1 INTRODUCTION Nepal is a landlocked country of 27 million people residing across a land area of more than 147,000 km 2 . Nepal is divided into three major geographic regions (Figure 1.1): the Terai and inner Terai; (34,000 km 2 ); the Hills (61,000 km 2 ); and the Mountains (52,000 km 2 ). The Terai and inner Terai lie at the lowest altitudes and are the most productive regions agriculturally. Mostly they enjoy subtropical warm humid climatic conditions favorable for growing three crops in a year. There is considerable variation in climate from east to west and great diversity for water accord- ing to local topography and the availability of irrigation. When agroecology is considered, the hill regions are divided into three (Table 1.1) resulting in five distinct domains that vary in their resource endowment, land-use systems, farming systems, cropping patterns and intensities, and access to road and market networks. Population distribution in Nepal is related more to the proportion of cultivated land than to the geographical area. For example, Terai and inner Terai are home to 48.4 percent of the population and cover 56 percent of the total cultivated land; hills account for 44.3 percent of population and 37 percent of cultivated land while mountains accommodate 7.3 percent of the total population and have only 7 percent of the cultivated land (Ne- pal, National Planning Commission–WFP–NDRI 2010). FIGURE 1.1—MAJOR PHYSIOGRAPHIC REGIONS OF NEPAL Source: Nepal, National Planning Commission 2010a. TABLE 1.1—AGROECOLOGICAL DOMAINS OF NEPAL Source: Vaidya and Floyd 1997. Agroecological domain Altitude range (m) Local names for the region Importance in terms of agricultural production Terai and river basin 80–600 Terai, Tar, Bensi, Phant, Khonch, Kachad Most important Low hills 600–1,000 TalloPahad Third most important Middle hills 1,000– 1,600 Deurali, Hatiya, Madhya Pahad Second most important High hills 1,600– 2,300 Lekh, Kharka Fourth most important Mountains >2,300 Himal Least important PROJECT PAPER | DECEMBER 2012 Agriculture, seed, and innovation in Nepal: Industry and policy issues for the future K. D. Joshi, C. Conroy and J. R. Witcombe and 2 SUMMARY | APRIL 2010 Agriculture, seed, and innovation in Nepal: Industry and policy issues for the future Brief historical background to agriculture development in Nepal The predominant agriculture system in Nepal was pastoralism and in the 12th century the introduction of terracing, plowing methods and the construction of irrigation systems favored the intensification of agriculture. Rice, barley, several species of millets, and legumes were native to the region, while the introduction of maize and potatoes in the 18th century allowed an increase in cropping intensities par- ticularly in the Bari lands (upland) of higher altitudes. This resulted in expansion of the area under cul- tivation from the valley bottoms through to terraces on the valley slopes (Seddon et al. 1979). During the 19th century, population growth influenced the process of intensification of agriculture, and led to opening up of virgin lands for cultivation, especially in the Terai. In the 1950s, the eradication of malaria in many areas of the Terai and the clearance of forests accelerated the migration of hill peo- ple, as well as people from India, to the Terai. Current agricultural situation Nepal is predominantly an agrarian country with 88 percent of the population living in rural areas and 78 percent of the adult rural population engaged in agriculture (Nepal, Ministry of Agriculture and Co- operatives 2008). Nepal currently has 3.2 million hectares of cultivated land (and nearly 1.0 million hectares of additional land that could be brought into cultivation) (LRMP 1986). The 3.4 million total holdings average about 0.8 hectare of land per holding. In this mountainous country, only 16 percent of the land is agricultural land (Nepal, National Planning Commission–WFP–NDRI 2010). Because of rapid increase in population, land holdings are fragmented and scattered. As a result, agri- culture is mostly subsistence in nature. As per the census of 2001, 61 percent of the total population engaged in agriculture and 78 percent of farming households have their own agricultural land (Nepal, National Planning Commission 2002) but the holdings are small. Nearly half (47 percent) of all hold- ings are less than 0.5 hectare and 28 percent are between 0.5 and 1 hectare land area (so 75 percent of the total holdings are less than 1 hectare) (Nepal, National Planning Commission 2003). Given av- erage yields, a farm household would need at least 0.64 hectare of land in the mountains, 0.52 hectare in the hills, and 0.42 hectares in the Terai to produce enough food to feed a family of 6 members (Ne- pal, National Planning Commission–WFP–NDRI 2010). The share of cultivable land per person for the mountain region (0.3 hectare) comfortably fulfills this criteria, while the hill and Terai regions are facing extreme population pressure with implications for producing adequate food (Nepal, National Planning Commission 2006; Subedi 2003). Nepalese farmers grow crops under uncertain environments as rain- fed farming accounts nearly for two-thirds of the cultivated area; farmers cultivating rain-fed farms live in the most marginal, risk-prone, and poverty-stricken regions in the country (Nepal, National Planning Commission 2008). The agriculture sector accounts for one-third of the gross domestic product (GDP). Its contribution to employment, especially in rural areas, is quite considerable and there is a high dependence on agricul- ture as the major source of income. Returns from the agriculture sector have remained low and the per capita GDP is only US$140 per agricultural worker (Nepal, Ministry of Finance 2008). With little scope for increasing area under cultivation, population growth has led to falling average farm size and increasing fragmentation, which results in growing poverty. However, because there is cur- rently very low agricultural productivity, there is great potential for a rapid improvement through mod- ern technologies. To translate the potential into reality, huge investment is needed in areas such as ir- rigation, agricultural roads, land improvement, agricultural mechanization, equipment, and the use of purchased inputs. 3 SUMMARY | APRIL 2010 Agriculture, seed, and innovation in Nepal: Industry and policy issues for the future Poverty and human development in the context of current state of transition Nepal is in a state of transition. Following the recent cessation of civil unrest through a comprehensive peace accord, the country remains in the (long) process of constitution writing. This has important im- plications for the general peace and security and the pace of development in the country. Poverty can be measured in various ways. The Government of Nepal uses the cost-of-basic needs (BCN) 1 approach. Government statistics vary in their assessments of the extent of poverty. For exam- ple, the Nepal Living Standard Survey 2 (NLSS) estimated that nearly 31 percent of the total population was living below the national poverty line and 41 percent were consuming less than the minimum calo- rie requirements in 2010. In terms of purchasing power parity (PPP), the percentage below the poverty line would be more than 24 percent (Nepal, Ministry of Finance 2010). In contrast, the Multidimension- al Poverty Index 3 (MPI) indicates more than 18 million (66 percent) living in poverty, and 78 percent of them with an income of up to $2 a day (Table 1.2). Out of these, 67 percent are engaged in agro- based employment and 11 percent work as agricultural laborers. In spite of long social unrest and almost stagnant growth, there were positive changes in education and health indicators; average literacy and net enrollment rate in primary education and values for some of the health indicators improved compared with the previous decade (Nepal National Planning Commission, 2010b). Some of the indicators, such as average literacy and net enrollment rate in pri- mary education have achieved high levels (Table 1.2). The income gap between urban and rural, as well as among various geographical regions and groups, is still extremely wide in Nepal. The Gini coefficient, a widely used indicator for income inequality, was 0.34 in 1992–93 but reached 0.41 in 2003–04 suggesting that the income growth rate of the rich has been higher than that of the poor (Nepal, Ministry of Finance 2010). TABLE 1.2—DEVELOPMENT INDICATORS FOR NEPAL Indicator Value Year Source Population size (millions) 30.49 2011 Preliminary census results by CBS, NPC Population growth rate (%) 2.25 2009–10 CBS 2010 Life expectancy at birth (years) 63.7 2006 UNDP 2009 GNI per capita (US$) 472 2008–09 CBS 2009 Human Development Index (HDI) 0.509 2006 UNDP 2009 Real GDP growth rate (average %) 3.95 2008–09 CBS 2009 Literacy rate, 15–24 years (%) 86.5 2008 CBS 2008 Net enrollment rate in primary education (%) 93.7 2009 DOE 2009 Underweight children under age five (%) 38.6 2006 MOHP et al. 2007 Under-five mortality rate (per 1,000 live births) 50 2009 FHD 2009 Maternal mortality rate (per 100,000 live births) 229 2009 FHD 2009 HIV/AIDS prevalence, 15–49 years (%) 0.49 2007 NCASC 2009 Population below national poverty line (%) 25.4 2009 NPC 2010 Multidimensional Poverty Index (MPI)§ 0.35 2010 University of Oxford Proportion of MPI poor (H) 0.65 2010 University of Oxford Average intensity of deprivation (A) 0.54 2010 University of Oxford 1 The BCN approach establishes a poverty line that represents the level of per capita expenditure required to meet basic needs. For Nepal, an average of Rs.7, 696 (US$110) per year per person has been considered as the national poverty line and the minimum calorie requirement has been set at 2,144 kilocalorie per person per day (Nepal, National Planning Commission 2004). 2 The list of abbreviations and acronyms is annexed to this paper. 3 Multidimensional Poverty Index (MPI) was constructed by Oxford Poverty and Human Development Initiative (OPHI) for the UNDP’s 2010 Develop- ment Report (http://hdr.undp.org/en.). MPI = H*A Where H =Incidence of Poverty and A=Average intensity across the poor http://hdr.undp.org/en 4 SUMMARY | APRIL 2010 Agriculture, seed, and innovation in Nepal: Industry and policy issues for the future Number of MPI poor people (millions) 18.3 2010 University of Oxford Share of $1.25-a-day income poor (%) 55 2010 University of Oxford Share of $2-a-day poor (%) 78 2010 University of Oxford HDI category Medium 2010 University of Oxford Sources: As indicated. Note: GNI = gross national income. Poverty is not evenly spread across the country as a result of differences in various factors and their in- teractions. At 35 percent, the rural poverty level is more than three times as high as the urban poverty level in 2003–04 (Nepal, National Planning Commission 2005), and agriculture shows the highest con- centration of poverty (Gulati et al. 2010). The hills and mountains have a higher percentage of poor people than the Terai (Table 1.3), but in terms of absolute numbers the Terai has the highest concen- tration of poor people due to high population density. The Hills and Terai of the western and far western regions, as well as the hills and mountains of the eastern development regions, are the poorest with more than half of the population living below poverty. There was considerable reduction in poverty in the mountains and Terai from 1995–96 to 2003–04. However, in the hills there was a 13% increase in the proportion of poor people (Table 1.3) even though there was a decrease in the poverty head count during the same period. Despite overall reduction in poverty levels in the country, poverty for those self- employed in agriculture increased by 10 percentage points. One of the major contributing factors in poverty reduction has been the income from remittances, which accounts for more than 20 percent of GDP (Gulati et al. 2010). TABLE 1.3—POVERTY HEAD COUNT AND DISTRIBUTION OF POOR POPULATION IN NEPAL Source: Nepal, National Planning Commission 2005, adapted from Gulati et al. 2010. Food security in Nepal Nepal is one of the most food-insecure countries in Asia in large part because the average growth rate of major cereals is far below the average population growth rate. In fact, Nepal is the only country in South Asia where population growth rate surpasses the growth rate of cereals, although Pakistan is near equi- librium (Figure 1.2). Growth rate of rice, maize, and wheat production in Nepal is the lowest in South Asia (Table 1.4). Yields of cereals in Nepal are comparable with the Indian state of Bihar, which shares similar agroclimatic conditions and cropping systems. Description of poverty Poverty head count (%) Distribution of the poor population (%) 1995–96 2003–04 % change 1995–96 2003–04 % change Nepal 41.8 30.8 -26 100.0 100.0 - Urban 21.6 9.6 -56 3.6 4.7 31 Rural 43.3 34.6 -20 96.4 95.3 -1 Mountain 57.0 32.6 -43 10.7 7.5 -30 Hill 40.7 34.5 -15 41.9 47.1 12 Terai 40.3 27.6 -32 47.4 45.4 -4 Self-employed, agriculture 43.1 32.9 -24 60.7 66.9 10 Wage-earner, agriculture 55.9 53.8 -4 15.7 10.9 -31 Farmers with <0.2 hectare of land 48.0 39.0 -19 23.0 25.0 9 Farmers with 0.2–1.0hectare of land 45.0 38.0 -16 44.0 51.0 16 5 SUMMARY | APRIL 2010 Agriculture, seed, and innovation in Nepal: Industry and policy issues for the future FIGURE 1.2—AVERAGE ANNUAL GROWTH RATES OF POPULATION AND CEREAL PRODUC- TION IN SOUTH ASIAN COUNTRIES Source: Cereal production data from FAOSTAT 2010; population growth data from World Bank 2010. The combined impacts of stagnant economic growth (including agriculture), extreme variation in climatic con- ditions, and natural disasters have jointly contributed to escalating food prices and increased food insecurity (Nepal, National Planning Commission–WFP–NDRI 2010). The annual average price of food increased by nearly 17 percent in 2008–09 compared to a 10 percent rise in 2007–08. The price of some commodities, such as sugar and related products, increased by nearly 46 percent (NRB 2009). Food security in Nepal is characterized by considerable regional and seasonal variations. February to March and July to August, for example, are the hungry or agriculture-lean periods when people are unable to meet minimum nutrition and consumption needs over a sustained period of time. Chronic food insecurity often stems from extended peri- ods of poverty, restricted access to resources, and limited or no assets. Because of the longer crop-growing season, hungry periods last longer in the hills and mountains than in the Terai (Nepal, National Planning Commission–WFP–NDRI 2010). TABLE 1.4—AVERAGE PRODUCTIVITY OF RICE, MAIZE AND WHEAT IN SOUTH ASIA AND SOME OF THE INDIAN STATES Source: Country-level yields from FAOSTAT 2010; yields of Indian states from Ministry of Agriculture, Government of India. Country Average productivity (t/hectare) Rice Maize Wheat Bangladesh 4.0 5.8 1.9 India 3.3 2.2 2.7 Nepal 2.7 2.1 2.2 Pakistan 3.3 3.4 2.6 Sri Lanka 3.8 1.8 Bihar 2.1 2.3 1.9 Uttar Pradesh 3.3 1.4 2.7 6 SUMMARY | APRIL 2010 Agriculture, seed, and innovation in Nepal: Industry and policy issues for the future Poverty and food insecurity are always interrelated. On the one hand, poor people lack the means to produce suffi- cient food or purchase food in the market. On the other, food insecurity can also lead to poverty when people have to borrow or sell assets in order to buy food. The severity of hunger in Nepal is extreme. With its ranking of 20.6 on the 2008 Global Hunger Index (GHI), Nepal ranks 57th out of 88 developing countries and countries in transition. The highest prevalence of hunger can be found in the far- and mid-western hill and mountain regions. The Hunger Indexes (GHI) of 2008 in these parts of the coun- try are close to, or above, 30, indicating an extremely alarming situation (Nepal, National Planning Commission– WFP–NDRI 2010). In fact, not a single sub-region in Nepal falls within the moderate to low hunger categories (WFP 2009). The situation had improved a little by 2011 as the GHI fell slightly to 19.9 (using data from 2004 to 2009). This gave Nepal a rank of 54th out of 129 countries and the hunger category was reduced to serious rather than alarming (IFPRI 2011). Nepal’s recurrent natural disasters, such as the Koshi floods of 2008, and floods in central and far western Terai in 2009 and 2010, have had an immense impact on crop production, livestock, and other productive and household as- sets. The disruption of market supply chains for long periods due to political unrest, bandhs (virtual closures of transport and business) and more localized strikes can also cause transitory food insecurity. Various risks jeopardize food and nutritional security and worsen vulnerability to food insecurity, including unfavorable geographic location, as well as social, economic, and cultural exclusion and inequalities (Nepal, National Planning Commission–WFP– NDRI 2010). Whatever the case, the people of Nepal are known for their resilience. To cope with seasonal food insecurities, many people—mostly men—migrate to obtain temporary employment, leaving families behind with little to eat (Ne- pal, National Planning Commission–WFP–NDRI 2010). In 2005, the government formed the Nepal Food Security Monitoring System (NeKSAP) 4 to monitor, collect, consoli- date, and analyze food security data. It uses the Food Security Phase Classification approach to distinguish five phases of food insecurity to aid decision making (Nepal, National Planning Commission–WFP–NDRI 2010). In this way, the NeKSAP helps achieve coordinated, appropriate, and timely action to prevent human suffering due to food insecurity. Background to the study and its purpose Farmers’ access to technology, technical knowledge, and institutional credit is severely restricted, particularly in the poorest and most food-insecure areas.One measure for increasing agricultural productivity would be greater applica- tion of modern technologies and inputs. Hence, this study analyzes the use of seeds, fertilizers, pesticides, and agri- cultural machinery with specific reference to three major cereal crops (rice, maize, and wheat). It seeks to under- stand how policies might be changed to promote the supply of these inputs.This study specifically addresses the fol- lowing questions:  What is the current structure and impact of the public agricultural research and extension systems in Nepal, and how do these public systems contribute to improving the ability of small-scale, resource-poor farmers to access new technologies?  What is the current structure of the seed and agricultural inputs industry in Nepal, particularly with respect to competition and innovation?  How does industry structure affect the ability of small-scale, resource-poor farmers to access new technolo- gies? 4Nepal Khadya Suraksha Anugaman Prarali (NeKSAP), meaning Nepal Food Security Monitoring System, is institutionalized within the Ministry of Agriculture and Cooperatives (MOAC) at the district level. It falls under the strategic guidance of the Food Security Monitoring Task Force, which was established by the National Planning Commission. The NeKSAP is comprised of representatives from district-based organizations, including govern- ment, nongovernmental organizations (NGOs), and civil society. 7 SUMMARY | APRIL 2010 Agriculture, seed, and innovation in Nepal: Industry and policy issues for the future  What policy options can be introduced to improve the ability of small-scale, resource-poor farmers to access, and benefit from, new technologies? 8 SUMMARY | APRIL 2010 Agriculture, seed, and innovation in Nepal: Industry and policy issues for the future METHODS, SOURCES OF DATA, AND LIMITATIONS The study was undertaken to understand the situation of agricultural input supply and associated policy regimes in the country using a combination of methods. Sources of data used in the study are given in Table 2.1. TABLE 2.1—A SUMMARY OF DATA SOURCES USED IN THE STUDY Sources: As indicated. Note: n/a = not applicable. The study mainly used three methods to collect and analyze the information:  Review and analysis of secondary data and information, including (i) various policy documents of the Nepal government related to agriculture and (ii) reports and journal articles on related work in Nepal and interna- tionally  Field survey of vendors of agriculture inputs (Agrovets) and private seed companies using a checklist  Case study of selected chronically food-deficit and food-insecure districts in the hills and mountains of Nepal The authors reviewed and analyzed various policy documents related to agriculture research, development, and ag- ricultural inputs in Nepal. This provided the basis for understanding overall agricultural policy regimes, their gaps, and deficiencies. They consulted with key people from the Department of Agriculture (DoA) and the Nepal Agricul- tural Research Council (NARC) within the Ministry of Agriculture and Cooperatives (MOAC) in Kathmandu Valley in Method of study Organizations as source of data and information and other past reports in the study Gov- ernment agencies Agrovets/ Cooperatives Private companies/ NGOs UN agencies and other international organizations Projects 1. Review and analysis of policy documents Number of policy documents reviewed Number of periodical report and other documents reviewed Number of interna- tional documents and journal articles reviewed 10 12 0 n/a 0 0 n/a 2 4 3 2 12 n/a 3 0 2. Key informant consultation Number of key informants consulted 22 10 6 0 3 3. Field survey Number of vendors studied n/a 49 11 0 0 4. Case studies Number of govern- ment line-agencies and vendors con- sulted 6 22 0 0 0 9 SUMMARY | APRIL 2010 Agriculture, seed, and innovation in Nepal: Industry and policy issues for the future October and November 2010, as well as with a number of district-level practitioners in selected districts during No- vember 2010. Farmers and private-sector organizations were consulted during the same period to get their perspec- tives on agricultural inputs and existing policy instruments, and to help set the overall framework for the study. Field Surveys Government departments and parastatals report only official sources of data. This creates a huge gap between offi- cial statistics and reality on the ground. Most of the agricultural inputs are dealt with by private-sector organizations, but none of these transactions are included in official government reporting. Not only are there gaps in data record- ing and reporting, but the reliability of data from government sources is often questionable due to a lack of scientific data-recording systems. The general tendency is to show an improved physical progress in terms of quantity and quality over previous reports (with the exception of a major census surveys). To address these gaps in the available data sets, field surveys were undertaken. The surveys provide not only miss- ing data in the government system, but also data not normally collected by government agencies, such as the infor- mal flow of fertilizers from India. The survey identifies: the trend in the quantity of fertilizers coming informally from India; the actors involved in this trade; the modes of transportation used; and other related issues, including the pos- sible involvement of government administrators, politicians and police from both countries in the trade. The field surveys were done in 10 out of 20 randomly selected Terai districts since these districts are the formal and informal gate ways for any fertilizer and other agricultural inputs from India. It was also important to document the major trends in the non-regulated flow of fertilizers as most of it comes through the open border with India all along Terai districts. The vendors (Agrovets) who were surveyed from November 2010 to January 2011 were selected on the basis of whether they were involved in fertilizer trade. The same vendors were also interviewed about the supply of cereal seeds and pesticides. All seed companies in the 10 randomly surveyed districts in the Terai were studied, as only 16 private seed compa- nies were registered in the country at the time of study. The survey covered 11 out of the 16 private seed companies because most happened to be in the survey districts. Information on agricultural machinery was collected mainly through secondary sources. However, in-depth inter- views with scientists from the Agricultural Implements Research Centre (AIRC), NARC, and Ranighat Parsa district provided a great deal of valuable information. Six hill and mountain districts were chosen for case studies (Figure 2.1). Of these, Darchula and Jumla represent mountain districts, while the remaining four are mid-hill districts. Darchula, Achham, Jumla, and Kalikot are among the least developed, remote, and vulnerable to food insecurity and also affected greatly by conflict. Sindhuli is also low on development indexes but better off in terms of food security; Palpa has quite high development indexes, in- cluding accessibility. Information collected from a sample of vendors—mainly Agrovets and fair-price cooperatives— formed the main source of information for the case studies in Palpa (six) and Sindhuli (four). Information was gath- ered from four vendors in Darchula and Achham and two vendors in Jumla and Kalikot. The field surveys were done by staff of SUPPORT Foundation, one of CARIAD’s collaborators. However, staff members of LI-BIRD and FORWARD (also CARIAD's collaborators) were also involved in a few instances. Surveys were arranged through partners mainly because these organizations had good rapport in the area and were fully aware of the situation related to agricultural input supply. The checklist and questionnaires were prepared for all the field surveys and shared with the International Food Policy Research Institute (IFPRI). All the enumerators were ori- ented on methodology and the checklist to be used in the survey. Data on the seasonal variation in the transaction volume and price of fertilizers were also collected. After the draft report was prepared, certain sections of the document, where independent review was deemed im- portant, were circulated to a selected professional for critical review. The section on agricultural machinery was re- viewed by Dr. Stephen Biggs (ex Research Fellow East Anglia University, United Kingdom), and pesticides by Dr. Yubak Dhoj G. C. (Programme Director, Plant Protection Directorate, DoA) and Sharala Sharma (Senior Scientist, 10 SUMMARY | APRIL 2010 Agriculture, seed, and innovation in Nepal: Industry and policy issues for the future Plant Pathology Division, NARC). The draft reports on Agrovets and private seed companies were reviewed by Laxmi Kant Dhakal (a successful seed entrepreneur and vice-chairperson, Seed Entrepreneurs' Association of Ne- pal). All the comments provided by the reviewers were considered while revising the report. Limitation of the study The study heavily draws on secondary data and information mainly from government sources. Consequently, it only presents official transactions by government agencies and parastatals, alongside government positions, rather than realistic situations on the ground. In many instances, government departments, NGOs, and projects work with the same set of community-based organizations, while their physical progress on the same sector is reported separate- ly. In this context, there is a danger of double accounting, particularly on statistics related to cereal seeds. The quan- tity of seed transactions reported for small seed enterprizes (SSEs), community-based seed producer (CBSP) groups, and District Seed Self-Sufficiency Programmes (DISSPROs) is one example. Information related to the informal flow of fertilizers from India also needs to be interpreted cautiously; it presents more of a trend than accurate data. This is understandable since it is an informal and illegal operation. Many times respondents hesitated to share information about this trade. Once the study’s objectives were shared with them, they were more comfortable responding to queries but often provided a range or just rough estimates. FIGURE 2.1—DISTRICTS OF NEPAL WHERE SURVEYS WERE COMPLETED ON AGRICULTURAL INPUTS Banke Rupandehi Parsa Chitwan Bara Dhanusa Kailali Dang JhapaMorangSurvey districts Jumla Darchua Palpa Kalikot Achham Case study districts Sindhuli Source: Produced by authors. The wealth of information collected in this report will be useful as raw material for government planning, particularly for developing (1) a new Agricultural Development Strategy (ADS), (2) a national food and nutritional security policy (NFNSP), and (3) a new fertilizer policy. 11 SUMMARY | APRIL 2010 Agriculture, seed, and innovation in Nepal: Industry and policy issues for the future ANALYSIS In Nepal, farmers’ access to agricultural inputs and modern technologies and technical knowhow is con- strained. An analysis of these constraints can provide evidence for practical policy recommendations to the Nepal government, which is currently developing a new Agricultural Development Strategy (ADS) to replace the existing Agriculture Perspective Plan (APP). The analysis was done with reference to three major cereal crops (rice, maize, and wheat) in four subsectors (crop-breeding research and seed supply, fertilizers, pesticides, and agricultural machinery). It was designed to gain information on the research, development, and innovation system and on the role of policy and organizations in providing agricultural inputs and innovations. This analysis addresses four major issues:  The research and extension system and how it affects resource-poor farmers  The structure of the seed and agricultural inputs industry  How these industries allow small-scale, resource-poor farmers to gain access to new agricultural technolo- gies  The available policy options that could improve access of resource-poor farmers to new agricultural technol- ogies The study synthesizes information and data from several sources:  Statistics published by government agencies  Interviews with key actors in the research, extension and agro-industry sectors conducted in October– November 2010  A field survey from November 2010 to January 2011 of 49 agricultural input vendors who sold fertilizer as part of their product range in 10 randomly selected Terai districts and all 11 registered private seed compa- nies in these districts  Six case studies in hilly and mountainous districts with four selected to represent chronically food-deficit dis- tricts and two selected to represent food-insecure districts 12 The public agricultural research and extension systems Nepal’s agriculture sector contributes to more than one-third of gross domestic product (GDP) but, over recent decades, agriculture development has failed to keep up with the increased demand from a growing population. In particular, the average annual growth in the production of the major cereals (rice, maize, and wheat), which are the staple foods in Nepal, is much lower than the population growth rate. Because of this lack of growth in agricultural productivity, Nepal already faces food shortages: 41 percent of the population consumed less than the minimum calorie requirements in 2010 (Nepal, Ministry of Finance 2010); 43 out of 75 districts are in food-deficit;and 10 are vulnerable to food shortages (NARC 2010). Rural poverty, food security, and agricultural productivity are closely linked: 78 percent of the population has an income of less than US$2 a day. Of these, 67 percent are smallholder farmers, while 11 percent are agricultural laborers. The rural poverty level of about 33 percent of people below the poverty line is 3 times as high as the urban poverty level (Gulati et al. 2010). Agriculture research is an engine for agricultural growth if it generates and disseminates improved agricultural technologies that farmers can accept and that increase food production. Only if such technologies are widely used will they help increase production and farm incomes at a national level, and hence lower food prices for the food-insecure. Agricultural research will only achieve this impact if it is demand-driven (so that its research products meet the needs of farmers), and results-oriented (so that the new technologies are put into use). The Nepal Agricultural Research Council (NARC) is the main agency responsible for agricultural research. In spite of a high degree of autonomy provided by its bylaws, political interventions greatly reduce its effectiveness. Private-sector companies and NGOs are also involved in agricultural research and extension, but on a limited scale. NARC’s breeding programs are heavily dependent on introducing new crop varieties from the Consultative Group on International Agricultural Research (CGIAR) centers and other international crop research programs; the national research system has not developed sufficient research capacity of its own. The Department of Agriculture is present throughout the country through its District Agriculture Development Offices (DADOs) and is the key government agency for agricultural extension. The extension system has changed from the “training and visits” system to its current supposedly “service on demand” approach, in which farmers are expected to visit the Agriculture Service Centers (ASCs) or Agriculture Sub-Service Centers (ASSCs) where the government agricultural technicians are stationed. However, on average, rural households require more than two hours to reach their nearest markets or government service centers because of difficult terrain and a lack of transport infrastruc- ture (Nepal, National Planning Commission 2004). Despite such difficulties, the case study in the hilly and mountain areas showed that DADOs were almost the only source of new seed varieties and technical knowhow. However, the stakeholders' comments revealed constraints other than travel:  The unavailability of the technician when they visited  The poor quality of inputs and services rendered by DADOs and their lack of timeliness  Poor commitment of staff members to their work, management’s failure to impose accounta- bility, and poor monitoring of the work of the DADOs There is a strong case for reviewing the current “service on demand” approach, particularly for remote and difficult areas because private-sector service providers provide a weak or almost non-existent alternative. The involvement of NGOs in supporting agricultural development is common, mainly through exten- sion and subsidized input supply in projects funded by donors. While few are also involved in agricul- tural research, they have made tangible contributions in participatory research and the scaling-up of its outputs, and hence have generated impacts on food security and productivity. However, this role in agricultural research and development may not be sustainable as most NGOs lack a continuing 13 source of funding. Sustainability can be provided only by a change in policy that allows the funding of NGOs through long-term, government-funded agricultural programs. The seed and agriculture inputs industry The Agriculture Perspective Plan (APP) was formulated in 1995, with its implementation starting in 1997, as a framework for accelerating broad-based agricultural growth to achieve food security and poverty alleviation. Most of the APP targets have not been met for a variety of reasons, including lack of coordination and resources. Its implementation focused on a “pocket package” approach where activities were concentrated in only a few pockets, mostly in high-potential production systems in the more accessible areas—that is, the Terai, low hills and valleys. This ignored the vast majority of small-scale, resource-poor farmers (smallholders) in the country. However, the government has recently begun to replace the APP with a new 20-year strategy—the Agricultural Development Strategy (ADS) (B. Aryal, pers. comm.). The new strategy is expected to be in place by 2013, nearly two years before the planned completion of the APP. The more recent National Agricultural Plan (NAP), 2005, and the Three-Year Interim Plan (TYIP), 2007–2008, were both politically motivated policy documents. Their main features were broad-based economic growth by modernization and commercialization of agriculture, while addressing the issues of social inclusion by incorporating concerns of ethnic minorities and women. Again, none of the plans have been very effective for smallholder farmers as the programs were focused on high-potential production systems and accessible areas. Research Within all these plans, the Nepal Agricultural Research Council (NARC) is the lead agency for agricultural research in Nepal and mainly works with open-pollinated and inbred varieties of all the food crops. It has released 140 varieties of rice, maize, and wheat since the 1960s. Only 60 percent of currently released varieties are in demand for rice, and only one-third for maize and wheat. Farm- ers are overwhelmingly growing old and obsolete varieties of these crops: 85 percent of all the foundation seeds demanded for rice in 2010 were for varieties released before 1995; for maize the share was 75 percent; and for wheat, around 36 percent. A few private-sector maize hybrids (from Indian and Chinese companies) have been formally registered with the regulatory authority, although there has been a well-established informal trade in hybrid maize seeds. NARC has released only one hybrid maize variety through the National Seed Board (NSB). Private companies have also registered some rice hybrids with the NSB (L. Acharya, pers. comm.). No genetically modified (GM) cereal varieties have been tested or released in Nepal. Seed industry Until 1998, the formal seed industry, comprising the development of new varieties and the supply of source seeds, was almost entirely in the hands of the NARC. The National Seed Policy (NSP) of 1999 for the first time had a provision for private-sector involvement in crop-variety development and the seed trade. The policy opened up the possibility of research on transgenics—genetically modified organisms (GMOs)—following the establishment of biosafety rules and regulations. The Plant Variety and Farmers Rights Protection Bill (currently in Parliament) has a provision, for the first time in Nepal, for plant breeders’ rights, and these will be compatible with international norms. The NSP also allows the use of data generated from participatory on-farm trials for the release or registration of crop varieties. This was in response to initiatives on participatory crop improvement and community-based seed production by nongovernmental organizations (NGOs) with international collaborators. Two types of seed supply systems are prevalent in Nepal: the informal farmers’ seed system, which supplies more than 90 percent of all the seeds used by framers in crop production (see Table 2.1), and the formal system, which supplies less than 10 percent. Official statistics for cereal seeds in Nepal reported only 4,000 tons per year (Gulati et al. 2010). This, however, is the amount for the National Seed Company Limited (NSCL) and does not take into account seed transactions by companies, cooperatives, and groups, some of which belong to the District Seed Self-Sufficiency Programme (DISSPRO), that are registered with their local DADO (Table 2.1). The performance of 14 the formal, largely public-sector, system has been poor, but recently private seed companies and small seed enterprises (SSEs) have been emerging as important actors for food-crop seeds. TABLE 3.1.SEED SUPPLY BY VARIOUS ACTORS OF CEREAL SEED INDUSTRY IN NEPAL, 2010 Sources: Data for area, total seed requirement, and supply by NSCL are from national statistics (NMoF, 2010); data for supply by private seed companies, cooperatives, groups, and farmers themselves are from surveys conducted for this study. Note: Throughout the document, “t” refers to tonne. Some recently established seed companies also have research components. Private companies from the neighboring countries of India and China have already registered hybrid maize and rice varieties with the National Seed Board (NSB) Nepal to market seeds in Nepal. This indicates that the environ- ment for competition and creation of crop innovations is becoming more favorable. However, it remains to be seen how well these new varieties cater to the needs of smallholders. Cereal seeds for open-pollinated varieties (OPVs) of maize and inbred varieties of rice and wheat are not commonly imported from India. DADOs make the most officialdemands (called indents) for foundation seed across the three major cereals, although the private sector is the most important for rice; DADOs are the sole actor in the mountainous region, and the most important one in the hilly region. Private-sector organizations are concentrated in the Terai where business opportunities are better, while DADOs provide source seed in the difficult and remote areas. Private organizations from just a few districts placed most of the indents for foundation seed for the Terai. These districts were Dang, Rupandehi, Chitwan, Rautahat, and Sunsari, demonstrating their emerging importance as centers for seed production. Fertilizer supply Until 1997, the supply of subsidized fertilizer was entirely a public- sector responsibility. In that year, the Nepal government decided to deregulate the fertilizer trade and open it to private-sector participa- tion. The Fertilizer Control Order 1999 and National Fertilizer Policy 2002 provided a regulatory framework for the liberalization of this essential commodity. For a variety of reasons, deregulation did not have the desired effect in improving fertilizer supply and hence the policy had no impact on agricultural production and productivity. Not only is there an acute shortage of fertilizers in the country, the deficiency of major nutrients, such as phosphorus, is becoming evident in areas with higher cropping intensities. This is mainly because farmers do not use balanced dosages of plant nutrients. After withdrawing its fertilizer subsidy from 1997 to 2008, the government reverted to subsidizing limited quantities of fertilizers in 2009; once again, it made the National Fertilizer Company Limited (NFCL) the sole importer and distributer of such fertilizers. There is a big mismatch between the estimated current fertilizer demand of 586,000 tons annually and supply by NFCL of 100,000 tons in 2009–2010 and around 180,000 tons in 2010–2011. Nepal is in a dilemma in terms of fertilizer supply. Policy analysis in the past (ANZDEC 2002; IDL Group 2005; Thapa 2006) indicated that Nepal would need US$41 million annually to subsidize fertilizer purchases, which is a substantial amount. None- theless, the Nepal government provides a huge subsidy (US$170 million) on petroleum products (diesel, petrol), sometimes even by drawing on the development budget, but this subsidy is largely enjoyed by the urban elite (D.B. Chhetri pers. com 2011; Sharma 2011; Ghimire 2011). Similarly, the Nepal government provides US$25 million to World Food Programme (WFP) Nepal to support 1.8 Indicator Seed supply Rice Maize Wheat Area (million ha) 1.55 0.82 0.73 Total seed requirement (t) 93,000 20,000 87,600 Supply by National Seed Company Ltd. (%) 0 0 4 Supply by 11 private seed companies (%) 2 2 2 Supply by cooperatives and groups (%) 3 7 2 Farmers' own seed (%) 96 92 92 15 million people across the country through a range of programs focusing on preventing hunger and meeting food and nutrition needs (WFP 2011). The existence of these large subsidies makes it more difficult for the government to find resources for a fertilizer subsidy (or for other inputs such as seed), even though this would be more equitable (as it would improve agricultural productivity and reduce external food supply and stabilize food prices). However, one can argue that $4.1 million funds can be found by revisiting overall subsidy regime of the government and by restructuring and privatizing the import of petroleum products which is the major factor affecting development budget. Official statistics based on aggregate/national-level statistics put fertilizer use as low as 19 kg per hectare per year (Gulati et al. 2010). Fertilizer-use level in Nepal falls far below than that of other South Asian countries, which range from 136 kg per hectare (India) to 295 kg per hectare (Sri Lanka), (Gulati et al. 2010). Comparing these figures with neighboring states of India would be a more relevant comparison, but reliable data are not available. The Ministry of Agriculture and Cooperatives (MoAC) has no system of regularly collecting farm-level data on the use of fertilizers. Citing 2001 data from the Central Bureau of Statistics (CBS) in a study commissioned by Nepal’s Ministry of Finance, Thapa (2006) reported the average use of fertilizers in Nepal in 2004–2005 to be 125 kg per hectare. Nepal’s Fertilizer-Use Baseline study reported that the average nitrogen–phosphorus–potassium (NPK) nutrient application-rate on cereals in 2001–02 was 63 kg per hectare (Agrifood Consulting International 2003). In 2009, a large, randomized survey of more than 1,600 households in 18 districts of the Terai found entirely different patterns of fertilizer use in Nepal. 5 For example, the average use of fertilizers in rice was 206 kg per hectare, in maize 204 kg per hectare, and wheat 194 kg per hectare (the UK DFID Research into Use Programme). In addition, the study also found that sampled households were using on average 11 tons per hectare of farmyard manure on all three cereals. These fertilizer figures seem more than reasonable. To support a rice yield of 3 tons per hectare, 100 kg nitrogen is required. To support the national average rice yield of 2.7 t per hectare, 90 kg nitrogen per hectare is needed. These farm-level data show much higher fertilizer use than other studies that relied on secondary statistics from government sources. The higher use is probably because it takes into account both officially obtained fertilizer and also that informally imported from India. In the survey of agricultural input vendors, 60 percent of fertilizer sold had been imported unofficially. The proportion of unofficial imports is even higher when fertilizers are directly imported by farmers rather than passing through Agrovets 6 , and when other private fertilizer dealers are taken into account. Those involved in this unofficial trade were private fertilizer dealers (bulk suppliers), Agrovets, and individuals (including politicians) bringing fertilizers into the country using a variety of means of transportation, but the bicycle was reported to be the most important. Trucks, motorcycles, and four- and two-wheeled tractors were also used. Spurious fertilizers are a common occurrence when imported informally from India as it is impossible for purchasers to complain about a product that is imported illegally. Case studies from six selected districts (four chronically food-deficit and two food-insecure) in the hilly and mountain regions, where more than half of the population are poor (Nepal, National Planning Commission–WFP–NDRI 2010), showed the supply of fertilizer (as well as improved seed and pesticides) by the private sector, AICL, and NFCL to smallholders was woefully inadequate. Fertilizers were not available when needed and never in quantities sufficient to have a visible impact on overall agricultural production. Even delivery by the World Bank–funded Nepal Safety Net project was unable 5 A baseline survey covering more than 1,600 randomly sampled households from 180 wards of 90 village development committees (VDCs) of 18 Terai districts was done in 2009. For the survey, VDCs were randomly drawn to represent three agroecological categories based on drainage patterns, moisture regime, soil fertility, and crop productivity patterns from north to south, representing uplands, medium lands, and low lands. In addition, at least one group discussion was conducted at each participating VDC to triangulate results from household surveys (LI-BIRD-FORWARD-SUPPORT Founda- tion-CARIAD 2011). 6 Agrovets refer to private entrepreneurs in Nepal that deal with agricultural inputs and veterinary medicines http://www.wfp.org/countries/Nepal/Operations 16 to meet modest targets. For example, of the 74-ton fertilizer-supply target for Manang district, only 13 tons were delivered, and there was a complete failure to deliver the targeted 8 tons of improved crop seed (Tiwari, 2010). Pesticide supply All told, 97 pesticide suppliers have been registered with the Plant Protection Directorate (PPD). Pesticides are distributed through a large network of Agrovets —around 2,000 (including non- registered ones)—located in the Terai and more accessible areas in the hills (Nepal, Ministry of Agriculture and Cooperatives 2009). There is high competition in the sale of pesticides. The use of pesticides is increasing in the Terai and more accessible areas, while smallholders had a limited access to these inputs in the hills and mountains. There is also a lack of proper mechanisms in place for the safe handling of hazardous chemicals. The survey of vendors found 60 percent of people engaged in the pesticide business had no technical background to handle pesticides, but many of them were giving advice to farmers. Although, study by Nepal Economic Agriculture, and Trade (NEAT-a United States Agency for International Development (USAID) funded project report that people running Agrovets and agro-dealers are educated (NEAT, 2011) (which is true for town centers and market places), however, majority of those operating in the rural areas are neither run by educated people nor are registered. Discussion with the Agrovets and other vendors revealed a lack of regular coordination with con- cerned lineagencies and vendors. This related particularly to keeping vendors up to date on govern- ment policy for import and trading of pesticides and on newer and safer pesticides. Private-sector agencies, particularly Agrovets, are the major players in the trade of pesticides, micronutrients, other agrochemicals, and in advising on their use. However, the majority of vendors do not have an agricultural science background so the content and quality of information passed on to clients with regards to pesticides is probably poor. Areas of great concern include a lack of knowledge about the quantity of pesticide used per unit of water, a safe waiting period, and safety measures during the storage, application of pesticides. The source of information used by vendors to pass on to clients was reported as largely coming from their own experience—which could not possibly cover the range of products actually supplied—rather than from credible sources such as DADOs or manufac- turers. This may have far-reaching consequences in terms of correct use of pesticides, their effective- ness, and impact on the environment. There is, hence, strong evidence of the need for capacity building of people on safe handling of pesticides and other agrochemicals particularly for those who are in the trade but do not have relevant qualifications. Reviewing the criteria for issuing new licenses to trade in pesticides may also be important. To address this, PPD is making a 35 days intensive training mandatory for those who are interested to start business on pesticides. Training covers the entire aspect of import, safe storage and application of pesticides before a trading license can be issued (Yabak Dhoj GC, pers. comm.). Similarly, coordination between the concerned government line agencies and the vendors while importing the pesticides is important to ensure that safer pesticides are brought into the market. There were also cases where the pesticide content in the consignments did not conform to the weight stated on the label. Both vendors and clients were fully convinced about the need for strong quality monitor- ing mechanism by the responsible government line agencies. There is a clear lack of a recording system for these potentially hazardous chemicals in terms of storage, sale, and use. It was beyond the scope of this study to compile complete records on the pesticides in Nepal. It is possible to trace legally imported pesticides through records in customs offices, but this will not provide the complete inventory. Although the PPD has recently started documenting the various pesticides imported in the country, it is far from being complete (Nepal, Ministry of Agriculture and Cooperatives 2005). In the absence of such records, not much can be said about whether demand matches supply, and whether farmers in Nepal are benefiting from recent advances in this area of science. 17 During the discussion with the PPD, it was learned that Nepal does not have a pesticide accreditation laboratory and the required human resources to analyze pesticide residues. The PPD saw this as one of the constraints for the national programs to comply with international treaties and conventions signed by the Nepal government. This has also adversely affected the export of high-value crops and other agricultural commodities. Rural and agricultural mechanization There has been considerable progress in agricultural mechanization in Nepal with various types of machinery being adopted, primarily through imports by the private sector and its engagement with farmers. Most agricultural mechanization is taking place in the Terai and, to a lesser extent, in the lower hills and valleys. There are suppliers of all major makes of tractors, power tillers, and pump sets in Nepal. The machinery-supply industry is becoming more competitive as the number of actors increases along with the demand for machinery. However, there is a lack of policy to support agricul- tural mechanization. On the contrary, some government policies have had a negative effect on the development of local manufacturing, such as a high duty on importing raw materials needed for local manufacture of agricultural tools and machines. Agricultural machinery is having a positive impact on smallholders since small-scale threshers, pump sets, and tillage equipment are now more widely available, less expensive, and suitable for small- holders. Smallholders can also avoid capital investment as, increasingly, Nepalese machinery owners provide custom hiring machines particularly tillage equipments, threshers and combine harvesters. Custom hiring by Indian business people in declining more recently, although, there is a legal provi- sion to pay a daily customs duty during the temporary import of machines. In the hills and mountain districts, mechanization is low given the difficulties of transporting heavy machinery and using it on small terraces. The predominant service providers for agricultural hand tools are DADOs, which procure the tools on demand for farmers. The private sector is almost non- existent in the mountain districts and thin in the hilly districts. Increasingly, Nepalese agriculture is facing a shortage of labor as a result of outmigration of youths in search of employment opportunities. This is also partly due to the lack of attention in reducing drudgery in agricultural and rural operations for poorer farmers and laborers. This labor shortage has had a distinct influence on the mechanization of agriculture in Nepal, which has accelerated in the last two decades. Several power-operated agricultural machines are now in use in Nepal (Biggs and Justice 2011), including the following: water pumps; tractors both 4 wheel and 2 wheel; harrows; rotavators; seed drills; threshers; combine harvesters; agricultural processing machines; rice, oil, and pulse mills; and laser land-levelers. Some agricultural mechanization had begun in Nepal in the 1960s and 1970s with the introduction and promotion of four-wheel tractors (4WTs) on the Terai (Pudasaini 1976). This could be attributed to government- and donor-supported policies, as well as to the interest and engagement of the private sector in bringing in machines for agricultural use. Mechanization technologies and machines availa- ble across the long open border with India greatly influenced mechanization patterns in the Nepal Terai (Biggs et al. 2011). Two-wheel tractors (2WTs) were promoted during the mid-1970s and early 1980s, with two Japanese aid programs importing approximately 2,000 tractors. Initially, the spread of 2WTs was limited to the Kathmandu and Pokhara Valleys, where they were used for transport and tillage. Further registration of two-wheelers was banned in both valleys as the vehicles were contributing to growing traffic congestion. One can still see some of the 2WTs in use even 30 years after their first introduction in Nepal (Biggs et al. 2011). Minimum tillage by power-tiller drills has been popular among small- to medium-farmers as the drills perform three operations simultaneously: soil tilling, seed sowing, and planking. It saves on cost and overcomes the problem of poor plant-stand that can result from poor tilth and manual broadcasting. Minimum tillage by participatory technology development has produced an extra 600 kilograms per 18 hectare mean grain-yield of wheat compared with typical yields based on farmers’ practice (Manandhar et al. 2009; Pariyar et al. 2001). From the 1970s onwards, all the major Indian tractor companies established sales agencies across Nepal. The latest are multinational companies such as John Deere and New Holland, which have manufacturing bases in India. Current estimates of the total number of tractors are 42,000, with 30,000 (71 percent) four-wheelers and 12,000 (29 percent) two-wheelers. Many business people and engineers feel the demand will continue to grow as high as 40,000– 50,000. Despite the ban still in place in the Kathmandu and Pokhara Valleys on 2WTs, the growth in these peri-urban areas contin- ues: farmers and other rural entrepreneurs who own the vehicles evade bans by registering their tractors in other places (Justice and Biggs 2011). Trailers are an important feature for improving the income-earning potential of tractors. Tractors are used for three–six months a year in agricultural work, but in rural and peri-urban areas, tractor-trailers generate much work for hauling and transport. Justice and Biggs (2011) indicate a lack of statistics on the number of trailers, but estimate nearly 80 percent of owners of 2WTs in Nepal also own a locally manufactured trailer (9,600 tractor-trailers). From the mid1990s on, large Indian combine harvesters were seen in Nepal, and today many Nepali combine owners and also Indians work on a contract basis in various districts in the Terai, majority of which are concentrated in Rupandehi, Bara, Parsa, Rautahat, Kapilvastu and Kailali districts. Accord- ing to the NARC Scientist of AIRC Ranighat, Birganj, the machines are popular because they halve the costs of harvesting and threshing (G. Sah, pers. comm.). Small field sizes and maintenance problems are the major constraints to increased use of combine harvesters. Both wheat and rice threshers are becoming popular in most parts of the Terai because farmers see a clear advantage in terms of saving time, resources, and labor. The use of machines to thresh wheat is much higher than for rice, although rice threshers are beginning to spread in Nepal’s eastern Terai. These are generally powered by the diesel engines of pump sets of Indian and, increasingly, Chinese origin. At present, only 17 percent of the total cultivated area has year-round irrigation, while 42 percent has irrigation of some sort. In the Terai, out of 889,000 hectares of irrigated area, 82 percent is covered by surface irrigation and the rest by ground water (tube wells) (Nepal, Ministry of Energy 2002). The agricultural census reveals that 14 percent of holdings in the Terai use pump sets for pumping water mainly from shallow tube-wells. Indeed, water-lifting machines have one of the fastest growing adoption rates (Fig. 3.1). Farmers now have access to small-scale irrigation technologies such as Chinese pump sets, lay-flat pipes and the custom hiring of irrigation services. These are helping to alleviate the effects of more unpredictable and erratic rains, and more frequent long dry spells. In general, five-horse-power pump sets are the most common, and farmers prefer water pumps made in India for their durability, availability of spare parts, and convenience of repair and maintenance services. Chinese water pumps are also becoming popular as they are cheaper to purchase or custom hire. Another contributing factor to the adoption of pump sets may be subsidies for them, and subsidies for the use of electricity. There are now at least 100,000 small-scale irrigation pump sets in use. In addition, more than 100,000 treadle pumps have been installed by small farmers for irrigating mainly vegetable crops in the Terai. Pump sets are the most popular agricultural machinery in Nepal followed by threshers and winnowing fan. Use of four-wheel tractors is as long established as pump sets, but their growth has been rather slow, probably due to high demand for capital. Around 15–20 percent of tillage is now mechanized, largely by four-wheel tractors. Two-wheelers have been around since 1990, but have been increasing considerably in number only over the last decade (Fig. 3.1). Another striking change is the growth of processing mills in the last decade. In addition, the rapidly growing horticulture, poultry, dairy, animal- feed and other “value-added chain” industries, based on agricultural and other rural resources are using mechanical equipment of one type or another in production, processing, transport, and market- ing activities. 19 FIGURE 3.1—SPREAD OF AGRICULTURAL MACHINERY, NEPAL, 1970–2010 1970 1980 1990 2000 2010 Year 0 20 40 60 80 100 120 140 N u m b er i n t h o u sa n d Pumpset 4 W heel Tract ors Threshers 2 W heel Tract ors Mills Hand Cranked W innowing Fan Source: Adapted with permission from Justice and Biggs, from a forth coming book on agricultural mechanization to be published by FAO. Using drip irrigation technology, fields can be irrigated with less water, making it more effective in areas having water scarcity. It has been very successful in hilly regions and the northern belt of the Terai region. To date, more than 189,000 biogas plants have been installed with more than 95 percent in operation. Favorable policies for subsidies, public-private partnership, comprehensive quality standards, and a strong monitoring mechanism are the major pillars of this program. More than 66 percent of biogas plants are connected to toilets. Biogas is widely recognized as a sustainable and clean energy source to improvethe quality of life and protects the environment (Manandhar et al. 2009). An improved water-mill program is being implemented in 16 hill and mountain districts of Nepal through 16 service centers and eight Ghatta Owners’ Associations.7By the end of June 2007, the program had helped install more than 2,700 improved water mills. Of these, the vast majority is of short shaft and used for efficient grinding; about 300 units are of long shaft and used for other purposes such as rice mills, saw-mills, oil expellers, lokta (Daphe spp.) 8 beaters, and chiura (beaten rice) making. The improved water mills have replaced diesel-powered mills and, hence, contribute directly to reduced carbon dioxide emissions, saving on scarce foreign exchange to import expensive fossil fuels and reducing dependency on them. Gravity ropeway technology was transferred from northern India to Nepal in collaboration with the International Centre for Integrated Mountain Development (ICIMOD) and private manufactur- ers/suppliers. An initial study showed that the gravity ropeway system reduced transportation costs of agro products by at least half. It gave villagers the confidence to supply vegetables, milk, and other perishable agricultural and forest products in larger amounts and to enter the competitive market in cities. It also improved socioeconomic conditions (health, education, and convenience), created employment opportunities, and supported the business of local manufacturers and service providers (Biggs and Justice 2011). Some agricultural machinery developed by NARC’s Agricultural Engineering Division (AED)—hand maize sheller, coffee pulper, thresher-cum-peeler for millet, and low-cost solar dryers—are commer- 7 Ghatta is a traditional water mill used for milling cereal grains. 8 Lokta (Daphne spp.) is an endemic plant bark which is used for making Nepali paper. 20 cialized and adopted by farmers in various parts of Nepal. The hand maize sheller reduces drudgery for women farmers. The coffee pulper has been adopted on a commercial scale at the community level, and has increased capacity and efficiency by 50 percent compared to manual operations related to these commodities (NARC 2004). The Agricultural Engineering Unit was established in 1953 under the Ministry of Agriculture to develop infrastructure, farm buildings, farm roads, and irrigation facilities in various newly established agricul- tural farms and stations. Later, its role was redefined to develop appropriate technologies in agricul- tural engineering for various agroecological regions of the country. The Agriculture Implements Research Centre (AIRC), also known as Agricultural Research Station– Agriculture Implements) Ranighat, was established in 1965. It is responsible for designing, testing, modification, and promotion of various types of tools and equipment. The Agricultural Tools Factory (ATF) was established in 1968 and was later privatized due to its poor performance as a public-sector organization. As a private company, it was constituted with 65 percent private shareholding and 35 percent shareholding from the Nepal government. ATF performed only for about six months after privatization and then it stopped functioning for various reasons (Adhikary 2003). Mechanization will become more important in addressing the emerging issue of all the young men who have migrated out of the rural areas. With male outmigration, women may become more respon- sible for making on-farm production decisions. This suggests the need for targeted interventions to address rural women’s needs, for example, special credit programs to help rural women purchase small-scale tractors and pump sets (Manandhar et al. 2009). NARC scientists and staff working in the engineering workshop revealed in discussions that many of the tools and machines developed by NARC were still on the shelves due to lack of a favorable policy for enterprise and marketing. NARC stations do not have a mandate to produce and sell tools and machines on a large scale—their role is limited to research. Local manufacturers are de-motivated, as there is a high customs duty on importing raw materials for manufacturing. Conversely, all kinds of tools and machines that are ready to use are exempt from customs duty. Hence, engineering firms are not interested in commercializing NARC prototypes, and small holder farmers and other stake- holders have not benefited from these interesting and potentially useful innovations. Since the late 1990s, there has been collaborative research between NARC and CGIAR centers (particularly CIMMYT and IRRI) to develop conservation-agriculture (CA) technologies with the National Wheat Research Programme (NWRP), AIRC, and other NARC Centers. More recently, the Cereal Systems Initiative for South Asia (CSISA) project has also begun to validate and promote CA technologies again by providing a few units of various types of machines to groups of farmers in six selected districts of the Terai. CA-based technologies promoted by CSISA are considered to maintain soil health and are profitable as they reduce the cost of cultivation and enhance productivity. They also reduce drudgery, and are sustainable and environmentally sound. As important interventions, CSISA chose the laser land leveler (LLL); direct-seeded rice (DSR); zero tillage on wheat, maize, and lentil crops; un-puddled transplanted rice (UTPR); and reduced tillage (RT) on wheat, lentils, and winter maize after rice. Currently, two units of laser land-leveler have been introduced and are in operation. Apart from LLL, the project also introduced multi-crop seeders for rice, maize, wheat, lentils, kidney beans, and barley. The Turbo Happy Seeder was mentioned to be particularly suitable for plots that use combine harvesters; the seeder is designed to chop straw that is quite long and spread on the ground (D. P. Sherchan, pers.Comm.). However, since many rural engineering innovations are already spreading in use, and have been tested against the harsh economic and technical realities in contrasting areas of Nepal, it would be best to assess if, and where, these CSISA interventions are already in use. 21 At Ranighat, in the Terai, a long-term experiment is assessing the relative advantages of the raised bed over the flat-bed system for rice, wheat, and mungbean. It includes experiment with and without mulch (rice straw for wheat, and wheat straw for rice and mungbean) in terms of net return per unit area, as well as long-term sustainability of cropping patterns. Direct seeding is done for all the three crops. Reshaping of beds is done after rice, as some of the soil from the beds also goes down to the furrows. Preliminary results suggest that raised beds with mulch gave much higher combined yield and returns compared to flat beds with or without mulch. Weed suppression, moisture conservation, and possibly the addition of extra organic matter to the soil through mulch may have contributed to this increased yield and returns compared to no mulching. According to the NARC scientist at AIRC, Ranighat farmers consider raised beds less user-friendly for crop production; according to them, this is a tedious method as it requires planting on the rows remaking the beds, and use of zero or minimum tillage every year, for which they may not have access to necessary tools and machinery. Capacity building of in-country agro-related metal works industries Local manufacture is crucial for the growth and development of agricultural mechanization. Manandhar and Adhikari (1999) estimated that more than 85 percent of hand tools and implements are made by local artisans (blacksmiths and other people engaged in this business). However, such artisans are often poor, illiterate or poorly educated, socially excluded (particularly in the case of blacksmiths), and lack any access to new knowledge and technologies and financial resources. Therefore, modernization of this important trade is yet to be achieved. These constraints could be removed by training artisans in manufacturing skills development and business development skills following the model used in the development of small seed enterprises (SSEs). Other lessons from the SSEs also apply, such as the need for support for working capital—in this case to avoid the usurious interest rates of money lenders—and the development of cooperatives. Lack of technical and safety standards Although NARC is involved in the development and testing of new prototypes of agricultural tools and machines, there is no institution responsible for overseeing quality control, standardization processes, and safety measures. As a result, there have been several accidents while operating agricultural machinery. Enforcing quality and safety standards on agricultural machinery production and import are vital (G. Shah, pers. comm.). Focus on creating awareness about agricultural mechanization Unless farmers and other entrepreneurs know about new tools, equipment, and machinery and their roles to improve livelihoods, it is unlikely they would adopt those technologies. Therefore, knowledge about agricultural machines and associated technologies and their potential contribution in poverty reduction should reach clients in the grassroots. Several studies on seed flow indicate that knowledge about the seeds travels far ahead of the adoption of new crop varieties (Witcombe et al. 2008), and this concept would also fit in the process of other technologies, including agricultural tools and machines. However, since these are more resource-intensive than other agricultural inputs, it would need more awareness creation and flexibility—such as custom hiring for smallholders—to be able to use them. Extension and promotion of agricultural machineries are currently weak in Nepal mainly due to lack of an organizational set-up and the necessary human resources. The Directorate of Agricultural Engi- neering (DoAE) was recently established with limited human resources for the dissemination of technologies. The extension of its presence to the district level in all the DADOs would improve access of farmers to these technologies. Although the testing and promotion of agricultural tools and equipment and associated technologies are taking place at the outreach sites of NARC, these activities are limited. They need to be made more participatory and client-oriented so NARC can learn which ones should be promoted, and so farming communities can learn more easily about new technologies and innovations, and then adopt them. 22 Inadequate facilities for service, repair, and maintenance of farm machinery; lack of availability of spare parts; and lack of training on operation and maintenance of farm machineries are also con- straining growth of this industry. Lack of rural infrastructure, particularly roads, is a major obstacle for the overall development of agriculture, including mechanization and market networks. Interestingly, in Bhutan, there is an integrated program of cheap rural roads for two-wheel tractors (2WT) as the transport device; the 2WT is also seen as a method to supply tillage and other services, such as haulage during slack periods. However, infrastructure development is also an important part of a wider national strategy for economic development, including agricultural mechanization. The failure of Nepal’s public sector to develop rural infrastructure, which was also on the priority list of APP, has adversely affected overall growth, including the process of mechanization. This is a higher-level policy issue outside the scope of this paper, but the need for concerned agencies to address this is beyond doubt. Focus on small-scale machinery and custom hiring There are strong arguments for the government to promote small-scale machinery in Nepal when the examples of China and Bangladesh (described below) are considered. This case is strengthened by the small landholdings and difficult terrain in most of Nepal, which would seem to particularly favor small-scale machinery. There needs to be a policy on how best to supply small-scale machinery to smallholder farmers. Small-scale machinery is not always the best choice. Large size machinery for threshing in the Terai for example is the best technology and custom hiring of such expensive machinery is the only viable economic option for the vast majority of farmers. In the past, this business (custom hiring of large equipments) was dominated by Indian machine owners but more recently, increasingly, Nepalese machine owners custom hire equipments for tillage, harvesting and threshing. There is a need to develop and implement an appropriate policy to facilitate custom hiring for the wider benefit of smallholders. Biggs et al. (2011) and Justice and Biggs (2011) cite examples from China and Bangladesh in favor of small machinery. In China, the government initially prioritized the production of small horsepower pumpsets and tractors at a lower level of technical efficiency, and a significantly lower cost than elsewhere. Lower-quality Chinese equipment turned out to be good enough to meet the basic eco- nomic and technical needs of two groups: small farmers with fragmented plots, and small-scale entrepreneurs who bought equipment to hire out as part of the increasing growth of rural agricultural service industry (Qiuqiong et al. 2007). Small-scale equipment entered Bangladesh in the 1990s following a major change in policy in the late 1980s resulting from the national food crisis. This change in policy combined with market liberaliza- tion, and lowering of tariffs resulted in massive imports of small pump-set engines for irrigation; later, 2WTs and other equipment mainly powered by small-scale and cheap Chinese diesel pump-set engines also arrived. It gave Bangladesh possibly the most mechanized and labor-intensive agricul- ture sector in South Asia with substantial employment and growth linkages to rural and urban sectors (Biggs et al. 2011a, Justice and Biggs 2011). The situation in India stands in striking contrast to that of China and Bangladesh. In spite of India being the number one producer of 4WTs in the world, only 22 percent of its area is under mechanized tillage (Kulkarni 2009) compared to 80 percent mechanization in Bangladesh and Sri Lanka (Biggs et al. 2011b). This may be due to a small number of 2WTs—110,000, a third of the number in Bangla- desh. Variation in mechanization between Indian states has been attributed again to government policy that favors large equipment. Policy gap It appears that not only there is no agricultural mechanization policy, but some government policies to address wider issues are counter productive. For example, the heavy duty on importing raw materials for construction and manufacturing has created a disincentive to all local engineering firms to engage in local production and sales of tools and equipment. On the other hand, imported ready-to-use 23 machinery that does not attract the same duty is not always suitable for smallholder farmers and the farming situations in the hills and mountains. One option could be to provide tax relief to local engi- neering firms and equipment manufacturers for the import of raw materials used for manufacturing agricultural machinery. There are a number of engineering workshops/metal craft workshops, mainly concentrated in the Terai, that are also involved in manufacturing agricultural tools, equipment, and machinery. They cannot survive on the business of agricultural tools and equipment alone so most have diversified to include non-agricultural manufacturing, which forms the major part of their production and sales. There is a lack of reasonable support by government for agricultural mechanization. Conversely, in Bihar, India, farmers get a 50–80 percent subsidy on various kinds of machineries (G. Sah, pers. comm.). The public sector could fund some low-cost research to find out what rural mechanization already exists in farmers’ fields, how mechanization processes are taking place, and who appears to be getting the benefits and why. The study would determine what processes for rural mechanization need to be promoted to help reduce rural poverty and increase worthwhile rural employment. It may well mean, say, the promotion of two-wheel tractor programs by offering credit to small farmers, who then hire the vehicles out to others for transporting or threshing grain, etc. Another opportunity highlighted in policy analysis on agricultural mechanization is to link this sector with occupational caste of Nepal to improve livelihoods and reduce poverty. This will help generate employment and poverty reduction in rural areas by engaging liberated Haliyas9, particularly mi10andLuhar, in manufacturing agricultural tools and equipment and helping to establish their cooperatives for the sale of such machinery. Favorable government policies (whether direct or indirect) are vital for the promotion of agricultural mechanization. Research and development on agricultural mechanization would need to be participatory and client-oriented so that appropriate farmer-friendly tools, equipment, and machinery could be developed and promoted. Industry structure and access to new technologies Most Agrovets (nearly 90 percent) are situated in the Terai and accessible areas in the low hills. The key informant survey with Agrovets in the Terai districts revealed no difference whatsoever in the access of farmers in different wealth categories to input suppliers. Small- and medium-sized farmers constituted nearly 84 percent of all Agrovet customers (63 percent medium and 21 percent poor). These figures largely agree with the official figures that 75 percent of all holdings are of less than 1 hectare of land. Most farmers (86 percent) accessed inputs from an Agrovet located within 15 km of their farms. The presence of private-sector agricultural inputs and service providers in the remote rural areas is almost non-existent. This is evidenced by the case study from the hills and mountainous districts, and an analysis of the organizations that placed demand for source seeds of three major cereals. When- ever Agrovets and other private-sector agencies are present in the hills, they are limited to the district headquarters for the obvious reason that there is an inadequate volume of business in rural areas. The overall industry structure can be summarized with the help of the following matrix (Table 3.2). It clearly demonstrates the public sector is still a major player for agricultural research, extension, and seed supply both in accessible areas and remote regions. Private-sector actors are slowly becoming important players in accessible areas for most agricultural services, except research and extension. There is a lack of competition and plurality when one considers all the agroecological regions and service areas. 9 Haliyas are traditionally bonded plowmen from the Dalit families responsible for all plowing works in the hills for so-called upper-caste people. All families with such arrangements have been liberated by a government order without proper arrangement for their settlements and livelihood securities. 10 Kami and Luhar are an occupational caste mainly responsible for making agricultural tools and equipment. 24 TABLE 3.2.INDUSTRY STRUCTURE IN NEPAL IN RELATION TO SEED AND AGRICULTURAL INPUTS Source: Analysis by the authors based on this study. The majority of smallholders in remote rural areas of hills, mountains and Terai have limited access to NARC research. They neither have a say in the processes of research priority-setting and technology development nor access to new technologies once they have been developed by NARC. The analysis of overall effects of industry structure on the ability of smallholders to access new technologies revealed that most of the desirable technologies are not readily available. Of these, new seed varieties and other agriculture-related technologies are more appropriate for smallholders (Table 3.3). TABLE 3.3.THE EFFECT OF INDUSTRY STRUCTURE ON THE ABILITY OF SMALL-SCALE RESOURCE-POOR FARMERS TO ACCESS NEW TECHNOLOGIES Source: Analysis by the authors based on this study. Note: IPM =integrated pest management. Policy options to improve technology access Policy options If anything is dynamic in Nepal, it is the policy environment. Policies keep changing before being fully implemented. Currently, preparations are underway to revise or replace the following policy instru- ments: Agriculture Perspective Plan (APP) The major policy framework for agriculture growth in Nepal is undergoing a revision to develop an Agricultural Development Strategy (ADS). Major gaps were identified in the implementation of the APP, e.g. poor coordination between government line-departments resulting in weak implementation, Region/agricultural services Involvement of different sectors Overall structure Public Private Civil Society Degree of competition Plurality Terai, river basins and low hills (< 1,000 m) Agricultural research High Low Medium Medium High Agricultural extension High Low High Medium High Seed supply Medium High Medium High High Fertilizer supply High Medium None Low Low Pesticides supply Low High None Low Low Machinery supply Low High Very low Low Low Hills and mountains (> 1,000 m) Agricultural research High Low Medium Low Low Agricultural extension High Low High Medium Medium Seed supply High Low Medium Low Low Fertilizer supply Medium Low None Low Low Pesticides supply Low Low None Low Low Machinery supply Low Low Very low Low Low New technologies Accessibility to smallholders Existence of new technolo- gies in Nepal (Yes/No) Appropriateness (High/Medium/Low) Availability (High/Medium/Low) Affordability (High/Medium/Low) Seeds of new varieties Yes Medium Medium Medium to High Fertilizers/nutrient management technologies Yes High Low Medium Pesticides/IPM Yes Low Low Low to Medium New machinery or equipment Yes Low Low Low 25 a lack of financial resources, and a lack of needed institutional reforms. All of these gaps would need to be addressed by the new strategy. New fertilizer policy of Nepal The fertilizer policy is also undergoing change. A new fertilizer policy must first address the need to supply Nepal’s full fertilizer requirements (a demand for 580,000 tons per year), and those figures do not even consider additional fertilizer needed for increased crop intensification and increased com- mercialization of agriculture. One strategy is to try to strengthen existing agricultural production systems by building on and improving existing on-farm practices for soil fertility management using the concept of Integrated Plant Nutrient Management (IPNM) that integrates all available means of soil, nutrient, and crop manage- ment to achieve optimum productivity. These include organic, mineral fertilizers, and biological sources of plant nutrients and better crop husbandry (Subedi and Weber 2001). This is particularly needed for the hills and mountains where it is more difficult to deliver fertilizers. However, both IPNM and targeted fertilizer-dosage involves a quite complex extension message, and would need to have an efficient extension system to disseminate it. A second strategy to address issues related to fertilizers is to be more context-specific and account for the high agroecological diversity in Nepal—this is an approach that current policy does not cover. Soil-nutrient status across agroecological domains needs to be determined to identify and quantify nutrient-deficiency problems. There has been no such study since the mid-1980s, but it would allow recommendations for economical fertilizer doses for various cropping patterns across agroecological regions. There are vast differences across Nepal in production potential, access to inputs, markets, and technologies so a future policy on fertilizer should not have a blanket recommendation. Separate policy options are needed for accessible versus inaccessible areas and high-potential production systems with high cropping intensities versus marginal agriculture. Although this would help increase the efficiency with which fertilizers are used, it does not solve the problem of inadequate supply. A third strategy is to invest in manufacturing phosphatic fertilizers over the medium-term horizon (three–five years) in areas such as Baitadi and Bajhang where considerable phosphorite deposits have been identified. Laboratory analysis indicated 15–32 percent P2O5 in the ore (Pradhanag 1986). A fourth strategy is to initiate bilateral talks between the Governments of India and Nepal to include fertilizer as a commonly traded commodity. As a short-term option, which could then be linked with a bilateral agreement, the Nepal government could encourage farmers to bring quality fertilizers from India to meet their needs. It could also explore the possibility of small traders, Agrovets, and private companies bringing fertilizers using a variety of means for use in the Terai and other accessible areas in the hills. Given the lack of fertilizers supplied on time and in adequate quantities through govern- ment parastatals, an informal flow of fertilizers from India (which are also cheaper due to substantial government subsidies) fills the gap. Good quality fertilizers so imported would always contribute to improve the crop productivity of Nepal. Nepal's fertilizer market is merely one percent to that of India (Thapa 2006) and even if fertilizer is made a commonly traded commodity, it would hardly have any adverse impact on the overall fertilizer situation in India. A fifth strategy is based on extending government subsidies to cover sufficient fertilizer to meet needs. This involves a broader policy consideration by the Nepal government. For example, what should be the overall policy regime on government subsidies? What priority should be assigned to subsidies on petroleum products versus fertilizer (and diesel) for agricultural operations? Currently, such policy guidelines are only evidenced by the subsidies in place. If fertilizer subsidies are extended, then the existing problems of delivery have to be avoided. Current- ly, the National Fertilizer Company Limited (NFCL) is the sole agency for handling subsidized fertiliz- ers, which is sold through the network of cooperatives (owned by the government) and fair-price cooperatives. This system is highly criticized as fertilizers are never imported and delivered on time and there is inefficiency and financial misappropriation. The private sector could be encouraged to procure and sell subsidized fertilizers, although they would have to be assured of having equal terms 26 to that of NFCL. There may be a need for an initial investment in this sector as a joint venture (public- private partnership) if fertilizer is imported in large consignments from the international market and not just India. Alternatively, many Agrovets already deal in fertilizer and are capable of expanding sales if they can officially import it from India. Another way of delivering subsidized fertilizer on a pro-poor basis is the introduction of a targeted voucher system for distributing fertilizers (Denning et al. 2009). However, such a system is not without difficulties such as high administrative costs, and the reselling of vouchers by the target group. Also, the voucher system was designed to encourage private input-dealers in Malawi and there is less need to do this in Nepal. Nonetheless, if a voucher system were to be used in Nepal, the vouchers would have to be redeemable in both the private sector and government outlets. Improvement in seed regulatory frameworks Encouraging private investment in plant breeding and the seed trade would contribute to food security and poverty reduction by increasing the availability and choice of better technologies for smallholders. However, amendments to the variety release process are needed to encourage substantial private sector investment on plant breeding:  An independent authority should generate data on distinctness, uniformity, and stability (DUS). Currently this is done by the Nepal Agricultural Research Council (NARC), which is al- so the major crop-breeding and variety-releasing institution in Nepal. Hence, there is a conflict of interest  Representation of stakeholders from outside of NARC on the Variety Approval, Release, and Registration Committee (VARRC) should be increased. Currently, the committee is comprised of a majority of NARC scientists, and again there is a conflict of interest  The provision of “truthfully labeled” seeds should be popularized to help in the deregulation of the market and to give greater flexibility to the private sector  A strong quality-monitoring mechanism should be put in place to discourage the sale of spuri- ous seeds falsely packaged with the branding of reputable companies. The slow pace of introducing new varieties delays their benefits for smallholders: policies are needed to encourage seed production of new varieties. A modest subsidy on new varieties, as applied in India, would help. The government often approves bilateral projects that involve the distribution of agricultural inputs such as the European Union–funded and FAO-managed Food Facility project, and the World Bank–funded Nepal Food Crisis Response Programme. They should insist that in such projects, priority is given to newer varieties adapted to specific ecological conditions. It would raise awareness of the importance of varieties and avoid mistakes that have been made in distributing seed of varieties un-adapted to the local environment. Since most varieties officially released by NRRP have had limited testing with farmers, feedback from farmers should be solicited for the initial seed distribution of the newest varieties. Policy on agricultural mechanization Agricultural mechanization can contribute to the growth of agricultural production and rural develop- ment while reducing rural poverty. Correct and timely government interventions are crucial to achieve these kinds of objectives. For example, the development of minor irrigation—particularly private, investment-based expansion of shallow tube wells—contributed to the impressive performance and expansion of boro rice in Bangladesh. This was an outcome of the government’s market liberalization policy for irrigation equipment in the late 1980s. This policy promoted rapid expansion of irrigated boro-rice farming in the dry season (Hossain 2009). Similar interventions are needed in agricultural mechanization in Nepal. In the same vein, linking rural and agricultural mechanization with occupa- 27 tional castes of Nepal, particularly liberated Haliyas 11 , Kami 12 and Luhar for manufacturing, importing, and marketing agricultural tools and equipment can contribute to poverty reduction in rural areas. This can be done by setting up cooperatives to increase employment opportunities, improve livelihoods, and reduce poverty. In the Terai, large businesses are dealing with agricultural machinery; there is little need to change policy to encourage these enterprises. However, these businesses do not operate in the low- and mid-hills and an alternative outlet is needed such as cooperatives. These could be linked to the businesses in the Terai by, for example, operating on a commission basis or by linkages for spare parts, repairs, and maintenance. Although the cooperative movement in Nepal is not without its problems, there are success stories as well. To encourage their establishment, the policy should be to promote training in business development services (BDS) and marketing skills, using experts from successful businesses or professionals from universities. Concentrating on selling agricultural machinery may be too restrictive and cooperatives would need to deal with other products. Revisiting public investment priorities and strengthening implementation: Analysis of return on public investment in various sectors may be needed for refining overall govern- ment investment plans. However, research by IFPRI has indicated that public investment on infra- structure such as rural roads, and in agriculture research and development (R&D) and education, can significantly reduce poverty. Increasing share of funding to impoverished and previously un-served regions can be truly a pro-poor public investment, generating income in rural areas by creating positive impacts on agricultural profitability, agricultural wages, rural non-farm employment, and overall poverty reduction (Renkow 2010). The importance of agricultural R&D for the nation's development has been highlighted in a number of important policy documents, but instability at the ministerial level and a lack of political will power continue to pose a major challenge; witness the failure of implementing policies, programs, and activities. The country needs to put effective and efficient policy implementation bodies and instru- ments for agricultural R&D in place if it is to succeed in enhancing smallholder product