Executive Summary. This paper discusses the current state of agricultural mechanization in Africa and its potential contribution to agricultural and broader

372 KB – 58 Pages

PAGE – 1 ============
Agricultural Mechanization and Agricultural Transformation Paper written by Xinshen Diao, Jed Silver and Hiroyuki Takeshima INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE BACKGROUND PAPER FOR African Transformation Report 2016: Transforming Africa™s Agriculture FEBRUARY 2016 Joint research between: African Center for Economic Transformation (ACET) and Japan International Cooperation Agency Research institute (JICA-RI) www.acetforafrica.org

PAGE – 2 ============
Agricultural Mechanization and Agricultural Transformation Xinshen Diao, Jed Silver and Hiroyuki Takeshima (International Food Policy Research Institute)February 2016Background Paper for African Transformation Report 2016: Transforming Africa™s Agriculture Joint research betweenAfrican Center for Economic Transformation (ACET) andJapan International Cooperation Agency Research institute (JICA-RI)

PAGE – 3 ============
1Contents Executive Summary 31. Introduction ..52. Definitions and Concepts ..6 Definitions of Mechanization 6 Mechanization and Agricultural Intensification .73. Demand for Mechanization .8 Sufficient demand as a precondition for successful mechanization .8 Components of Mechanization Demand 8 Farm Size and Mechanization .9 Labor Saving Effects of the Mechanization of Certain Operations ..9 Market Demand ..12 Complementary Technologies .13 Utilization and Income Generation Potential ..14 Learning and Demonstration of Mechanization™s Benefits ..14 Sequences of Mechanization Demand ..15 Demand for mechanized plowing .15 Demand for Animal traction 15 Demand for large and small tractors 17 Demand for post-harvest mechanization 19 Demand for mechanized harvesting ..19 Spatial Variations of Mechanization Demand 204. Mechanization Supply Chains in Africa .22 The Supply Chain as a Framewor k 22 Development of Supply Chains ..22 Key Players and Their Functions in Mechanization Supply Chains .23 Importation .25 Manufacturing of Agricultural Machinery ..29 Ownership and Service Provision 29

PAGE – 4 ============
2 Farmer-ownership characteristics .30 Farmer to Farmer Service Provision .3 Specialized Service Provision ..36 Fabricators, Spare Parts and Maintenance ..39 Rural repair services 39 Locally fabricated implements and small machines ..395. Roles of Government in Promoting Mechanization ..41 Providing Public Goods to Create an Enabling Environment for the Private Sector .41 Research on Mechanization Demand and Adoption .41 Support to R&D ..42 Infrastructure Investments 43 Mechanization Credit Schemes ..43 Eliminating Distortions 44 Potential Roles of Donors in Mechanization Promotion .45 Opportunities for Private-Sector Promotion of Mechanization 466. Main Messages and Conclusions ..46References 49

PAGE – 5 ============
3Executive SummaryThis paper discusses the current state of agricultural mechanization in Africa and its potential contribution to agricultural and broader economic transformation. This background paper reviews the factors likely to influence farmer demand for mechanization in Africa and details different existing and potential mechanization supply models. Although an empirical analysis of mechanization demand and the effectiveness of supply chains is beyond the scope of this paper, in part due to data limitations, this paper suggests that demand for mechanization may be emerging in some parts of Africa. It also suggests that private sector-driven supply models are better positioned to meet this demand than direct government involvement and certain types of subsidized programs. The paper then identifies possible areas for government support to complement private sector leadership in developing mechanization supply chains.A renewed focus on agriculture™s potential contribution to economic transformation in Africa has resulted in increased attention paid to agricultural mechanization. Nevertheless, African agriculture still relies predominantly on human muscle power, in contrast to other developing regions that have experienced rapid increases in agricultural mechanization over the past few decades. Efforts to promote mechanization in previous decades largely consisted of state-led interventions, which failed due to the lack of demand for mechanization among farmers (Pingali et al 1987).This paper attempts to overcome some of the misconceptions that drove these programs by reviewing definitions of agricultural mechanization and its role in agricultural intensification processes. The paper draws on Boserup (1965) and Ruthenberg™s (1980) theory of agricultural intensification and Hayami and Ruttan™s (1970; 1985) induced innovation theory. According to demand for agricultural products. This in turn prompts mechanization, both through the adoption of existing and the development of new technologies. Essentially, mechanization can be expected absence of such conditions. The paper then delves further into the components of demand for mechanization, specifically in Africa. Specifically, it discusses the effects of farm size, labor saving, market demand, the availability of complementary technologies, and demonstration on developing demand. It also describes the sequential nature of mechanization demand as postulated by Pingali et al (1987): in which power-intensive operations (plowing, threshing and harvesting) are mechanized before control-intensive ones (planting, weeding, winnowing) and animal power (where feasible) is adopted before the transition to mechanized power. As a result of the components and sequences spatial variation, meaning that existing national surveys and other data may fail to accurately capture patterns of demand. Our analysis then turns to mechanization supply chains in Africa, focusing on manufacturing, importation, service provision, and ownership. The analysis suggests that where demand has emerged, the private sector has been relatively responsive to meet demand. Private dealers tend to import the types of machinery demanded by farmers, for which markets for spare parts and repairs may be relatively developed, while governments tend to import the brands of machinery accessed

PAGE – 6 ============
4through concessional loans, which do not possess these advantages. Moreover, government importation and subsidization of machines may produce distortionary effects on the private importation and distribution channels. Similarly, individual machine owner-operators appear to have an advantage over government-subsidized service-provision enterprises, as they are usually their own farmers. Despite the apparent advantages for the private sector in machine distribution and service Perhaps the most significant of these roles is providing public goods, including infrastructure, technical R&D, and economic research. Other potential roles include capacity building, removing distortionary policies, facilitating access to credit and formulating viable strategies. However, the private sector is still better positioned to drive mechanization in areas where demand has emerged and government policies should aim to play a facilitative and supportive role. Despite a history of disappointment, agricultural mechanization may finally be in position to contribute to an agricultural transformation in parts Africa. Demand for mechanization appears to have emerged in certain systems and where it has, the private sector has often demonstrated its potential to efficiently supply machines and hiring services. However, the evidence base to better understand the changing nature of mechanization demand in Africa and the extent and effectiveness of different supply models in meeting it.

PAGE – 8 ============
6been conducted in other African countries and national-level data does not adequately cover variation, there do appear to be case of emerging private demand for mechanization among farmers of different sizes. Farmer to farmer service provision appears to be the most effective method of meeting this demand where it exists, while government direct and subsidized service provision may encroach on the private sector™s comparative advantage in providing mechanization services. However, the paper discusses a number of constructive roles for governments to play in promoting mechanization in the context of overcoming the constraints to broader agricultural transformation. These roles include the provision of public goods such as training, research and knowledge, favorable trade policy, infrastructural development, and facilitating access to credit. In the following section (Section 2) we give a set of definitions of mechanization drawn from the literature. The definitions are given from the perspective of the role of mechanization in 3, in which we discuss the main factors contributing to demand for mechanization among different types of farmers. We also attempt to describe observable demand patterns of mechanization in Africa. In Section 4, we focus on the supply side of mechanization, considering mechanization supply as a chain with many different actors including manufacturers, importers and distributors, as well as mechanization service providers and machinery maintenance providers. In Section 5 we focus on the role of governments in mechanization and provide an overview existing and potential of government policies affecting mechanization. We also try to identify potential areas in which government intervention could help to promote mechanization as a facilitator of agricultural transformation. We provide a set of concluding recommendations in Section 6. FAO defines mechanization as fithe application of tools, implements and machinery in order to achieve agricultural productionfl (Clarke 1997). These can all be operated by manual, animal or engine (fossil fuel or electric) power. Essentially, agricultural mechanization represents technological change through the adoption of non-human sources of power to undertake agricultural operations. Mechanized agricultural operations can be grouped into power and control intensive functions. Mechanization of power intensive agricultural operations, such as land preparation, threshing, grinding and milling, is characterized by non-human sources of energy input to replace human and animal ones required in the operations. On the other hand, mechanized control intensive operations, such as planting, weeding, winnowing, fruit harvesting, require greater human judgment and mental input in addition to energy (Pingali et al 2007). Grain harvesting can be thought of as both a power and control intensive operation (Binswanger 1986; Pingali 2007). Some literature also separates stationary operations, such as milling, water lifting, and threshing, from mobile operations, which include plowing, weeding, and harvesting (Rijk 1999). Distinctions between power and control intensive operations, and stationary and mobile operations, are important for understanding the demand for mechanization. As shown in later sections, mechanized operations often have certain sequential patterns. Engine-powered irrigation and transport are two activities that are sometimes included in mechanization. However, in this paper we exclude them in most cases, except where they complement the mechanization of other agricultural operations.

PAGE – 9 ============
7 Ruthenberg (1980), and from the perspectives of long-term evolution of farming systems and through reducing fallow periods (intensified use of land). However, in agricultural economics literature that does not focus on the long-term evolution of farming systems, agricultural intensification is exclusively referred to the intensified use of inputs, while the intensified use of land that often leads to the expansion of cropping areas by reducing forest or fallow land is Treating farming system evolution as well as technologies used under different farming systems as endogenous processes, Boserup (1965), (and later further formalized and tested by Pingali et al. with increased demand for agricultural products. Such increased demand is the result of population growth and improved market access, including both domestic and international market access (which expands agricultural demand beyond farmers™ own subsistence needs).In response to the increases in demand for agricultural products, intensification in agricultural production is also a process of agricultural technological changes, in addition to the increased use of agricultural inputs. Mechanization is a key component of the technology that allows agricultural market demand, or when existing land has to be more intensively cultivated, which requires more labor use per unit of land, mechanization will be adopted. Therefore, the dynamic relationship process, is another key factor influencing mechanization. Thus, mechanization, particularly the demand for mechanization, can also be explained by applying the induced technical change framework developed by Hayami and Ruttan (1970, 1985). Beginning in the early 1970s, Hayami and Ruttan (1970, 1985) and Binswnger and Ruttan (1978) formulated a model of induced technical change in which the development and application of new technology is endogenous to the economic system (Ruttan 2002). This framework allows us to assess emerging demand for mechanization as part of a technology adoption process. The induced technical change model emphasizes agricultural technology innovation and adoption as a continuous sequence often biased toward saving the limiting factor Œ land or labor – as the relative Ruttan 1970). In this model, alternative agricultural technologies are developed (and adopted by farmers) to facilitate the substitution of relatively abundant (cheap) factors for relatively scarce (expensive) factors (Ruttan 2002). Mechanical technology is designed to substitute power and machinery for labor and is filabor savingfl, while biological and chemical technology is filand saving.fl Moreover, changes in land and labor productivity are relatively independent (Griliches 1968), indicating that adoption of labor-saving technology by farmers is not necessarily driven by an incentive to improve land productivity, which is the case for the adoption of biological technology. Given the relatively high land-to-labor ratio on agricultural endowments in many African countries, mechanization may play a greater role in African agricultural intensification than

PAGE – 10 ============
8it did in the intensification processes observed in the Asian Green Revolution (GR). In most Asian countries during the GR, land-to-labor ratio was low and rural non-farm employment opportunities were few. In some parts of African countries, including parts of Ghana, Nigeria, Senegal and Zambia, land is more abundant than in many Asian countries that underwent the Green Revolution (Nin-Pratt and McBride 2014). In addition, the urbanization process is more rapid in many African countries in recent years, as many of them rely on natural resource exports, and also as the service sector accounts for a much larger share of their economies than most Asian countries did at similar levels of per capita income. Such structural characteristics of some African economies also lead to the development of ficonsumption citiesfl in which urbanization occurs without industrialization (Gollin et al 2013). Migration to the urban areas as well as increased employment opportunities in non-farm services in the rural areas could create pressure on rural wages (Byerlee 1974), even though agricultural land productivity, measured by yield, is still much lower in most African countries than in the post-GR Asian countries. As a result, demand for labor- saving technology could become a necessary condition for further agricultural intensification in some African countries at least in areas with better market access and higher opportunity costs of rural labor. Thus, it is possible that the lack of labor-saving technology limits the potential returns to certain types of land saving technology such as improved seeds, and application of fertilizers and pesticides, unless such labor constraints can be overcome through mechanization (Nin-Pratt and McBride 2014). 3. Demand for MechanizationIn general, demand for mechanization emerges at the point when it becomes cost effective for farmers to use it over other available options. Thus, policy interventions aimed at promoting mechanization must first confirm whether sufficient demand is indeed present. Nearly all of the 30 mechanization schemes in Africa from 1945-1987 studied by Pingali et al (1987) failed to recognize the lack of demand amongst farmers. This is seen as one of the key factors that led the governments in these countries to introduce tractors at an inappropriate stage. Without demand in place, tractor hiring services quickly collapsed under these programs, and machines were often left idle, scrapped, or abandoned (Pingali et al. 1987). It is reasonable to argue that promoting mechanization when demand is insufficient tends to be socially suboptimal and can also have adverse equity effects. As Pingali (2007) argues that where the potential and demand for aggregate land expansion is limited, increased tractor use by medium and large scale farmers is likely to displace tenant farmers or hired labor provided by landless farmers. While this effect has been observed in South Asia (Lockwood et al 1983; Jabbar et al 1983; Singh et al 2013), there is little evidence of it occurring in Africa. Although displacement and adverse equity effects do not appear as strong as suggested by a wave of cross-continental ILO studies in the early 1970s, they nevertheless remain important concerns going forward (Mrema et al 2008). Components of Mechanization DemandAs mentioned before and further discussed below, demand for mechanized agricultural operations sequences of mechanization demand. Moreover, such demand sequences are also affected by

PAGE – 11 ============
9different agro-ecological conditions and the availability of technologies and the ability to maximize their utilization.Farm Size and Mechanization In a society where there are both large and small farmers, tractors can be essential for expanding the aggregate area cultivated by large farms, for whom hired labor represents a high proportion of their production cost. The economies of scale associated with a large machine such as a tractor have also made mechanization a more attractive technology to such farms (Binswanger 1986). As a provide hiring services to non-owners when it helps them maximize their tractors™ utilization. Such trends have been observed in Asia. In Punjab, India, tractor owners typically own 4 times as much land as tractor hirers (Singh et al 2013); in Thailand, tractor hirers also had much greater share of mechanization in Asia has been adopted by smallholders (as described later) cases from Asia show that mechanization is often driven by large farm sizes and enables farmers to further Mechanization in many parts of Africa appears to be following this pattern. North African agricultural has long been characterized by large farm sizes. Meanwhile, Ghana and Zambia are Sub-Saharan examples of countries with land dynamics suitable to mechanization, with rapidly rising farm sizes in recent years leading to medium-scale farmers (5-100 ha) cultivating the largest share of national cropland (Jayne et al 2014). In the 2013 IFPRI/SARI survey of medium-to-large farmers in Northern Ghana, over half of tractor owners cited land expansion as the primary motivation for their investment (Chapoto et al 2014). In the same survey, farmers who hired in tractor services also expanded their land more than those who did not use tractors, though at a much lower rate than tractor owners did (Houssou et al 2015). Likewise, land expansion appears to be a major factor for the adoption of mechanization in Southern Nigeria, mostly to expand the input area of input-intensive rice cultivation, but does not seem to apply in the North, where it substitutes for household labor on small, intensively-cropped farms and allows for the pursuit of off-farm income generation (Takeshima et al 2013). In general, the presence of households with large farm sizes and the potential for area expansion suggests that mechanization is more likely to these conditions. Labor Saving Effects of the Mechanization of Certain Operations typically results in small decreases in the labor required for weeding and harvesting (Pingali et al 1987). Where hired labor represents a relatively large share of production costs, even smaller farmers, begin to demand mechanization technology when labor cost starts to rise in order to reduce their labor and total production costs. For example, according to data from GLSS 5 (Ghana Living Standards Survey 5) for 2005-06, hired labor represents 40% of paid input costs on average in Ghana, compared to 28% for fertilizers (Figure 1a) 2. This cost share is even higher in the relatively land abundant northern regions of Ghana, even though per capita income is lower in the north than in the south in Ghana; the share of hired labor costs is around 50% in Northern 2 Note that Ghana™s fertilizer subsidy was not in effect during the GLSS5 period, but was during the GLSS 6 period

372 KB – 58 Pages