Ten striking facts about agricultural input use in Sub-Saharan Africa Megan Sheahan and Christopher B. Barrett Presentation for the workshop on Structural Transformation in African Agriculture and Rural Spaces (STAARS) African Development Bank Headquarters, Tunis, Tunisia, November 11-12, 2014 A summary of work prepared under the “Myths and Facts” project Introduction Improved agricultural productivity is a primary pathway by which societies can begin down the path of economic transformation and growth and out of subsistence level poverty. Expanded use of modern agricultural inputs, embodying improved technologies, is often seen as a prerequisite to increasing agricultural productivity. Asia and Latin America enjoyed tremendous increases in agricultural productivity through rapid and widespread uptake of yield-enhancing modern agricultural inputs. Benefits accrued to both producers and consumers, helping stimulate historically unprecedented economic growth and poverty reduction in east and southeast Asia. Introduction What about Sub-Saharan Africa? Prevailing wisdom = “African farmers use few modern inputs” Well-perpetuated claim grounded in: • Macro-statistics (e.g., FAOStat and World Bank’s Development Indicators) • Studies derived from micro-data with small or purposively chosen samples • Case studies with limited statistical underpinnings • Data collected 10-20 years ago Major changes in SSA in last 10-20 years: • High and volatile food prices • Urbanization and growth of a middle class • Increased investments in agricultural sector (including fertilizer subsidy programs) • New technologies available to farmers (cell phones) • Changing bio-physical environment (climate change, soil erosion) Introduction It’s time to update our understanding of the agricultural input landscape in Sub-Saharan Africa. 1. 2. 3. 4. 5. 6. 7. Large cross section of SSA’s population Cross-country comparable Strong focus on agricultural data collection Plot, household, and community level information Nationally-representative statistics as well as within-country (and even within-household) analysis Statistics derived from farmers’ accounts Coupled with growing collection of geo-referenced data sets Living Standards Measurement Study Integrated Surveys on Agriculture Burkina Faso Ethiopia Malawi Mali Niger Nigeria Tanzania Uganda Introduction We use one cross section of LSMS-ISA data collected between 2010 and 2012 in each of six countries (Niger, Nigeria, Ethiopia, Malawi, Tanzania, Uganda), including over 22,000 cultivating households and 62,000 agricultural plots • Objective: update the basic facts on agricultural input use in SSA through descriptive statistics • Not our objective: uncover casual pathways for these conditions Focus on fertilizer, improved seed varieties, agro-chemicals (pesticides, herbicides, fungicides), irrigation, and mechanization Huge number of descriptive statistics included in Sheahan and Barrett (2014) World Bank Policy Research Working Paper No. 7014 10 most important facts presented here… Sample selection and variable creation Sample: any sampled household cultivating at least one agricultural plot in the main growing season (mostly rural but not exclusively) Country Year Season # hh # plots Ethiopia 2011/12 - 2,852 23,051 Malawi 2010/11 Rainy 10,086 18,598 Niger 2011/12 Rainy 2,208 6,109 Nigeria 2010/11 - 2,939 5,546 Tanzania 2010/11 Long rainy 2,372 4,794 Uganda 2010/11 First 1,934 3,349 Variable creation: • Variables created and data “cleaned” using the same rules across all data sets and countries • Use of imputed plot size values to limit known reporting bias • Household sampling weights as well as calculated plot level weights 1 of 10 “striking” facts Modern input use may be relatively low in aggregate, but is not uniformly low across these six countries, especially for inorganic fertilizer and agro-chemicals. • Average inorganic fertilizer use rates > widely quoted 13 kg/ha statistic in 3 of 6 countries, simple six country average nutrient application rate of 26 kg/ha Share of cultivating households (%) using input on fields 100 77 80 40 • Application rates are highest in Malawi and Nigeria, both with government input subsidy programs, and Ethiopia 56 60 33 31 41 17 20 3 13 8 17 11 3 0 Ethiopia Malawi Niger any agro-chemical Nigeria Tanzania Uganda inorganic fertilizer 121 Inorganic fertilizer application (nutrients in kg/ha) 120 100 80 • Relatively high shares of households use inorganic fertilizer, with 3 of 6 countries > 40 percent • Where > 30 percent of households use agro-chemicals on plots (others used in storage), any implications for human health? 64 56 60 40 33 26 25 23 20 0 2 1 6 8 7 Micro data (LSMS-ISA, 2009-2011) 12 13 1 2 Macro data (World Bank, 2010) 2 of 10 “striking” facts The incidence of irrigation and mechanization, however, remains quite small. Water control is limited 10 Mechanization is proceeding slowly 9 9 Ownership or rental: • 8 7 7 – Traction animal ownership >20 percent in all countries except Malawi – 1-2 percent of households own a tractor, not many more rent – 32 percent of households own and 12 percent of households rent some type of farm equipment that could be used for mechanization 6 5 4 3 2 1 0 5 4 4 4 4 3 1 0.4 1 2 2 0.2 Use: • % of all cultivated land under irrigation by smallholders – ~50 percent of households in Nigeria used a mechanized input or animal power on their plots – >50 percent of households in Ethiopia used oxen to prepare their plots % of households with at least some irrigation on farm • 5 percent of households use some form of irrigation, covering only about 2 percent of land under cultivation 3 of 10 “striking” facts Considerable variation exists within countries in the prevalence of input use and of input use intensity conditional on input use. Agro-chemicals Inorganic fertilizer Suggests need for research to understand drivers of within-country agricultural input use variation. 4 of 10 “striking” facts There is surprisingly low correlation between the use of commonly “paired” modern inputs at the household- and, especially, the plot-level. Ethiopia: household level Ethiopia: plot level Raises important questions about prospective untapped productivity gains from coordinated modern inputs use. 5 of 10 “striking” facts Input intensification is happening for maize in particular. Plots with mostly maize are among those most likely to receive a modern input and with the highest application amounts, including agro-chemicals Related: plots that include a major cash crop (<25 percent of all plots) are generally no more likely to receive modern agricultural inputs 25-40 percent of maize cultivating households purchased new maize seed ~25 percent of maize cultivating households in Ethiopia and >50 percent in Malawi used an improved variety 6 of 10 “striking” facts An inverse relationship consistently exists between farm or plot size and input use intensity. Nigeria: farm level Nigeria: plot level Local polynomial smooth kg/ha of inorganic fertilizer applied to field Local polynomial smooth 200 150 100 50 0 -50 0 200 150 100 50 0 1 2 3 Total hectares of land under cultivation 95% CI lpoly smooth kernel = epanechnikov, degree = 1, bandwidth = .34, pwidth = .51 4 0 .5 1 1.5 plot size in hectares 95% CI lpoly smooth kernel = epanechnikov, degree = 1, bandwidth = .2, pwidth = .3 In most cases, this relationship is more pronounced at the plot level, therefore inter-household variation cannot explain relationship. Suggests need to better understand intra-household agricultural input allocation decisions. 7 of 10 “striking” facts Farmers do not significantly vary input application rates according to perceived soil quality. Simple descriptive statistics: farmers do not appear to adjust input application rates to accommodate their perceptions of plot soil quality (Malawi, Tanzania, Uganda) “Within household” regression analysis: plots deemed “average” or “poor” quality are more likely to receive inorganic fertilizer applications, however only explains a tiny amount of variation Farmers do not make different input use decisions across eroded and non-eroded plots (Niger, Uganda, Malawi, Tanzania), including with respect to organic fertilizer Suggests a need for extension programming around soil fertility and input use and the need to invest in inexpensive soil quality tests 8 of 10 “striking” facts Few households use credit to purchase modern inputs. In all countries except Ethiopia, less than one percent of cultivating households used credit— either formal or informal—to purchase improved seed varieties, inorganic fertilizer, or agro-chemicals. In Ethiopia, where there exist widespread input credit guarantee schemes operated by cooperatives, nearly 25 percent of cultivating households claimed to receive some type of “credit service,” although we cannot be sure whether this is for agriculture or other household purchases. Reinforces widespread perceptions of the weakness of agricultural input credit markets in the region. Much scope remains for deepening rural financial markets, despite recent advances in money transfer systems based on mobile phone platforms, the proliferation of microfinance institutions, etc. 9 of 10 “striking” facts Gender differences in input use exist at the farm and plot level. • Male headed households are more likely use modern inputs across almost all countries and input types • Plots managed or owned by men (88 percent of all plots), are more likely to receive inorganic fertilizer and in higher amounts; almost always holds when controlling for gender of household head • Male headed households are more likely use modern inputs across almost all countries and input types • Related to work on “gender gap” in ag input productivity • Plots managed or owned by men (88 percent of all plots), are more likely to receive inorganic fertilizer and in higher amounts; almost always holds when controlling for gender of household head • Related to work on “gender gap” in ag input productivity 10 of 10 “striking” facts National-level factors explain nearly half of the farm-level variation in inorganic fertilizer and agro-chemical use. Variation in household-level inorganic fertilizer use Categories of variables Bio-physical variables: rain, soil, elevation, maximum greenness, agroecological zones Socio-economic variables: consumption level, sex of household head, household size and dependency ratio Farm operation characteristic variables: farm size, number of crops, type of crops Market and accessibility variables: distance to market and road, prices of fertilizer and main grain Country dummy variables Shapley value 24 4 16 11 45 • Ultimately interested to learn where most of the variation in input use comes from: biophysical, infrastructure, market, socioeconomic, or policy-specific variables? • Binary use at household level (avoids bias from survey design) • R2 decomposition using Shapley-Owen values • 45 percent of variation in inorganic fertilizer use can be explained by country level • Similar for agro-chemical use (43 percent) Suggests the policy and operating environments facilitated by governments and regional processes (e.g., CAADP) are critically important for ushering in a Green Revolution in Sub-Saharan Africa. Conclusions Modern agricultural input use in Sub-Saharan Africa is far more nuanced and varied than current claims suggest. Confirmed longstanding conjectures: • • • Irrigation and mechanization remain limited Women farmers use fewer inputs than men Agricultural input credit use is virtually non-existent New findings that suggest more policy-relevant research opportunities: • • • • • • • More agro-chemical use by smallholder farmers than commonly thought Huge amount of across and within country variation in fertilizer and agro-chemical use Input use is no higher on cash crops; maize is receiving a fair amount of input use Very little pairing of inputs with bio-physical complementarities at the plot level Little correlation between farmer-perceived plot quality and input use Input use intensity if more related to plot than farm size Country-level factors explain large amount of variation in household fertilizer and agro-chemical use