statistics (1987) and personal database.

made in these environments the production impacts will be small. If yield breakthroughs are made in these areas, the main impact would be on farmer incomes rather than on aggregate production.

Shifting the yield frontier

What are the likely breakthroughs in yield potential on irrigated and rain-fed lowlands? Two approaches are being used to attempt to improve rice yield potential by 20-25 per cent within the next 5-10 years: changes in plant architecture and the exploitation of hybrid vigour. Donor varieties for large panicles, very sturdy stems and low tillering have been identified and these are being used for developing new plant types with higher yield potential. This is a top-priority breeding activity in the irrigated rice programme at IRRI, but progress is likely to be gradual. Although it is extremely difficult to assess both the probabilities of success in breeding and the duration of time to get a successful variety into farmers' fields, it would appear that yield gains in farmers' fields from this type of plant architecture breeding are likely to be 7-10 years away (IRRI, 1990). Hybrid rice in China, cultivated on 15 million hectares during 1989, gives 15-20 per cent higher yield relative to semi-dwarf varieties (Lin and Yuan, 1980: He et al., 1987; Yuan and Virmani, 1989). The development of hybrids adapted to tropical conditions is a priority activity at IRRI. Experimental evidence (Virmani et al., 1982; Virmani and Edwards, 1983; Virmani, 1987, 1990) indicates that a similar increase in rice varietal yield potential is possible by developing suitable rice hybrids outside China. Hybrid rices for tropical conditions may be available within the next 5-10 years (IRRI, 1990).

678 M. W. Rosegrant and P. L. Pingali Increasing yield stability Three approaches are being used for improving yield stability through durable resistance to diseases and insects, including alien gene transfer, use of novel genes and horizontal resistance. Useful alien genes for resistance to diseases and insects are being transferred from wild species into cultivated rice. Viable technology for these more durable resistance genes should be available within the next 3 years. Protocols have been developed to transform rice with novel genes. Novel genes such as the Bt gene for insect resistance and coat protein genes for tungro resistance are likely to become available within the next 5 years. When introduced into rice they should impart high levels of resistance. Genes for partial or horizontal resistance to blast have been incorporated into improved breeding materials, which should also contribute to more durable resistance (IRRI, 1990).

Improving rice quality In addition to improving yield potential and stability, breeding can improve the quality of rice, increasing the value to both the farmer and the consumer, thereby improving the incentives for rice production. Toquero (1991) used a hedonic price model to identify quality characteristics that command a price premium across several South-East Asian countries. Results highlight the wide variation in consumer demand for quality characteristics and the economic values attached to each attribute across countries, regions, rural and urban areas, income groups and seasons. Given this heterogeneity in preferences, quality enhancement has generally been the concern of individual national programmes. International breeding efforts have concentrated on providing breeding material with distinct quality traits that can be incorporated on demand. International efforts are also leading to the use of biotechnology tools, such as molecular market-based selections, for increasing the probability of success in selecting for difficult traits such as aroma. The physical and chemical quality characteristics considered relevant by consumers include colour, shape, size and starch content. Physical properties influence milling recovery, while the physicochemical properties of the starch determine its cooking quality. Characterizing rice grains by their amylose content and gelatinization temperature provides an index of cooking quality. Most rice varieties grown in tropical Asia have a high amylose content (greater than 25 per cent), low gelatinization temperature and cook dry andfluffy.However, the preferred rice varieties have an intermediate level of amylose (20-23 per cent) and intermediate gelatinization temperature; they cook moist and remain soft when cool. While the early generation modern rice varieties, such as IR-8, has high amylose content and low temperature, more recent varieties, such as IR-64, are intermediate in amylose content and gelatinization temperature (IRRI, 1991). Aroma is also an important quality characteristic of high-quality rices. Domestic demand for aromatic rices of the basmati and jasmine type has been increasing across Asia. On the international market, aromatic rices command premium prices. International research efforts are currently under way to introduce aroma in high-yielding cultivars (Khush and Juliano, 1991). This effort will result in breeding lines that could be used by national programmes if they desire to introduce aroma traits in locally bred varieties.

Rice Policy and Technology 679 Fertilizer Subsidized fertilizer prices have induced increased use of fertilizers without encouraging efficiency of use, and have tended to favour the use of nitrogen fertilizers over other nutrients. As will be discussed later, the reduction and eventual removal of fertilizer price subsidies can improve the efficiency of fertilizer use. Non-price policies are also important, including location-specific research on soil fertility constraints and agronomic practices, improvement in extension services, development of improved fertilizer supply and distribution systems and development of physical and institutional infrastructure (Desai, 1986, 1988). There are two main instruments for increasing the efficiency of fertilizer use in Asia, the first of which is improving the balance of fertilizer applications in order to deal with soil fertility constraints. The most widespread soil fertility problem in Asia appears to be phosphorus deficiency. For example, in China it is estimated that about two-thirds of agricultural land is deficient in phosphorus, while in India nearly one-half of the districts have been classified as low in available phosphorus (Stone, 1986; Tandon, 1987; Desai and Gandhi, 1989). There is also growing evidence in Asia of sulphur and micro-nutrient deficiency problems. The second major instrument for increasing the efficiency of fertilizer use is the improvement of the placement and timing of fertilizer. Basal nitrogen application and a second nitrogen application just before panicle initiation have a higher yield response for transplanted rice than later applications. Efficient basal application of nitrogen requires thorough incorporation without standing fioodwater during final harrowing and land-levelling before transplanting. Further efficiency gains can be made by deep placement of fertilizer rather than surface application. Further work needs to be done, however, to generate economically viable deep-placement technologies.

Crop Protection Efficiency gains in the utilization of pesticides appear to be possible through the use of integrated pest management (IPM). Adoption of IPM can increase the efficiency of pesticide use in two ways: by lowering the amount of pesticides applied without a consequent yield decline and by promoting non-chemical methods of pest control. IPM uses resistant varieties, predator management and cultural practices along with judicious pesticide use to provide long-term control of pest losses. A major force that has shaped the evolution of IPM field implementation on rice is the FAO Inter-Country Program for Integrated Pest Management for South and South-East Asia. The programme has encouraged both technical accomplishments and significant policy changes in several governments in the form of official sanctions of IPM as the means of national plant protection. Official decrees supporting IPM implementation have been promulgated in the Philippines, Indonesia, India, Sri Lanka and Malaysia (Teng, 1990). IPM technology is still single-pest oriented, and further work needs to be done to produce the knowledge for farmers to manage pests in an integrated manner. Inappropriate use of IPM can actually increase the probability of pest damage and increase the use of pesticides. Because of this, improved extension and training is crucial with this technology, as with techniques for increased fertilizer efficiency.

680 M. W. Rosegrant and P. L. Pingali Extension and Education Empirical evidence on the economic returns to extension is rare, but Rosegrant and Evenson (1992) find a marginal rate of return to extension in India of 52 per cent. The impact of education on long-term productivity growth has also been difficult to measure. However, Rosegrant and Evenson (1993) show that improvement in rural literacy (which is heavily infiuenced by education, as well as investment in health and infrastructure) has been one of the main sources of agricultural productivity growth in both India and Pakistan. Returns to extension and education in the future may be even higher because of the increasing importance of efficient use of inputs as opposed to input and crop variety promotion. Technologies to implement IPM and to improve the nutrient balance and the timing and placement of fertilizer applications are highly complex, knowledge intensive and location specific. Because new technologies are more demanding for both the farmer and the extension agent, they require more information and skills for successful adoption than the initial adoption of modern varieties and fertilizers. In addition, these improved technologies do not give as clear gains in yield and income as the initial adoption of new technology. The increase in income from these technologies is in fact highly sensitive to the farmer's skills and efficiency in using them (Byerlee, 1987). As a result of the greater complexity of new rice technologies, increased investment in extension, education and human capital more generally is likely to have high payoffs. To provide the necessary information dissemination and training for these new technologies, extension services will have to be upgraded. The poor performance of many extension services can be attributed to inadequate training, inappropriate organization and lack of incentives. Possible options for reform are privatization of extension services, the training and visit extension system and decentralization of existing public systems. Privatization of extension for rice farming through contracting to private companies can introduce incentives for higher efficiency, but may be difficult to manage and monitor and may not generate adequate profits for the private company because of the large number of widely dispersed small farmers growing many different cropping systems. Privatization of extension is more likely to be successful when extension is linked to delivery of a specific technology, such as hybrid corn, to larger, more homogeneous, farmers. The training and visit (T&V) extension programme has shown mixed success in Asia. Feder et ai (1987) found significant productivity gains from T&V in India, while other researchers have found no impact on productivity (Khan et ai, 1984). Byerlee (1990) suggests that the problems with T&V may be because of excessive promotion of inputs rather than focusing on efficiency enhancement. Decentralization of existing extension services could also help farmers cope with the additional complexity of efficiency-enhancing technology. Thimm (1990) recommends integrated national planning of research and extension to ensure the appropriate budgetary mix for proper operation, regular interdisciplinary evaluation of onfarm benefits of recommended technology and establishment of a goal-oriented organizational structure which encourages a bottom-up fiow from fanners to extension and research. The last-mentioned, when combined with adaptive, location-specific research, is particularly important in the transfer of complex technologies.

Rice Policy and Technology 681 Irrigation Given the significant investment decline and degradation of existing irrigated area, what will be the contribution of irrigation to future rice production growth? Rosegrant and Svendsen (1993) have shown that the slowing of investment in new irrigation is a rational response to declining rice prices and rising capital costs, and that continued increases in capital costs of new irrigation construction make a return to the investment levels of the 1970s highly unlikely. With expansion of irrigation into new areas likely to be slow, future increases in food production must come mainly from improvement in the productivity of the existing irrigated land base. This implies both the need to increase the efficiency of water use and the need to improve the quality of the resource base in irrigated areas, reversing the trends towards increased degradation through waterlogging, salinization of soil, degradation of water quality and groundwater mining. However, attempts to improve the performance of existing systems, through rehabilitation of system facilities on the one hand, and expansion of farmer participation through water user associations or other cooperative groups on the other, have often produced results that fall far short of expectations. It has been difficult to establish whether increasing farmer involvement through water user associations has met the objective of improving the quality of irrigation service and increasing system agricultural outputs (Svendsen, 1990). The record of rehabilitation projects, many of which are large-scale and capital intensive, shows highly variable, and often low, net benefits (Rosegrant and Svendsen, 1992). Successful rehabilitation requires greater focus on choice of appropriate systems for intervention, selection of the appropriate intervention points within these systems and identification of low-cost rehabilitation options to implement at these intervention points. Particular attention should be given to the development of cost-effective projects designed to reduce waterlogging and salinity problems, including canal lining and drainage, field drainage and conjunctive use of surface and groundwater. Perhaps more importantly, policies for improving the incentives facing user groups and individual irrigators through the establishment of efficiency prices which reflect the scarcity value of water should be examined. Virtually free delivery of water has been a major source of inefficiency and wastage of irrigation water (Sampath, 1992). Establishment of more appropriate incentives through administered efficiency prices have been most often recommended, but establishment of markets in tradable water rights could have several major advantages over administered prices and other allocation mechanisms. Tradable water rights formalize and secure the existing water rights held by farmers, economize on transactions costs, induce irrigators to consider the full opportunity cost of water including its value in non-rice crops and other uses and provide incentives for irrigators to internalize many of the externalities inherent in irrigation, possibly reducing degradation of the irrigated environment (Rosegrant and Binswanger, 1993). Improvement in the structure of property rights for water could also hasten the expansion of private investment in irrigation. The highly successful deregulation and privatization of the tubewell sector in Bangladesh in the mid-1980s suggests that the major role of the government is as a facilitator, through provision of public goods, and as a regulator, through enforcement of legal rights, rather than as an active participant. Governments should seek to reduce subsidies on irrigation

682 M. W. Rosegrant and P. L. Pingali equipment, credit and energy which tend to encourage high-cost and inefficient investments and which create continuing pressure on ordinary budget resources of the government. Pricing policy Output pricing policy As we noted above, the profitability of rice production declined during the 1980s, in large part because of the falling world price of rice. Long-term differences between domestic and world prices are costly to the economy. If the domestic price of a commodity is lower than the world price, farmers are penalized and incentives to production reduced. Raising the price to the world level will induce increased production and reduce imports or increase exports of the commodity. If the domestic price is higher than the world price, consumers are penalized, and excess productive resources are devoted to production of the commodity. The country can therefore gain by allowing imports, which will reduce the domestic price, improve consumer welfare and release productive resources to other crops with a comparative advantage in production. David (1990) shows that countries in Asia have mostly followed long-term world price in setting domestic rice prices. The constraints on pricing policy imposed by the economic costs of significant departures from world prices limit the fiexibility of the governments in using rice price policy to infiuence long-term production growth rates. Although the scope for infiuencing long-term productivity growth through manipulation of the rice price is limited, the domestic rice price is also infiuenced indirectly through government policies that affect the real exchange rate. In most Asian countries, the indirect effects of trade and macroeconomic policies have caused overvaluation of the real exchange rate, which in turn lowers the effective rice price and other agricultural prices (Bautista, 1990). In the Philippines, for example, overvaluation of the exchange rate arising from protection of domestic industry has lowered rice and other agricultural prices by 30 per cent over the past several years (David, 1989). Policies in Thailand, Sri Lanka and Pakistan similarly induced overvaluation of the real exchange rate by 15-25 per cent during the 1980s (Bautista, 1990). Reform of trade and macroeconomic policy regimes which have penalized agriculture would provide a significant stimulus to the rice sectors of several Asian countries. Such reforms would not only improve short-term input use efficiency and output, but even more importantly would encourage long-term investment and technological change in the agricultural sector. Fertilizer price policy Governments have intervened in the setting of fertilizer prices in support of a number of broad-based policy goals. Many countries have protected domestic fertilizer production by restricting imports or maintaining import tariffs. In the Philippines, for example, domestic fertilizer prices were maintained well above world prices until the mid-1980s by a combination of import controls and subsidies to domestic fertilizer plants. More commonly, governments have subsidized farmlevel fertilizer prices in support of several objectives, including income support for farmers and provision of incentives to increase the rate of adoption and level of

Rice Policy and Technology 683 fertilizer use, to increase crop production and to balance other taxes against agriculture. Fertilizer subsidies can become extremely costly to government treasuries. In Indonesia, subsidies have been maintained both for farmers and for the domestic fertilizer industry. The total costs for the two types of fertilizer subsidy were about Rupiah 670 billion (US $407 million) in 1986-87, representing nearly one-sixth of total government development expenditures for agriculture and irrigation. Since 1986, Indonesia has been slowly phasing out fertilizer subsidies, but they remain substantial. Fertilizer subsidies in Bangladesh in 1983-84 accounted for about 14 per cent of its budget allocation to agriculture (Rosegrant and Ahmed, 1990). The economic costs can be even greater, as subsidies soak up funds that could be used for alternative investments, and can induce the overuse of fertilizers relative to socially optimal levels. To the extent that subsidies are not fully funded to provide enough fertilizer to meet demand at the subsidized price, excess demand will be created, which can contribute to non-price rationing, non-availability of fertilizer, black markets, poor logistics and untimely delivery of fertilizer. Given the negative effects of subsidies, are there appropriate uses for fertilizer subsidies to farmers? Fertilizer subsidies to farmers may be cost-effective in stimulating farmers to adopt and utilize fertilizer appropriately together with new production technology. Temporary subsidies during the early stage of adoption of fertilizer may be effective in overcoming the fixed costs related to the adoption of new technology and in inducing farmer experimentation and learning during periods of rapidly changing technological potential. However, such temporary subsidies should be phased out as adoption and appropriate use of fertilizer become widespread as in many Asian countries. At high levels of fertilizer use, the budgetary cost of the fertilizer subsidy becomes prohibitively expensive, and the subsidy induces inefficient use of fertilizer beyond appropriate levels. Removal of fertilizer subsidies, in conjunction with appropriate extension policies, would encourage more efficient use of fertilizer. 5

CONCLUSIONS

This paper has shown that a combination of price-induced and intensificationinduced developments in the 1980s caused a sudden decline in the growth in rice yields to a rate below projected demand growth in the next two decades. These developments include reduced growth in research expenditures and output, reduced irrigation investment and degradation of irrigated land, a declining yield gap and stagnant technological yield frontier, declining growth in fertilizer use and efficiency and increased crop losses due to pests. This combination of factors will increase the difficulty in maintaining yield growth in rice, and innovative policies will be required to sustain cost-effective growth. Future growth in rice productivity will increasingly come from improved management and efficiency of use of the scarce resources consumed in rice production, in contrast to rapid dissemination of modern technology, which has been dominant in the past. Thus, there is a shift in fertilizer policy from a sole focus on increasing the level of use of fertilizer to improving the efficiency of nutrient balance and the timing and placement of fertilizers; there is a shift in crop protection policy from

684

M. W. Rosegrant and P. L. Pingali

dissemination of chemical pesticides to utilization of integrated pest management techniques; and a shift of emphasis in irrigation policy from investment in new systems to improved water-use efficiency and productivity in existing systems. The foundations of increased efficiency in rice production are greater investment in research, extension and education to upgrade human capital, combined with the establishment of economic incentives that reflect the opportunity costs of scarce resources. Increased funding of research and extension should be combined with decentralization based on agroecological characterization and increased input from farmers into the technology generation and dissemination process to generate technology better adapted to diverse production environments. Appropriate economic incentives should be established through rice price policies which keep domestic rice prices in line with long-term world rice price trends; through establishment of efficiency pricing of irrigation water by administered pricing or markets in tradable water rights; and by elimination of subsidies to fertilizer and pesticides. Finally, reform of trade and macroeconomic policy regimes which have penalized agriculture should stimulate rice production by improving short-term input-use efficiency and output and encourage long-term investment and technological change in the agricultural sector.

REFERENCES Ahmed, N. (1985). 'Fertilizer efficiency and crop yields in Pakistan', Phosphorus in Agriculture, 89, 17-32. Aluwihare, P. B. and Kikuchi, M. (1990). 'Irrigation investment trend in Sri Lanka: new construction and beyond', Sri Lanka: International Irrigation Management Institute. Azam, Q. T., Bloom, F. A. and Fvenson, R. F. (1991). 'Agricultural research productivity in Pakistan', Discussion Paper No. 644. Yale University: Fconomic Growth Center. Azarcon, Y. J. (1990). 'Public investments in irrigation in the Philippines'. Unpublished M.S. Thesis. Ithaca: Cornell University. Barker, R. (1988). 'Methods for setting agricultural research priorities'. Working Paper in Agricultural Fconomics No. 88-3. Cornell University: Department of Agricultural Fconomics. Barker, R. and Chapman, D. (1988). 'The economics of sustainable agricultural systems in developing countries' (mimeo). Cornell University. Barker, R., Herdt, R. W. and Rose, B. (1985). The Rice Economy of Asia. Washington, DC: Resources for the Future. Barrow, C. J. (1991). Land Degradation: Development and Breakdown of Terrestrial Environment. Cambridge: Cambridge University Press. Bautista, R. M. (1990). 'Development strategies, foreign trade regimes, and agricultural incentives in Asia', Journal of Asian Economics, 1, 115-134. Bouis, H. F. (1989). 'Prospects of rice supply/demand balances in Asia', final report submitted to the Rockefeller Foundation. Washington, DC: International Food Policy Research Institute. Bouis, H. F. (1990a). 'Measuring the sources of growth in rice yields: are growth rates declining in South and Southeast Asia?' (mimeo). Washington, DC: International Food Policy Research Institute.

Rice Policy and Technology 685 Bouis, H. E. (1990b). 'Rice in Asia: is it becoming a commercial good?, (mimeo). Washington, DC: International Food Policy Research Institute. Bowonder, B. and Ravi, C. (1984). 'Waterlogging from irrigation projects—an environmental management problem'. Hyderabad: Center for Energy and Technology, Administrative Staff College of India. Brundtland, G. H. and Khalid, M. (1987). Our Common Future. Report of the World Commission on Environment and Development, presented to the chairman of Intergovernmental Intersessional Preparatory Committee, UNEP Governing Council. Oxford: Oxford University Press. Byerlee, D. (1987). 'Maintaining the momentum in post-green revolution agriculture: a micro-level perspective from Asia', International Development Paper No. 10. Michigan State University. Byerlee, D. (1992). 'Technological change, productivity and sustainability in irrigated cropping systems of South Asia: emerging issues in the post-green revolution era'. Journal of International Development, 4, 447-496. Crosson, P. and Anderson, J. R. (1992). 'Global food-resources and prospects for the major cereals'. Background Paper No. 19, World Development Report. Washington, DC: World Bank. Dalrymple, D. G. (1986). 'Development and spread of high-yielding rice varieties in developing countries', Washington, DC: Metrotec. David, C. C. (1989). 'Philippines: price policy in transition'. In Sicular, T. (ed.), Food Price Policy in Asia. Ithaca, NY: Cornell University Press. David, C. C. (1990). 'Determinants of rice price protection in Asia', Social Science Division Paper No. 90-22. Los Banos, Philippines: International Rice Research Institute. Desai, G. (1988). 'Development of fertilizer markets in South Asia: a comparative overview' (mineo). Washington, DC: International Food Policy Research Institute. Desai, G. M. (1986). 'Fertilizer use in India: the next stage in policy', Indian Journal of Agricultural Economics, 41, 248-270. Desai, G. M. and Gandhi, V. (1989). 'Phosphorus for sustainable agricultural growth in Asia: an assessment of alternative sources and management', paper presented at the symposium on phosphorus requirements for sustainable agriculture in the Asia and Pacific region, March 6-10. Los Bafios, Philippines: International Rice Research Institute. El-Ashry, M. T. (1991). 'Policies for water resource management in semiarid regions'. International Journal of Water Resources Development, 7, 230-234. Feder, G., Lau, L. J. and Slade, R. H. (1987). 'Does agricultural extension pay? The training and visit system in Northwest India', American Journal of Agricultural Economics, 69, 677-686. Flinn, J. C. and De Datta, S. K. (1984). 'Trends in irrigated-rice yields under intensive cropping at Philippine research stations'. Field Crops Research, 9, 1-15. Gulati, A. (1991). 'Development cost of irrigation in India and subsidies therein' (mimeo). Washington, DC: IFPRI. Hayami, Y. (1975). 'A century of agricultural growth in Japan', Tokyo: University of Tokyo Press; Minneapolis, MN: University of Minnesota Press. Hayami, Y, and Morooka, K. (1987). 'The market price response of world rice research'. Agricultural Economics Department Paper No. 87-21. Los Bafios, Philippines: International Rice Research Institute. He, G. T., Zhu, X. and Flinn, J. C. (1987). 'A comparative study of economic efficiency of hybrid and conventional rice production in Jiangsu Province, in China', Oryza, 24,285-296.

686

M. W. Rosegrant and P. L. Pingali

Heong, K. L. (1990). 'Brown planthopper problem in Thailand', Trip Report August 2 2 25. Los Banos, Philippines: International Rice Research Institute. Herdt, R. W. and Capule, C. (1983). Adoption, Spread, and Production Impact of Modern Rice Varieties in Asia, Los Banos, Philippines: International Rice Research Institute. IFPRI (1992). 'Food supply, demand, and trade in Asia: regional trends and projections' (mimeo). Washington, DC: IFPRI. IRRI (1967). Annual Report-1967. Los Banos, Philippines: IRRI. IRRI (1967). World Rice Statistics. Los Banos, Philippines: IRRI. IRRI (1989). IRRI Toward 2000 and Beyond. Los Banos, Philippines: IRRI. IRRI (1990). 'Irrigated rice program: 1991 budget presentation'. Los Banos, Philippines: IRRI. IRRI (1991). 'Rice grain marketing and quality issues', Los Banos, Philippines: IRRI. IRRI (1992). 'Yield stagnation, yield decline, and the yield frontier of irrigated rice', report presented to the Program Committee of the Board of Trustees of the IRRI. Los Banos, Philippines: IRRI. Joshi, P. K. (1987). 'Effect of surface irrigation on land degradation—problems and strategies', Indian Journal of Agricultural Economics, 42, 416-423. Joshi, P. K. and Agnihotri, A. K. (1984). 'An assessment of the adverse effects of canal irrigation in India', Indian Journal of Agricultural Economics., 39, 528-536. Joshi, P. K. and Jha, D. (1991). 'Farm-level effects of soil degradation in Sharda Sarayak irrigation project'. Central Soil Salinity Research Institute, Indian Council of Agricultural Research and International Food Policy Research Institute. Working Papers on Future Growth in Indian Agriculture, No. 1. Washington, DC: IFPRI. Kayasseh, M. and Schenck, C. (1989). 'Reclamation of saline soils using calcium sulphate from the titanium industry', Ambio, 18, 124-127. Khan, M. J., Sharif, M. and Sarwar, M. (1984). 'Monitoring and evaluation of training and visit system of agricultural extension in Punjab, Pakistan', Publication No. 212. Lahore: Punjab Economics Research Institute. Khush, G. S. and Juliano, B. O. (1991). 'Research priorities for improving rice grain quality'. In Rice Grain Marketing and Quality Issues. Los Banos, Philippines: IRRI. Kovda, V. A. (1983). 'Loss of productive land due to salinization', Ambio, 12, 91-93. Kumar, P. and Rosegrant, M. W. (1993). 'Productivity and source of growth in rice: India' (mimeo). Washington, DC: International Food Policy Research Institute. Lai, R., Hall, G. F. and Miller, F. P. (1989). 'Soil degradation: I. Basic processes'. Land Degradation and Rehabilitation, 1, 51-69. Lin, S. C. and Yuan, L. P. (1980). 'Hybrid rice breeding in China'. In Innovative Approaches to Rice Breeding. Manila, Philippines: International Rice Research Institute, pp. 35-51. Middleton, N. (1988). Atlas of Environmental Issues. Oxford: Oxford University Press. Ministry of Food and Agriculture (Pakistan). (1988). Report of the National Commission on Agriculture. Islamabad. Ministry of Water Resources and Electric Power (China). (1987). Irrigation and Drainage in China. Beijing. Moya, F. P. and Pingali, P. L. (1989). 'Can we close the yield gap between the "best" and "ordinary" farmers in Luzon?', Social Science Division Paper No. 89-03. Los Bafios, Philippines: IRRI. Oram, P. (1990). 'Agricultural productivity growth and the structure and organization of agricultural research', policy brief presented at the International Food Policy Research Institute policy seminar on technology for sustainable agricultural growth. The Hague, July 2 - 3 .

Rice Policy and Technology 687 Otsuka, K., Cordova, V. G. and David C. C. (1990). 'Green revolution, land reform and household income distribution in the Philippines', Social Science Division Paper No. 9 0 08. Los Banos, Philippines: IRRI. Pardey, P., Roseboom, J. and Anderson, J. (1991). Agricultural Research Policy: International Quantitative Perspective. Cambridge: Cambridge University Press. Pardey, P. G., Lindner, R. K., Abdurachman, E., Wood, S., Fan, S., Fveleens, W. M., Zhang, B. and Alson, J. M. (1992). The Economic Returns to Indonesian Rice and Soybean Research (mimeo). The Hague: ISNAR. Pingali, P. L., Moya, P. F. and Velasco, L. F. (1990). 'The post-green revolution blues in Asian rice production—the diminished gap between experiment station and farmer yields'. Social Science Division paper No. 90-01. Los Banos, Philippines: IRRI. Postel, S. (1990). 'Saving water for agriculture'. State of the World, Washington, DC: WorldWatch Institute. Rhoades, J. D. (1982, 1987). 'The problem of salt in agriculture'. In Yearbook of Science and the Future. Chicago: Encyclopaedia Britannica. Rosegrant, M. W. and R. E. Evenson (1992). 'Agricultural productivity and sources of growth in south Asia', American Journal of Agricultural Economics, 74, 757-761. Rosegrant, M. W. and Evenson, R. E. (1993). 'Agricultural productivity growth in Pakistan and India: a comparative analysis', Pakistan Development Review, 32. Rosegrant, M. W. and Svendsen, M. (1993). 'Irrigation investment and management policy for Asian food production growth in the 1990s', Food Policy, 18, 13-32. Rosegrant, M. W. and Binswanger, H. P. (1993). 'Markets in tradable water rights: potential for efficiency gains in developing-country irrigation' (mimeo). Washington, DC: IFPRI. Rosegrant, M. W. and Ahmed, R. (1990). 'Fertilizer and output pricing policies for sustainable agriculture'. Policy brief presented at the International Food Policy Research Institute policy seminar on technology for sustainable agricultural growth. The Hague, July 2-3. Rosegrant, M. W. and Roumasset, J. A. (1988). 'Fconomic feasibility of green manure in rice-based cropping system'. In Green Manure in Rice Farming. Los Banos: IRRI. Rosegrant, M. W. and Mongkolsmai, D. (1990). 'Trends and determinants of irrigation investment in Thailand' (mimeo). Washington, DC: IFPRI. Rosegrant, M. W. and Pasandaran, E. (1990). 'Irrigation in Indonesia: trends and determinants' (mimeo). Washington, DC: IFPRI. Rosegrant, M. W. and Svendsen, M. (1992). 'Irrigation investment and management in Asia: trends, priorities, and policy directions' (mimeo). Washington, DC: IFPRI. Sampath, R. K. (1992). 'Issues in irrigation pricing in developing countries'. World Development, 20(7), 967-77. Sarma, J. S., and Gandhi, V. P. (1990). Production and Consumption of Foodgrains in India: Implications of Accelerated Economic Growth and Poverty Alleviation, Research Report No. 81. Washington, DC: IFPRI. Stone, B. (1986). 'Chinese fertilizer application in the 1980s and 1990s: issues of growth, balances, allocation, efficiency and resonse'. In China's Economy Looks Towards the Year 2000, Vol. l . , p p . 453-493. Svendsen, M. (1990). 'Substainability in irrigated agriculture', Entwicklung landicher

Raum, lA, 12-15. Svendsen, M. and Ramirez, R. (1990). 'Determinants of irrigation investment in the Philippines' (mimeo). Washington, DC: IFPRI.

688

M. W. Rosegrant and P. L. Pingali

Tandon, H. L. S. (1987). 'Phosphorus research and agricultural production in India'. New Delhi: Fertilizer Development and Consumption Organization. Teng, P. S. (1990). 'IPM in rice: an analysis of the status quo with recommendations for action'. Report to the International IPM Task Force (FAO/ACIAR/IDRC/USAID/ NRI). Los Banos, Philippines: IRRI. Thimm, H. V. (1990). 'Agricultural extension and the diffusion of new technologies'. Policy brief presented at the International Food Policy Research Institute policy seminar on technology for sustainable agricultural growth. The Hague, July 2 - 3 . Tolba, M. K. (1978). 'Welcoming address'. Proceedings of the Water Management for Arid Lands in Developing Countries Workshop., Vol. 13. Oxford: Pergamon Press. Toquero, Z. F. (1991). 'Consumer demand for rice grain quality'. In Rice Grain Marketing and Quality Issues. Los Banos, Philippines: IRRI. Virmani, S. S. (1990). 'Hybrid rice—prospects and limitations'. Paper presented at International Conference on Seed Science and Technology, New Delhi, India, Feb. 21-25. Virmani, S. S. and Edwards, I. B. (1983). 'Current status and future prospects for breeding hybrid rice and wheat'. Adv. Agron., 36, 145-214. Virmani, S. S., Aquino, R. C. and Khush, G. S. (1982). 'Heterosis breeding in rice, Oryza sativa\ L. Theor. Appl. Genetic, 63, 373-380. Yuan, L. P., Virmani, S. S. and Mao, C. X. (1989). 'Hybrid rice—achievement and future outlook'. In International Rice Research Institute, Progress in Irrigated Rice Research. Manila, Philippines: IRRI, pp. 7-24.