EPTD DISCUSSION PAPER NO. 94 Environment and Production Technology Division International Food Policy Research Institute 2033 K Street, N.W. Washington, D.C. 20006 U.S.A. September 2002 EPTD Discussion Papers contain preliminary material and research results, and are circulated prior to a full peer review in order to stimulate discussion and critical comment. It is expected that most Discussion Papers will eventually be published in some other form, and that their content may also be revised. AGRICULTURAL RESEARCH AND URBAN POVERTY IN INDIA Shenggen Fan i ABSTRACT Using a similar analytical approach to a study in China, this paper analyzes the impact of agricultural research on urban poverty reduction in India. State level data from 1970 to 1995 were used in the empirical analysis. It is found that in addition to its large impact on rural poverty reduction, agricultural research investments have also played a major role in the reduction of urban poverty. Agricultural research investments increase agricultural production, and increased production in turn lowers food prices. The urban poor often benefit proportionately more than the non-poor since they spend 50-80% of their income on food. Among all the rural investments considered in this study, agricultural research has the largest impact on urban poverty reduction per additional unit of investment. The results from this study are similar to earlier findings for China. Today, urban poverty still accounts for one quarter of total poverty in India, and this share is expected to rise in the future. Policymakers cannot afford to be complacent about this trend and continued investments are still needed to keep food prices low. Among all government policy instruments, increased agricultural research is still the most effective way to achieve this objective. KEYWORDS: developing countries, India, agricultural research, urban, poverty, food price . ii TABLE OF CONTENTS 1. Introduction......................................................................................................................1 2. Agricultural Research, Technology, Productivity..............................................................3 3. Urban Poverty ..................................................................................................................8 4. Econometric Model .........................................................................................................10 5. Data and Model Estimation ............................................................................................11 6. Contribution of Agricultural Research to Urban Poverty Reduction .................................17 7. Conclusions.....................................................................................................................19 References...........................................................................................................................21 AGRICULTURAL RESEARCH AND URBAN POVERTY IN INDIA Shenggen Fan1 1. INTRODUCTION The debate on the role of agricultural research in poverty alleviation dates back to the green revolution in South Asia and Mexico in the late 1960s (Pinstrup-Andersen and Hazell, 1985). A general consensus has emerged that not only did research-led technology prevent widespread starvation; it also contributed to significant national economic growth and saved huge areas of forest, hillsides and other environmentally fragile lands from conversion to agriculture. For example, the green revolution contributed to more than a doubling of the aggregate food supply in Asia over a 25-year period. More importantly, it achieved this output increase with only a 4 percent increase in the net cropped area (Rosegrant and Hazell, 2000). There is also a large empirical literature on the economic returns to agricultural research investment in developing countries. Alston et al. (2001) reviewed 292 studies (more than 1886 rates of return estimates) and obtained an average rate of return of 100 percent to agricultural research investment with a median rate of return of 48 percent. Fan, Hazell and Thorat (2000) were the first to directly link agricultural research to rural poverty reduction. Their results for rural India indicate that agricultural research has the largest productivity impact of all kinds of government investments included in their study. This growth impact has also trickled down to the rural poor. In fact, 1 Shenggen Fan is Senior Research Fellow in the Environment and Production Technology Division at IFPRI. 2 agricultural research has the second largest impact on rural poverty reduction in India, second only to investments in rural roads. Using provincial-level data for China, Fan, Zhang, and Zhang (2002) reached a similar conclusion that agricultural research has the largest productivity effect on agricultural production, and also has the second largest poverty impact in rural China. Only investments in rural education have a larger poverty impact. The links between agricultural research and food price benefits for consumers have also been quantified, using the consumer surplus as a welfare measure (Akino and Hayami 1975; Mellor 1975; Scobie and Posada 1978; and Pinstrup-Andersen 1979). But little work has been done on quantifying the impact of agricultural research on urban poverty reduction, despite the fact that rapid urbanization is increasing the incidence of urban poverty in developing countries (Haddad et al. 1999; Ravillion 2000). Fan, Fang, and Zhang (2002) were the first to develop a model formally linking agricultural research with urban poverty reduction and applied it to China. This paper uses India as a case to reinforce the findings of the China case study. But the India case has its own merits. First, India is largely a market-driven economy, in contrast to the centrally-planned economy practiced by China until the late 1970s. The distorted nature of food prices in China makes it difficult to fully capture the impact of agricultural research on urban poverty reduction by lowering urban food prices. Second, despite considerable success reducing poverty, India today still has more than 70 million urban poor, accounting for one third of India�s total poor. India also accounts for a large share of the total global 3 urban poor (more than 40%). Developing a national strategy to prevent further increases in urban poverty is more urgent than ever. The paper is organized as follows. We first review historical trends in agricultural research investment, technology development, and productivity growth in Indian agriculture, followed by a brief discussion of changes in urban poverty. Second, a conceptual framework and model are developed and adapted for the analysis of how agricultural research affects the urban poor, and the estimation procedures and results are discussed. We then conclude with some policy implications. 2. AGRICULTURAL RESEARCH, TECHNOLOGY, PRODUCTIVITY Government spending on agricultural research in India has increased significantly over the past four decades, but not without substantial year-to-year variations (Table 1). Investment in agricultural research was quite modest during the 1960s, ranging from 1.6 to 1.9 billion Rupees (all values in 1995 prices). During the 1970s, expenditures on agricultural research increased dramatically to 4.0 billion Rs around 1980, more than doubling in the decade. This was the period when many agricultural universities and national research institutions were set up (Evenson, Rosegrant, and Pray, 1999). These were the driving force behind the green revolution that more than doubled the yields of rice and wheat within a decade. During the 1980s, research expenditures continued to increase to 7 billion Rs in 1990. But in the 1990s, research expenditure increased only modestly to 7.3 billion Rs by 1995, which is worrying given their importance to national food security and poverty alleviation. 4 As a percentage of agricultural gross domestic product (AgGDP), agricultural research investment was relatively low at 0.20% during the 1960s, but it increased dramatically to more than 0.40% in the 1970s. In the 1980s, the percentage continued to rise, to a peak of 0.50% in 1987. But the percentage has gradually declined to below 0.43% in recent years. This indicates that government investment in agricultural research has increased in absolute terms over the past decade, but has declined relative to the size of the agricultural sector. One of the most significant changes in Indian agriculture in recent decades has been the widespread adoption of high-yielding varieties. During the green revolution of the 1970s, the crop area planted to high-yielding varieties (HYVs) for five major crops (rice, wheat, maize, sorghum, and pearl millet) increased from about 20 percent to 40 percent (Table 2).3 Even after the green revolution, the percentage of the crop area planted with HYVs continued to increase. It reached 53 percent by 1990, and 59 percent by 1995. This has been one of the major engines of productivity growth in Indian agriculture. As a result of the rapid adoption of new technologies and improved rural infrastructure, agricultural production and factor productivity have both grown rapidly in India. For all India, agricultural production grew at 2.11 percent per annum between 1970 and 1995 (Table 2). In the 1970s, production growth was comparatively low, growing at an average annual rate of only 1.95 percent. In the 1980s, it grew at 3.82 3 High-yielding varieties (also referred to as modern varieties) are those released by the Indian national agricultural research system and the international agricultural research centers. The yields of these varieties are usually substantially higher than those of traditional varieties. The percentage of cropped areas with HYVs is calculated as the ratio of areas planted with HYVs for 5 major crops (rice, wheat, maize, sorghum, and pearl millet) to the total cropped areas of these five crops. 5 percent per annum, a much higher growth rate than most other countries achieved during the same period. Since 1990, production growth has slowed, growing at only 2.09 percent per annum. Total factor productivity for India grew at an average annual rate of 0.69 percent between 1970 and 1995 (Table 2). In the 1970s, total factor productivity grew at 1.37% per annum. But it grew fast in the 1980s, at 1.99 percent per annum. Since 1990, total factor productivity growth in Indian agriculture has declined, at a rate of �0.59 percent per annum. 6 Table 1--Agricultural research expenditures in India, 1964-95 Research Expenditures a/ Research Intensity Ratio b/ million 1990 Rs million 1990 PPPs % 1964 1,629 378 1965 1,581 367 0.21 1966 1,869 434 0.25 1967 1,590 369 0.18 1968 1,684 391 0.19 1969 1,879 436 0.20 1970 1,902 441 0.20 1971 1,886 438 0.21 1972 1,973 458 0.22 1973 1,741 404 0.17 1974 2,504 581 0.26 1975 3,178 737 0.33 1976 3,471 805 0.38 1977 3,965 920 0.38 1978 4,407 1,022 0.43 1979 4,148 962 0.45 1980 3,982 924 0.38 1981 4,128 958 0.39 1982 4,292 995 0.41 1983 4,695 1,089 0.40 1984 4,978 1,155 0.43 1985 4,572 1,061 0.39 1986 5,115 1,186 0.44 1987 6,011 1,394 0.50 1988 6,517 1,512 0.48 1989 6,507 1,509 0.46 1990 7,085 1,643 0.48 1991 6,873 1,594 0.46 1992 6,754 1,567 0.44 1993 7,280 1,689 0.44 1994 7,246 1,681 0.42 1995 7,293 1,692 0.43 a/ Agricultural research expenditures were obtained from the State Planning Commission, Government of India. The GDP deflator was used to deflate expenditures to 1995 prices. We then used the1995 exchange rate based on purchasing power party (PPP) to convert expenditures into 1995 international dollars. b/ The agricultural research intensity ratio is defined as agricultural research expenditure as a percentage of agricultural GDP. 7 Table 2--Agricultural technology, production and productivity growth in India, 1970 � 95 HYV Adoption Production Growth Productivity Growth Urban Food Price Index % % % % 1970 21 100 100 100.00 1971 24 100 99 98.88 1972 23 93 92 n.a 1973 25 97 98 101.23 1974 26 101 100 102.44 1975 29 114 113 n.a. 1976 31 105 103 n.a. 1977 34 115 112 n.a. 1978 36 119 114 97.37 1979 37 119 113 n.a. 1980 41 120 112 n.a. 1981 40 127 116 n.a. 1982 43 125 110 n.a. 1983 41 135 118 97.31 1984 45 131 114 n.a. 1985 44 141 120 n.a. 1986 46 133 114 n.a. 1987 48 136 114 95.33 1988 47 152 130 95.78 1989 53 168 134 95.78 1990 53 152 121 n.a. 1991 57 152 119 95.78 1992 56 153 118 94.44 1993 57 156 118 96.02 1994 64 165 118 n.a. 1995 59 n.a. n.a. n.a. Annual Growth Rate (%) 1970-79 6.25 1.95 1.37 1980-89 3.10 3.82 1.99 1990-95 2.10 2.09 -0.59 1970-95 4.19 2.11 0.69 Sources: HYV (high yielding variety), production and productivity growth data are from Fan, Hazell, and Thorat, (1999). Food price index is from the Indian Statistical Abstract. 8 3. URBAN POVERTY In the early 1970s, both rural and urban poverty rates were high with 57% of the rural population and 47% of the urban population living under the poverty line (Table 3). Due to the high growth in agriculture, the rural poverty rate declined to 45% by the mid-1980s. The urban poverty rate also declined to 36%. In addition to growth in urban income, the decline in real food prices relative to nonfood prices may have played a large role in this reduction. From the mid-1980s to 1987, rural poverty continued to decline to 39%, but urban poverty changed very little. The reduction in rural poverty during this period is mainly due to the development of rural nonfarm employment and increases in rural wages. The so-called �trickle down� benefits of agricultural growth for the rural poor were almost nonexistent since both agricultural production and productivity growth was largely stagnant. The impact of agricultural growth on urban poverty through lower food prices was also absent. There was a relatively rapid reduction in rural and urban poverty during the second half of the 1980s. The rapid increases in agricultural production and productivity is the major reason behind this reduction in rural poverty. The growth in agricultural production and productivity may have also contributed to urban poverty reduction by keeping food prices low. If fact, the relative food price index dropped by 2 percentage points during this period. In summary, whenever there is higher growth in agricultural production and productivity, rural poverty declines. But it is also true that urban poverty falls when agricultural growth is high. 9 Table 3--Poverty in India, 1970 - 95 Rural Poverty Urban Poverty Urban Poor Rural Poor Share of Urban Poor % % million million % 1970 57.61 47.16 51.69 256.53 16.77 1971 54.84 44.98 51.12 248.99 1972 n.a. n.a. n.a. n.a. n.a 1973 55.36 45.67 55.97 260.99 17.66 1974 55.72 47.96 61.07 267.46 18.59 1975 n.a. n.a. n.a. n.a. n.a 1976 n.a. n.a. n.a. n.a. n.a 1977 n.a. n.a. n.a. n.a. n.a 1978 50.60 40.50 59.95 259.54 18.77 1979 n.a. n.a. n.a. n.a. n.a 1980 n.a. n.a. n.a. n.a. n.a 1981 n.a. n.a. n.a. n.a. n.a 1982 n.a. n.a. n.a. n.a. n.a 1983 45.31 35.65 62.36 253.06 19.77 1984 n.a. n.a. n.a. n.a. n.a 1985 n.a. n.a. n.a. n.a. n.a 1986 n.a. n.a. n.a. n.a. n.a 1987 38.81 34.29 67.73 232.36 22.57 1988 39.60 35.65 72.55 241.10 23.13 1989 39.06 36.60 76.71 241.77 24.09 1990 34.30 33.40 72.06 215.79 25.03 1991 36.43 32.76 72.72 232.89 23.79 1992 40.00 33.50 74.36 259.76 22.26 1993 36.66 30.51 71.63 241.73 22.86 1994 41.00 33.50 80.88 274.36 22.77 1995 37.15 28.40 70.54 252.15 21.86 Sources: Rural and urban poverty rates are from Datt (1998), and the number of rural and urban poor was calculated by the author using rural and urban population data from FAO (2002). 10 4. ECONOMETRIC MODEL To analyze the links between agricultural research and urban poverty, we developed an econometric model in which an agricultural production function, price determination function, and urban poverty equation were estimated. This is because agricultural research investments affect poverty through changes in food prices, it is difficult to capture this link using a single equation approach. (1) TFP = h(RDE, RDE-1,� RDE-2, RDE-I, IR, ROADS, PVELE, LITE, GCSHEL, GERDEV, GCSSL, RAIN) (2) FP = g(TFP, GDP, POP,WPI, S) (3) UP = ƒ(FP, M, GINI, Z) Equation (1) models the determination of TFP growth in agriculture. The TFP growth index is the ratio of an aggregated output index to an aggregated input index. The following variables are included in the equation: current and lagged government spending in agricultural research and extension (RDE, RDE-1,... RDE-i); percentage of irrigated cropped area in total cropped area (IR); literacy rate of the rural population (LITE); road density (ROADS); percentage of villages electrified (PVELE), capital stocks of government investments in health (GCSHEL), rural development (GERDEV), and soil and water conservation (GCSSL); and annual rainfall (RAIN). The first seven variables should capture the productivity-enhancing effects of technologies, infrastructure, education, and other government spending in rural areas. The rainfall variable should capture weather effects. Inclusion of other public goods and government 11 spending variables will avoid overestimating the effects of agricultural research, and will allow comparing the effects of these public investments with agricultural research. Equation (2) models the determination of food prices (FP). Food prices are measured as a ratio of food prices to nonfood consumer prices. Growth in agricultural productivity (TFP) increases the supply of agricultural products and hence is expected to contribute to lower food prices. Per capita GDP (GDP) and population size (POP) are used to capture demand-side factors in the food markets. Food prices in India may have also been affected by international market prices (WPI), although during most of the study period the share of imports and exports in total domestic consumption was small, often less than 3%. Variable S, which consists of a set of state dummies, is intended to capture the effect of all other factors on changes in food prices. Equation (3) models the determinants of urban poverty (UP)2. Urban poverty is expected to be positively relate to food price increases relative to nonfood prices (FP) and to inequality in urban incomes (GINI), and negatively related to the per capita income of urban residents (M). Variable Z (which comprises year and province dummies) is included to capture the effects of all other omitted variables. 5. DATA AND MODEL ESTIMATION DATA State level data from 1970 to 1995 were used in the model estimation. Most of the data are taken from the official sources of the Indian Government (Fan, Hazell and Thorat, 2000). The head-count ratio data used in this analysis were constructed by Datt, and are published in a World Bank publication (World Bank 1997). Datt used the poverty line originally 2 To simplify the presentation, we have omitted to include subscripts to indicate observations in year t and at the province level. The variables with subscript "-1,...-j" indicate lagged observations for years t-1,...t-j. 12 defined by the Planning Commission, and more recently endorsed by the same agency, which is based on a nutritional norm of 2,400 calories per person per day. It is defined as the level of average per capita total expenditure at which this norm is typically attained, and is equal to a per capita monthly expenditure of Rs 57 at October 1973-June 1974 all-India urban prices. The mean income and Gini coefficients are also taken from Datt (1998). Our measure of total factor productivity growth has already been defined. But because of concerns that the measure of TFP used may be sensitive to the cost data used in aggregating inputs, a primal approach was also tried. By first estimating a production function for Indian agriculture using district level data, production elasticities for key inputs like land, labor, fertilizer, machinery, and animals were obtained and then used to construct an estimate of TFP growth at the state level. The results were similar to those obtained by using the cost shares (a dual approach). But the dual approach is preferred here because the elasticities used in the primal approach do not vary by states. The road density variable is defined as the length of road per unit of geographic area. Education is measured as the literacy rate, defined as the percentage of literate people in the total rural population above 7 years old. The irrigation variable is defined as the percentage of the total cropped area under irrigation. The electrification variable measures the percentage of all villages that have access to electricity. These variables were aggregated from district level data, which were obtained from the Planning Commission through the National Center for Agricultural Policy and Economics Research, New Delhi. The food price variable is measured as the change in food prices relative to nonfood prices in urban areas. GDP and population data are from World Bank database (2002). The world food price index is a weighted average price index for rice, wheat, and maize in the 13 international market, and the international prices of these commodities are taken form FAO (2002). Functional Form and Estimation Technique We used double-log functional forms for all equations in the model. More flexible functional forms, such as the translog or quadratic, impose fewer restrictions on estimated parameters, but many coefficients are not statistically significant due to multicollinearity problems among the many interaction variables. Lags and Distributions of R&D Investments Government investments in R&D can have long lead times in affecting agricultural production, as well as long-term effects once they kick in. One of the thornier problems to resolve when including agricultural research investments in a production function concerns the choice of an appropriate lag structure. Most past studies use stock variables which are usually weighted averages of current and past government expenditures on R&D. But what weights and how many years lag should be used in the aggregation are currently under hot debate.3 Since the shape and length of these investment lags are largely unknown, we use a free form lag structure in our analysis, i.e., we include current and past government expenditures on R&D in the production function. Then we use statistical tools to test and determine the appropriate length of lag for R&D expenditure. Various procedures have been suggested for determining the appropriate lag length. The adjusted R2 and Akaike's Information Criteria (AIC) are often used by many economists (Greene 1993). In this report, we simply use the adjusted R2. The optimal lag length is determined by the length of lag that maximizes the adjusted R2. The AIC is similar in sprit to the adjusted R2 in that 3Alston et al. (1998) argue that research lags may be much longer than previously thought, perhaps even infinite. But this argument may be less relevant for most developing countries since their national agricultural research systems are much younger and their research tends to be more applied and hence has shorter useful life. 14 it rewards goodness of fit, but it penalizes for the loss of degrees of freedom. The lag determined by the adjusted R2 approach is 13 years. Another problem related to the estimation of the lag structure is that the independent variables (RDE, RDE-1, RDE-2, ... and RDE-i) are often highly correlated, making the estimated coefficients statistically insignificant. Several ways of tackling this problem have been proposed. The most popular approach is to use what are called polynomial distributed lags, or PDLs. In a polynomial distributed lag, the coefficients are all required to lie on a polynomial of some degree d. In this study, we use PDLs of degree 2. In this case, we only need to estimate three instead of i+1 parameters for the lag distribution. For more detailed information on this subject, refer to Davidson and MacKinnon (1993). Once the lengths of lags are determined, we estimate the simultaneous equations system with the PDLs and appropriate lag length for research investment. Estimation Results The estimated model is presented in Table 4. Since we used double-log functional forms, the estimated coefficients are in elasticity form. The estimated agricultural productivity function (equation (1)) confirms that agricultural research, improved roads, irrigation, access to electricity, and education all contributed significantly to agricultural production over the sample period. The coefficient reported for agricultural R&D is the sum of the past 13 years coefficients from the PDLs distribution. The significance test is the joint t test of the three parameters of the PDL. The estimated food price equation (equation (2)) indicates that increases in agricultural output do exert a strong downward pressure on food prices with an elasticity of 0.231. Per capita GDP and total population size have positive, but statistically insignificant impacts on agricultural 15 prices. World food prices have a significant impact on domestic food prices, indicating that domestic urban food prices are linked with the international market. The estimated poverty equation 3 shows that food prices have a very significant impact on urban poverty. For every one percent decline (increase) in food prices, urban poverty is reduced (increased) by 0.35%. Growth in per capita income has also contributed significantly to rapid reductions in urban poverty while a worsening income distribution in urban areas has worked to increase urban poverty. 16 T ab le 4 -- E st im at es o f t he S im ul ta ne ou s E qu at io n Sy st em (1 ) TF P = -0 .0 26 + 0. 25 5 TR D E + 0. 21 5 IR + 0. 24 2 R O A D S + 0. 06 2 PV EL E + 0. 70 8 LI TE + 0. 01 2 G C SH EL (-0 .7 8) (1 .8 2) * (1 .8 3) * (2 .4 3) * (0 .6 0) (1 .9 5) * (0 .3 9) + 0. 02 2 G ER D EV + 0. 00 15 G C SS L + 0. 27 2 R A IN R 2 =0 .3 01 (0 .6 3) (0 .3 7) (5 .4 7) * (2 ) FP = 0. 02 5 - 0. 23 1T FP + 0. 11 2 G D P + 0. 03 4 PO P + 0. 27 1 W PI R 2 =0 .3 63 (2 .2 2) * (-3 .0 3) * (1 .5 6) (1 .6 7) (8 .0 3) * (3 ) U P = 7. 07 - 1. 63 7 M + 1. 00 3G IN I + 0. 35 0 FP R 2 =0 .9 11 (2 1. 15 )* (-2 3. 89 )* (1 5. 15 )* (1 .7 8) * N ot e: T he e st im at ed fi rs t e qu at io n is fr om F an e t a l, 20 00 . A st er is ks in di ca te st at is tic al si gn ifi ca nc e at th e 5% le ve l. T he c oe ffi ci en t f or R D E is th e su m o f t he co ef fic ie nt s f or th e pa st 1 3 ye ar s, an d th e t-v al ue o f t he c oe ff ic ie nt is th e jo in t t -v al ue o f t he c oe ffi ci en ts fo r t he p as t 1 3 ye ar s. 17 6. CONTRIBUTION OF AGRICULTURAL RESEARCH TO URBAN POVERTY REDUCTION By totally differentiating equations (1) - (3), the impact of government investment in agricultural R&D in year t-i on poverty at year t can be derived as: (4) dUP/dRDE-i = (∂UP/∂FP) (∂FP/∂Y) (∂Y/∂RDE-i). By aggregating the total effects of all past government expenditures on R&D over the lag period, the sum of marginal effects is obtained for any particular year. This is equivalent to the marginal impact of a change in the �stock� of R&D investment at time t, where the stock RS is measured as: RSt = atREt +at-1REt-13 + �, +at-13REt-13, and at-i coefficients are the estimated parameters in the production function (equation 1). The estimated elasticity of urban poverty to agricultural research is �0.021. That is, for every one percent increase in agricultural research investment, urban poverty declines by 0.021%. Using this elasticity and the values of the relevant variables for specific periods of time, we can calculate the number of poor urban people raised above the poverty line for an additional 1 million Rs increase in the stock of agricultural research investment. Similarly, we can calculate the total number of urban poor who were lifted out of poverty each year as a result of actual investments in agricultural research. The results are shown in Table 5. Each additional million Rs increase in the 1970 stock of agricultural research investment lifted 196 urban people out of poverty. This figure had declined to 72 people by 1995. Given actual levels of investment in agricultural research, then 1.21 million urban people were lifted out of poverty in 1970 and 1.70 million in 1995. This suggests that although marginal impact of 18 agricultural research on urban poverty reduction is declining, the total number of rural poor lifted out of poverty by agricultural research actually increased over time. Table 5: Impact of agricultural research on urban poverty Number of Poor Reduced per Million Rs (1995 price) Total Number of Poor Reduced (million) 1970 196.26 1.21 1971 215.87 1.32 1973 229.58 1.30 1974 166.07 1.35 1978 102.47 1.46 1983 103.03 1.57 1987 85.10 1.66 1988 73.73 1.56 1989 74.52 1.57 1990 69.99 1.61 1991 69.49 1.55 1992 79.86 1.75 1993 64.61 1.52 1994 68.66 1.61 1995 72.11 1.70 The results obtained here for the urban poor are quite comparable with similar calculations by Fan, Hazell and Thorat (2000) of the impact of agricultural research investments on the rural poor (Table 6). For example, for every one million Rs increase in the stock of agricultural research investment, 84.5 rural people were raised out of poverty in 1995. The large impact on rural poverty arises not only from the direct impact of increased agricultural productivity on the poor, but also from indirect nonfarm employment effects. Among all types investments in rural areas, agricultural research has the largest impact on urban poverty, almost three times higher than road investments, which have the second largest impact. The total poverty effect (combining both rural and urban poor) of agricultural research 19 investment is also the largest. For every additional one million Rs spent on agricultural research, 157 poor people are lifted above the poverty line. Road investments have almost as a large an impact; each additional one million Rs spent raises 152 poor people above the poverty line. Table 6--Number of urban and rural poor reduced per million Rs, 1995 Urban Poor Rural Poor Total Poor Agricultural R&D 72.11 84.5 156.61 Irrigation 7.31 9.7 17.01 Rural Roads 28.39 123.8 152.19 Rural Education 7.43 41 48.43 Rural Electricity 1.44 3.8 5.24 Soil and Water Conservation 5.15 22.6 27.75 Rural Development 5.87 25.5 31.37 Rural Health 4.55 17.8 22.35 Note: The relationships between government investments and physical stocks for different types of government spending were taken from Fan, Hazell, and Thorat (2000). These relationships are used to calculate the marginal returns for poverty reduction. Source: The figures on the rural poor taken from Fan, Hazell and Thorat (2000); the figures on the urban poor are the author�s calculations. 7. CONCLUSIONS This study has estimated the impact of agricultural research investments on urban poverty in India using time series and cross-state data and an econometric modeling approach. The model explicitly tracks the causal links between agricultural research investments and subsequent production increases in agriculture, and how this impacts on food prices and the incidence of urban poverty. The results show that agricultural research has played an important role in reducing urban poverty in India. Without investments in agricultural research, urban poverty in India would be much higher today. Each one million Rs increase in the stock of agricultural research investment raises about as many urban people as rural people above the poverty line. 20 With rapid urbanization, agricultural research will still need to play a key role in supplying adequate food at affordable prices to ensure that urban and rural poverty remain low. But since 1990, agricultural research investment in India has stagnated. By 1997, government investment in agricultural research as a percentage of agricultural GDP was only about 0.4%. This is extremely low when compared with 2-3% in many developed countries, and is even lower than the average of 0.5% for all developing countries. One result of this stagnation in investment was that both rural and urban poverty declined at a slower rate in the 1990s than in the 1970s and 1980s. Today, the urban poor account for a quarter of India�s total poor. It is projected that more than half the Indian population will reside in urban cities by 2030 and the poor will be urbanized faster than the general population (Ravallion, 2001). India has made great success in feeding its large and growing population and in reducing both rural and urban poverty during recent decades through government investments in agricultural research, rural infrastructure, and education. But India cannot afford to be complacent. Continued government support for these investments is still needed, otherwise food insecurity, malnutrition, poverty, and social conflict will shadow India for a long time to come. 21 REFERENCES Akino, M. and Y. Hayami. 1975. Efficiency and equity in public research: Rice breeding in Japan's economic development. American Journal of Agricultural Economics 57(1): 1-10. Alston, J., C. Chan-Kang, M. Marra, P. Pardey, and T. Wyatt. 2000. A meta-analysis of rates of return to agricultural R&D, Ex Pede Herculem? Research Report 113, Washington, DC: International Food Policy Research Institute. Alston, J., B. Craig, and P. Pardey. 1998. Dynamics in the creation and depreciation of knowledge, and the returns to research. Environment and Production Technology Division Discussion Paper No. 35. Washington DC: International Food Policy Research Institute. Davidson, R. and J. MacKinnon. 1993. Estimation and inference in econometrics. New York and London: Oxford University Press. Evenson, R. E., and P. M. Flores. 1978. Social returns to rice research. In Economic Consequences of the New Rice Technology. Los Banos, Philippines: International Rice Research Institute. Evenson R., Pray, C., and M. Rosegrant. 1998. Agricultural Research and Productivity Growth in India. Research Report 109, Washington DC: International Food Policy Research Institute. Fan, S., C. Fang, and X. Zhang. 2001. How agricultural research affect urban poverty in developing countries: The case of China,� EPTD discussion paper #83, Washington DC: International Food Policy Research Institute. Fan, S., P. Hazell, and S. Thorat. 2000. Government spending, agricultural growth and poverty in rural India. American Journal of Agricultural Economics, 82 (4): Greene, W.H. 1993. Econometric analysis. Prentice-Hall, Inc. Haddad, L., M. T. Ruel, and J. L. Garret. 1999. Are urban poverty and undernutrition growing? Some newly assembled evidence. Food Consumption and Nutrition Division Discussion Paper No. No. 63. Washington, DC: International Food Policy Research Institute. Hazell, P., and L. Haddad. 2001. Agricultural research and poverty reduction. IFPRI 2020 Discussion Paper. Washington, DC: International Food Policy Research Institute, Washington D.C. 22 Mellor, J. W. 1975. The impact of new agricultural technology on employment and income distribution--Concepts and policy. Occasional Paper No. 2., Washington DC: USAID. Pinstrup-Andersen, P. 1979. Modern agricultural technology and income distribution: The market price effect. European Review of Agricultural Economics 6(1): 17-46. Pinstrup-Anderson, P. and P. Hazell. 1985. The impact of green revolution and prospects for the future, Food Review International, 1(1). Ravillion, M. 2000. On the urbanization of poverty. World Bank Discussion Paper. Washington, DC: World Bank. Scobie, G. and R. Posada. 1978. The impact of technological change on income distribution: The case of rice in Colombia. American Journal of Agricultural Economics, 60(1): 85-92. EPTD DISCUSSION PAPERS LIST OF EPTD DISCUSSION PAPERS 01 Sustainable Agricultural Development Strategies in Fragile Lands, by Sara J. Scherr and Peter B.R. Hazell, June 1994. 02 Confronting the Environmental Consequences of the Green Revolution in Asia, by Prabhu L. Pingali and Mark W. Rosegrant, August 1994. 03 Infrastructure and Technology Constraints to Agricultural Development in the Humid and Subhumid Tropics of Africa, by Dunstan S.C. Spencer, August 1994. 04 Water Markets in Pakistan: Participation and Productivity, by Ruth Meinzen-Dick and Martha Sullins, September 1994. 05 The Impact of Technical Change in Agriculture on Human Fertility: District-level Evidence From India, by Stephen A. Vosti, Julie Witcover, and Michael Lipton, October 1994. 06 Reforming Water Allocation Policy Through Markets in Tradable Water Rights: Lessons from Chile, Mexico, and California, by Mark W. Rosegrant and Renato Gazri S, October 1994. 07 Total Factor Productivity and Sources of Long-Term Growth in Indian Agriculture, by Mark W. Rosegrant and Robert E. Evenson, April 1995. 08 Farm-Nonfarm Growth Linkages in Zambia, by Peter B.R. Hazell and Behjat Hoijati, April 1995. 09 Livestock and Deforestation in Central America in the 1980s and 1990s: A Policy Perspective, by David Kaimowitz (Interamerican Institute for Cooperation on Agriculture), June 1995. 10 Effects of the Structural Adjustment Program on Agricultural Production and Resource Use in Egypt, by Peter B.R. Hazell, Nicostrato Perez, Gamal Siam, and Ibrahim Soliman, August 1995. 11 Local Organizations for Natural Resource Management: Lessons from Theoretical and Empirical Literature, by Lise Nordvig Rasmussen and Ruth Meinzen-Dick, August 1995. EPTD DISCUSSION PAPERS 12 Quality-Equivalent and Cost-Adjusted Measurement of International Competitiveness in Japanese Rice Markets, by Shoichi Ito, Mark W. Rosegrant, and Mercedita C. Agcaoili-Sombilla, August 1995. 13 Role of Inputs, Institutions, and Technical Innovations in Stimulating Growth in Chinese Agriculture, by Shenggen Fan and Philip G. Pardey, September 1995. 14 Investments in African Agricultural Research, by Philip G. Pardey, Johannes Roseboom, and Nienke Beintema, October 1995. 15 Role of Terms of Trade in Indian Agricultural Growth: A National and State Level Analysis, by Peter B.R. Hazell, V.N. Misra, and Behjat Hoijati, December 1995. 16 Policies and Markets for Non-Timber Tree Products, by Peter A. Dewees and Sara J. Scherr, March 1996. 17 Determinants of Farmers� Indigenous Soil and Water Conservation Investments in India�s Semi-Arid Tropics, by John Pender and John Kerr, August 1996. 18 Summary of a Productive Partnership: The Benefits from U.S. Participation in the CGIAR, by Philip G. Pardey, Julian M. Alston, Jason E. Christian, and Shenggen Fan, October 1996. 19 Crop Genetic Resource Policy: Towards a Research Agenda, by Brian D. Wright, October 1996. 20 Sustainable Development of Rainfed Agriculture in India, by John M. Kerr, November 1996. 21 Impact of Market and Population Pressure on Production, Incomes and Natural Resources in the Dryland Savannas of West Africa: Bioeconomic Modeling at the Village Level, by Bruno Barbier, November 1996. 22 Why Do Projections on China�s Future Food Supply and Demand Differ? by Shenggen Fan and Mercedita Agcaoili-Sombilla, March 1997. 23 Agroecological Aspects of Evaluating Agricultural R&D, by Stanley Wood and Philip G. Pardey, March 1997. 24 Population Pressure, Land Tenure, and Tree Resource Management in Uganda, by Frank Place and Keijiro Otsuka, March 1997. EPTD DISCUSSION PAPERS 25 Should India Invest More in Less-favored Areas? by Shenggen Fan and Peter Hazell, April 1997. 26 Population Pressure and the Microeconomy of Land Management in Hills and Mountains of Developing Countries, by Scott R. Templeton and Sara J. Scherr, April 1997. 27 Population Land Tenure and Natural Resource Management: The Case of Customary Land Area in Malawi, by Frank Place and Keijiro Otsuka, April 1997. 28 Water Resources Development in Africa: A Review and Synthesis of Issues, Potentials, and Strategies for the Future, by Mark W. Rosegrant and Nicostrato D. Perez, September 1997. 29 Financing Agricultural R&D in Rich Countries: What�s Happening and Why? by Julian M. Alston, Philip G. Pardey, and Vincent H. Smith, September 1997. 30 How Fast Have China�s Agricultural Production and Productivity Really Been Growing? by Shenggen Fan, September 1997. 31 Does Land Tenure Insecurity Discourage Tree Planting? Evolution of Customary Land Tenure and Agroforestry management in Sumatra, by Keijiro Otsuka, S. Suyanto, and Thomas P. Tomich, December 1997. 32 Natural Resource Management in the Hillsides of Honduras: Bioeconomic Modeling at the Micro-Watershed Level, by Bruno Barbier and Gilles Bergeron, January 1998. 33 Government Spending, Growth, and Poverty: An Analysis of Interlinkages in Rural India, by Shenggen Fan, Peter Hazell, and Sukhadeo Thorat, March 1998. Revised December 1998. 34 Coalitions and the Organization of Multiple-Stakeholder Action: A Case Study of Agricultural Research and Extension in Rajasthan, India, by Ruth Alsop, April 1998. 35 Dynamics in the Creation and Depreciation of Knowledge and the Returns to Research, by Julian Alston, Barbara Craig, and Philip Pardey, July, 1998. 36 Educating Agricultural Researchers: A Review of the Role of African Universities, by Nienke M. Beintema, Philip G. Pardey, and Johannes Roseboom, August 1998. EPTD DISCUSSION PAPERS 37 The Changing Organizational Basis of African Agricultural Research, by Johannes Roseboom, Philip G. Pardey, and Nienke M. Beintema, November 1998. 38 Research Returns Redux: A Meta-Analysis of the Returns to Agricultural R&D, by Julian M. Alston, Michele C. Marra, Philip G. Pardey, and T.J. Wyatt, November 1998. 39 Technological Change, Technical and Allocative Efficiency in Chinese Agriculture: The Case of Rice Production in Jiangsu, by Shenggen Fan, January 1999. 40 The Substance of Interaction: Design and Policy Implications of NGO-Government Projects in India, by Ruth Alsop with Ved Arya, January 1999. 41 Strategies for Sustainable Agricultural Development in the East African Highlands, by John Pender, Frank Place, and Simeon Ehui, April 1999. 42 Cost Aspects of African Agricultural Research, by Philip G. Pardey, Johannes Roseboom, Nienke M. Beintema, and Connie Chan-Kang, April 1999. 43 Are Returns to Public Investment Lower in Less-favored Rural Areas? An Empirical Analysis of India, by Shenggen Fan and Peter Hazell, May 1999. 44 Spatial Aspects of the Design and Targeting of Agricultural Development Strategies, by Stanley Wood, Kate Sebastian, Freddy Nachtergaele, Daniel Nielsen, and Aiguo Dai, May 1999. 45 Pathways of Development in the Hillsides of Honduras: Causes and Implications for Agricultural Production, Poverty, and Sustainable Resource Use, by John Pender, Sara J. Scherr, and Guadalupe Durón, May 1999. 46 Determinants of Land Use Change: Evidence from a Community Study in Honduras, by Gilles Bergeron and John Pender, July 1999. 47 Impact on Food Security and Rural Development of Reallocating Water from Agriculture, by Mark W. Rosegrant and Claudia Ringler, August 1999. 48 Rural Population Growth, Agricultural Change and Natural Resource Management in Developing Countries: A Review of Hypotheses and Some Evidence from Honduras, by John Pender, August 1999. EPTD DISCUSSION PAPERS 49 Organizational Development and Natural Resource Management: Evidence from Central Honduras, by John Pender and Sara J. Scherr, November 1999. 50 Estimating Crop-Specific Production Technologies in Chinese Agriculture: A Generalized Maximum Entropy Approach, by Xiaobo Zhang and Shenggen Fan, September 1999. 51 Dynamic Implications of Patenting for Crop Genetic Resources, by Bonwoo Koo and Brian D. Wright, October 1999. 52 Costing the Ex Situ Conservation of Genetic Resources: Maize and Wheat at CIMMYT, by Philip G. Pardey, Bonwoo Koo, Brian D. Wright, M. Eric van Dusen, Bent Skovmand, and Suketoshi Taba, October 1999. 53 Past and Future Sources of Growth for China, by Shenggen Fan, Xiaobo Zhang, and Sherman Robinson, October 1999. 54 The Timing of Evaluation of Genebank Accessions and the Effects of Biotechnology, by Bonwoo Koo and Brian D. Wright, October 1999. 55 New Approaches to Crop Yield Insurance in Developing Countries, by Jerry Skees, Peter Hazell, and Mario Miranda, November 1999. 56 Impact of Agricultural Research on Poverty Alleviation: Conceptual Framework with Illustrations from the Literature, by John Kerr and Shashi Kolavalli, December 1999. 57 Could Futures Markets Help Growers Better Manage Coffee Price Risks in Costa Rica? by Peter Hazell, January 2000. 58 Industrialization, Urbanization, and Land Use in China, by Xiaobo Zhang, Tim Mount, and Richard Boisvert, January 2000. 59 Water Rights and Multiple Water Uses: Framework and Application to Kirindi Oya Irrigation System, Sri Lanka, by Ruth Meinzen-Dick and Margaretha Bakker, March 2000. 60 Community natural Resource Management: The Case of Woodlots in Northern Ethiopia, by Berhanu Gebremedhin, John Pender and Girmay Tesfaye, April 2000. EPTD DISCUSSION PAPERS 61 What Affects Organization and Collective Action for Managing Resources? Evidence from Canal Irrigation Systems in India, by Ruth Meinzen-Dick, K.V. Raju, and Ashok Gulati, June 2000. 62 The Effects of the U.S. Plant Variety Protection Act on Wheat Genetic Improvement, by Julian M. Alston and Raymond J. Venner, May 2000. 63 Integrated Economic-Hydrologic Water Modeling at the Basin Scale: The Maipo River Basin, by M. W. Rosegrant, C. Ringler, D.C. McKinney, X. Cai, A. Keller, and G. Donoso, May 2000. 64 Irrigation and Water Resources in Latin America and he Caribbean: Challenges and Strategies, by Claudia Ringler, Mark W. Rosegrant, and Michael S. Paisner, June 2000. 65 The Role of Trees for Sustainable Management of Less-favored Lands: The Case of Eucalyptus in Ethiopia, by Pamela Jagger & John Pender, June 2000. 66 Growth and Poverty in Rural China: The Role of Public Investments, by Shenggen Fan, Linxiu Zhang, and Xiaobo Zhang, June 2000. 67 Small-Scale Farms in the Western Brazilian Amazon: Can They Benefit from Carbon Trade? by Chantal Carpentier, Steve Vosti, and Julie Witcover, September 2000. 68 An Evaluation of Dryland Watershed Development Projects in India, by John Kerr, Ganesh Pangare, Vasudha Lokur Pangare, and P.J. George, October 2000. 69 Consumption Effects of Genetic Modification: What If Consumers Are Right? by Konstantinos Giannakas and Murray Fulton, November 2000. 70 South-North Trade, Intellectual Property Jurisdictions, and Freedom to Operate in Agricultural Research on Staple Crops, by Eran Binenbaum, Carol Nottenburg, Philip G. Pardey, Brian D. Wright, and Patricia Zambrano, December 2000. 71 Public Investment and Regional Inequality in Rural China, by Xiaobo Zhang and Shenggen Fan, December 2000. 72 Does Efficient Water Management Matter? Physical and Economic Efficiency of Water Use in the River Basin, by Ximing Cai, Claudia Ringler, and Mark W. Rosegrant, March 2001. EPTD DISCUSSION PAPERS 73 Monitoring Systems for Managing Natural Resources: Economics, Indicators and Environmental Externalities in a Costa Rican Watershed, by Peter Hazell, Ujjayant Chakravorty, John Dixon, and Rafael Celis, March 2001. 74 Does Quanxi Matter to NonFarm Employment? by Xiaobo Zhang and Guo Li, June 2001. 75 The Effect of Environmental Variability on Livestock and Land-Use Management: The Borana Plateau, Southern Ethiopia, by Nancy McCarthy, Abdul Kamara, and Michael Kirk, June 2001. 76 Market Imperfections and Land Productivity in the Ethiopian Highlands, by Stein Holden, Bekele Shiferaw, and John Pender, August 2001. 77 Strategies for Sustainable Agricultural Development in the Ethiopian Highlands, by John Pender, Berhanu Gebremedhin, Samuel Benin, and Simeon Ehui, August 2001. 78 Managing Droughts in the Low-Rainfall Areas of the Middle East and North Africa: Policy Issues, by Peter Hazell, Peter Oram, Nabil Chaherli, September 2001. 79 Accessing Other People�s Technology: Do Non-Profit Agencies Need It? How To Obtain It, by Carol Nottenburg, Philip G. Pardey, and Brian D. Wright, September 2001. 80 The Economics of Intellectual Property Rights Under Imperfect Enforcement: Developing Countries, Biotechnology, and the TRIPS Agreement, by Konstantinos Giannakas, September 2001. 81 Land Lease Markets and Agricultural Efficiency: Theory and Evidence from Ethiopia, by John Pender and Marcel Fafchamps, October 2001. 82 The Demand for Crop Genetic Resources: International Use of the U.S. National Plant Germplasm System, by M. Smale, K. Day-Rubenstein, A. Zohrabian, and T. Hodgkin, October 2001. 83 How Agricultural Research Affects Urban Poverty in Developing Countries: The Case of China, by Shenggen Fan, Cheng Fang, and Xiaobo Zhang, October 2001. 84 How Productive is Infrastructure? New Approach and Evidence From Rural India, by Xiaobo Zhang and Shenggen Fan, October 2001. EPTD DISCUSSION PAPERS 85 Development Pathways and Land Management in Uganda: Causes and Implications, by John Pender, Pamela Jagger, Ephraim Nkonya, and Dick Sserunkuuma, December 2001. 86 Sustainability Analysis for Irrigation Water Management: Concepts, Methodology, and Application to the Aral Sea Region, by Ximing Cai, Daene C. McKinney, and Mark W. Rosegrant, December 2001. 87 The Payoffs to Agricultural Biotechnology: An Assessment of the Evidence, by Michele C. Marra, Philip G. Pardey, and Julian M. Alston, January 2002. 88 Economics of Patenting a Research Tool, by Bonwoo Koo and Brian D. Wright, January 2002. 89 Assessing the Impact of Agricultural Research On Poverty Using the Sustainable Livelihoods Framework, by Michelle Adato and Ruth Meinzen-Dick, March 2002. 90 The Role of Rainfed Agriculture in the Future of Global Food Production, by Mark Rosegrant, Ximing Cai, Sarah Cline, and Naoko Nakagawa, March 2002. 91 Why TVEs Have Contributed to Interregional Imbalances in China, by Junichi Ito, March 2002. 92 Strategies for Stimulating Poverty Alleviating Growth in the Rural Nonfarm Economy in Developing Countries, by Steven Haggblade, Peter Hazell, and Thomas Reardon, July 2002. 93 Local Governance and Public Goods Provisions in Rural China, by Xiaobo Zhang , Shenggen Fan, Linxiu Zhang, and Jikun Huang, July 2002.