Agric ricultura ral I Input S Subsidy bsidy P Pro rogra rams - PDF document

10/31/13 ¡ Agric ricultura ral I Input S Subsidy bsidy P Pro rogra rams s in T in Theory ry a and in d in P Pra ractic ice: The C Case se o of Z f Zambia bia Nicole M. Mason (MSU/AFRE), T.S. Jayne (MSU/AFRE), & Rhoda Mofya-Mukuka (IAPRI) Eye on Africa Seminar, MSU African Studies Center 31 October 2013 Input subsidy programs (ISPs) in Africa 1 § Cornerstone of many countries’ agricultural sector & poverty reduction strategies § 2011: 10 SSA countries spent US$1 billion on ISPs (29% of public spending on ag; 60% of SSA population) (Jayne & Rashid 2013a) § Fertilizer, hybrid seed at subsidized prices § ISP objectives: § Increase access to and use of modern inputs § Raise crop yields and production § Improve food security and reduce hunger § Raise incomes and reduce poverty 1 ¡

10/31/13 ¡ We’ve been here before … 2 § Universal ISPs common until 1980s/early 1990s § Scaled back – structural adjustment § Renaissance over last decade (targeted/universal) § 2003: AU Maputo Declaration / CAADP – 10% of national budget to agriculture § 2006: AU Abuja Declaration – “member states … must improve farmers’ access to fertilizer, by granting … targeted subsidies … with special attention to poor farmers” § 2007: Morris et al. – “smart subsidies” - “ if subsidies are to be used to promote fertilizer use, they should be market- smart ” Smart subsidy criteria (Morris et al. 2007) 3 1. ISP is part of a wider strategy 2. Support market development / private sector investment 3. Promote competition 4. Pay attention to farmer demand 5. Insist on economic efficiency 6. Put farmers in the driver’s seat 7. Have an exit strategy 8. Pursue regional integration 9. Ensure sustainability 10. Pro-poor 2 ¡

10/31/13 ¡ Outline 4 § Zambia’s ISPs over last decade § Objectives, design, implementation, impacts How do these compare to “smart subsidy” ¨ concepts and how has this affected outcomes? § How to redesign to better achieve objectives? § N. Mason, T.S. Jayne, and R. Mofya-Mukuka. 2013. “Zambia’s input subsidy programs.” Agricultural Economics http://onlinelibrary.wiley.com/doi/10.1111/agec.12077/abstract Zambia’s ISPs since 2002/03 5 1. 02/03-08/09: Fertilizer Support Programme (FSP) § 400 kg fertilizer, 20 kg hybrid maize seed § Uniform package § Farmers pay 20-50% of market price for inputs 2. 09/10-present: Farmer Input Support Programme (FISP) § Pack size halved § Rice, sorghum, groundnuts, cotton recently introduced (small quantities, few districts) Smart subsidy red flags? 3. 00/01-present: Food Security Pack Programme 3 ¡

10/31/13 ¡ Fertilizer distributed through FSP/FISP Smart subsidy red flags? 200,000 90 6 180,000 80 MT of subsidized fertilizer Fertilizer subsidy rate (%) 160,000 70 140,000 60 120,000 50 100,000 40 80,000 30 60,000 20 40,000 10 20,000 0 0 Subsidized fertilizer (MT) Fertilizer subsidy rate (%) Source: MAL (2012). Note: *2013/14 MT is preliminary estimate. Despite decline in MT, # of beneficiaries is still 900,000. 6 ISPs = large share of gov’t spending Smart subsidy red flags? 7 § 2011 : US$184 million (0.8% of GDP) Other 2% Other 14% FISP FISP 30% 34% FRA 64% FRA 56% % of ag sector Poverty % of total ag sector spending Reduction Program spending Sources: MFNP (2012), IMF (2012) 4 ¡

10/31/13 ¡ Objectives 8 § Improve household and national food security § Raise incomes and reduce poverty § Increase small-scale farmers’ access to inputs § Build private sector capacity Source: MACO (2008); statements by Min. of Ag. And Livestock Selection criteria for FISP beneficiaries 9 1. Capacity to cultivate 0.5 ha of maize 2. Small-scale farmer (cultivate <5 ha) 3. Ability to pay farmer share of inputs costs § US$15 for 10 kg of hybrid maize seed § US$38 for 200 kg of fertilizer (2012/13) 4. Cooperative member 5. Not receiving Food Security Pack  Will come back & evaluate how pro-poor Source: MAL (2012) 5 ¡

10/31/13 ¡ How FISP works – input distribution 10 § Input suppliers selected through national tender process à transport inputs to gov’t FISP depots § Through 2012/13, same 2 companies every year § Representatives from farmer cooperatives pick up inputs from gov’t FISP depots à to beneficiary farmers § Local agro-dealers not involved in FISP Smart subsidy red flags? Source: MAL (2012) What we did 11 1. Use nationally-representative HH survey data to: a. Compare HH characteristics to FISP criteria b. Identify the factors affecting kg of FISP fertilizer received (regression analysis) 2. Synthesize empirical evidence on FISP: 1. Targeting 2. Effects on farmer behavior & economic well-being 3. Political economy 6 ¡

10/31/13 ¡ FISP recipients vs. eligibility criteria (2010/11 agricultural season) 12 § 30% of smallholder HHs eligible for FISP § 30% of smallholder HHs received FISP fertilizer BUT 21% of recipients technically ineligible § Mainly due to not being coop member (53% of ineligible) and/or cultivating 5+ ha (33%) § Targeting performance fairly good based on eligibility criteria Source: Mason et al. (2013a) How well did FISP target poor HHs? (2010/11 agricultural season) 13 % of % of % of $1.25/day total HHs total total poverty Ha below FISP cultivated HHs rate (%) $1.25/day fertilizer 0-0.49 17.0 78.4 17.7 2.5 72.5 77.8 45.1 0.5-0.99 23.6 83.2 26.0 13.0 1-1.99 31.9 80.6 34.1 29.6 2-4.99 23.5 65.8 20.5 41.0 5-9.99 3.3 37.9 1.7 10.7 54.9 10-20 0.6 14.8 0.1 3.2 All HHs 100 75.5 100 100 Source: Mason et al. (2013a) 7 ¡

10/31/13 ¡ Better to target larger farms because they produce more maize per kg? No! 14 14 Farm AP of size (ha) fertilizer (kg/kg) 0-0.99 3.73 1-1.99 3.48 2-4.99 3.52 5-9.99 3.68 10-20 3.46 Also little effect on maize prices : on average, FISP reduces retail prices by < 3% Sources: Burke et al. (2012a), Ricker-Gilbert et al. (2013) Factors affecting receipt of FISP fertilizer 15 § HHs w/ more land and/or livestock get more § HHs closer to towns/roads get more § HHs in const. won by the ruling party get more Sources: Mason et al. (2012, 2013a) 8 ¡

10/31/13 ¡ Rural poverty rates, Zambia: 1996 - 2010 16 90 83 82 80 78 78 80 Rural poverty rate (%) 70 60 50 40 30 20 10 0 1996 1998 2004 2006 2010 Source: CSO (2009, 2011) Input subsidy impacts on smallholder behavior & economic well-being 17 9 ¡

10/31/13 ¡ Crowding out & diversion reduce FISP impact on total fertilizer use 18 § FISP fertilizer crowds out/displaces commercial fertilizer purchases § Crowding out less among poorer HHs § 1/3 of fertilizer intended for FISP diverted & resold on commercial markets § è gov’t injects 1 MT fertilizer for FISP into system à 0.54 MT increase in national fertilizer use § Crowding out also an issue w/ subsidized seed Sources: Mason & Ricker-Gilbert (2013); Mason & Jayne (2013) FISP fertilizer effects on fertilizer use, crop production, & fallow land 19 § Small, positive effects on: § Fertilizer application rate on maize (kg fert./ha maize) § Maize area planted, yields, & kg harvested (1.9 kg/kg) § No stat. sig. effect on area to other crops § Small, negative effect on area under fallow Sources: Mason et al. (2012); Mason et al. (2013a) 10 ¡

10/31/13 ¡ Why such low maize – subsidized fertilizer response rate (1.9 kg/kg)? 20 1. Crowding out (Mason & Jayne, 2013) 2. Late delivery (Xu et al., 2009) § to 20-30% of beneficiaries § Late delivery halves maize-N response rates 3. High soil acidity (Burke et al., 2012b) § > 90%+ of maize fields have pH < 5.5 § Maize-fertilizer response rates 50-67% lower on highly acidic soils FISP seed generally improves indicators of economic well-being 21 § FISP subsidized hybrid maize seed  § é maize production (106 kg) § é HH income (1.1%, US$32) § ê prob. that below US$2/day poverty line (0.7 p.p.) § ê severity of poverty (poverty gap 2 , 0.8 p.p.) § ê relative deprivation (income “inequality”, 0.4%) Source: Mason & Smale (2013). Changes are per 10 kg of subsidized hybrid maize seed. 11 ¡

10/31/13 ¡ Do FISP benefits outweigh the costs? 22 § Costs = farmers costs + net costs to government § Benefit-cost ratios : § Profitable if BCR > 1 è (Benefits > Costs) § Fertilizer (Jayne et al., 2013) ¨ 2006/07-2010/11 BCR = 0.90 § Seed (Mason & Smale, 2013) ¨ BCR generally > 1 Does FISP win votes for ruling party? 23 § No , but ruling party’s share of votes increases with reductions in: § Poverty § Income inequality § Unemployment § 2006 & 2011 elections Sources: Mason et al. (2013b) 12 ¡