1. Trends and Patterns in Fertilizer Use by Smallholder Farmers in Kenya, 1997-2007 Joshua Ariga, T.S. Jayne, Betty Kibaara, and J.K. Nyoro Paper presented at the Egerton University Tegemeo Institute Agricultural Policy Conference, 17 September 2008, Nairobi, Kenya

2. 2 Kenya (42.8, +40%) between 2007 and 1997 Swaziland (30.5, -40%) Malawi (30.8, +9%) Zimbabwe (48.3, +9%) > 25 kg/ha Uganda (0.6, +237%) Rwanda (1.8, +89%) Mozambique (3.2, +142%) Ghana (3.6, +68%) Chad (4.3, +93%) Cameroon (5.9, +77%) Togo (7.0, +30%) Cote d’Ivoire (11.8, +53%) Botswana (11.8, +294%) Senegal (13.2, +67%) Ethiopia (14.4, +71%) Benin (17.6, +76%) Lesotho (23.2, +35%) DRC (0.5, -47%) Angola (0.7, -69%) Niger (0.9, +5%) Guinea (2.0, -4%) Burundi (2.3, -6%) Madagascar (2.9, -8%) Mauritania (4.0, -64%) Tanzania (4.8, -47%) Gambia (5.2, +15%) Nigeria (5.6, -73%) Burkina Faso (5.9, -28%) Zambia (8.4, -34%) Mali (9.0, +7%) < 25 kg/ha > +30%< +30% % growth in fertilizer use intensity (kg/ha cultivated) (mean 1996-2002 / mean 1990-95) Intensity of fertilizer use (1996-2002)

3. 3 Kenya fertilizer use, 1990-2008

4. 4 Objectives: 1.Trends in fertilizer use on maize 2.Factors driving the increase in fertilizer use, 1997-2007 3.Household characteristics associated with fertilizer use 4.Impact on maize yields 5.Policy implications in light of higher food and fertilizer prices

5. 5 Trends in Fertilizer Use on Maize Objective 1

6. 6 % of Small-scale Farmers Using Fertilizer on Maize Agro-regional zone19961997200020042007 % of households using fertilizer on maize Coastal Lowlands003414 Eastern Lowlands2127254743 Western Lowlands215513 Western Transitional3941707181 High-Pot. Maize Zone8584908791 Western Highlands817591 95 Central Highlands8890 9193 Marginal Rain Shadow66121116 Total Sample5658646670

7. 7 Fertilizer Dose Rate (kgs/acre) on maize Agro-regional zone1997200020042007 Dose rate (kgs/acre) on fertilized maize fields Coastal Lowlands11537 Eastern Lowlands10181516 Western Lowlands24141012 Western Transitional54486271 High-Pot. Maize Zone65677475 Western Highlands31364647 Central Highlands6864 58 Marginal Rain Shadow121543 National sample56556059

8. 8 Factors driving the increase in fertilizer use, 1997-2007 Objective 2

9. 9 4 Reasons for the Upsurge in Fertilizer Use in Kenya 1.GoK has maintained a stable fertilizer policy stance since 1990  Eliminated import licensing quotas  Eliminated foreign exchange controls  Eliminated retail price controls  From 1990 to 2007, no market uncertainties introduced by large-scale subsidy programs

10. 10 Farmer fertilizer purchases, Malawi

11. 11 4 Reasons for the Upsurge in Fertilizer Use in Kenya 2.In response to stable input policy environment, private sector investment in fertilizer distribution expanded rapidly  10-11 importers  500 wholesalers  8,000 retailers

12. 12 4 Reasons for the Upsurge in Fertilizer Use in Kenya 3.In response to expansion of input stockists, small farmers’ are now much closer to fertilizer retailers  1997: 7.4kms  2000: 5.6kms  2004: 3.7kms  2007: 3.2kms

13. 13 Reasons for the Upsurge in Fertilizer Use in Kenya 4.Greater competition among importers and wholesalers has led to declining fertilizer marketing costs

14. 14 Price of DAP (Di-Ammonium Phosphate) in Mombasa and Nakuru (nominal Shillings per 50kg bag)

15. 15 Price of DAP (Di-Ammonium Phosphate) in Mombasa and Nakuru (constant 2007 Shillings per 50kg bag)

16. 16 Why have real fertilizer marketing margins declined in Kenya? 1.Greater competition has led to lower margins 2.Emergence of brokerage services for exploiting opportunities for cheaper backhaul transport, e.g., linking upcountry fertilizer supply with trucks transporting cargo from Rwanda and Congo to the port of Mombasa; 3.private importers are increasingly using international partners to source credit at lower interest and financing costs than are available in the domestic economy 4.mergers between local and international firms in which knowledge and economies of scope are being passed onto local firms to achieve cost savings in local distribution (e.g., Mea partnering with CONAGRA)

17. 17 Objective 3 Household characteristics associated with fertilizer use

18. 18 0 20 40 60 80 100 010203040 Total acres Scatterlowess Table 1a: Zone 1 Fertilizer rate vs. Household Acres 0 50 100 150 200 010203040 Total acres Scatterlowess Table 1a: Zone 2 Fertilizer rate vs. Household Acres 0 50 100 150 200 Fertilizer rate per acre 010203040 Total acres Scatterlowess Table 1a: Zone 3 Fertilizer rate vs. Household Acres 0 100 200 300 Fertilizer rate per acre 010203040 Total acres Scatterlowess Table 1a: Zone 4 Fertilizer rate vs. Household Acres Relationship between fertilizer use per acre and wealth

19. 19 Relationship between household wealth and fertilizer use Household Wealth Quartiles 1 (poorest) 234 (Highest) Assets (Kenya Shillings)2,98212,10625,633166,919 Dose Rate (kgs/acre) users only 59606668

20. 20 % of Small-scale Farmers Using Fertilizer on Maize Agro-regional zone19961997200020042007 % of households using fertilizer on maize Coastal Lowlands003414 Eastern Lowlands2127254743 Western Lowlands215513 Western Transitional3941707181 High-Pot. Maize Zone8584908791 Western Highlands817591 95 Central Highlands8890 9193 Marginal Rain Shadow66121116 Total Sample5658646670

21. 21 Objective 4 Impact on maize yields

22. 22 Not counting other crops grown on intercropped maize fields

23. 23 Yield includes quantity of other crops produced on intercropped fields converted to maize equivalents by price ratios

24. 24 Objective 5 Implications for policy under current world price conditions

25. 25 Profitability of using fertilizer: Farm-gate Maize Price Δkg maize ---------------------------------- * ------------ Farm-gate Fertilizer Price Δkg fert

26. 26 1.Reduce costs of supplying fertilizer to farm gate Port costs at Mombasa Improve rail / road infrastructure 2.Promote viable farm extension / service provision to raise efficiency of fertilizer use What about input subsidies?

27. 27 Summary of research evidence about fertilizer subsidies in Africa:  can help to raise production, but little sustained benefit after subsidies are withdrawn  (e.g., Zambia, Malawi)  Benefits tend to be disproportionately captured by better-off farmers, unless near universal coverage  Costly – foregone payoffs from alternative public investments

28. 28 Zambia Total Income AssetsLandholding size Fertilizer source: ‘000 kwacha per capita ha per capita Households not acquiring fertilizer: 266173.15 Source: Govereh et al, 2006

29. 29 Zambia Total Income AssetsLandholding size Fertilizer source: ‘000 kwacha per capita ha per capita Households not acquiring fertilizer: 266173.15 Cash purchases from private retailers: 774342.20 Source: Govereh et al, 2006

30. 30 Zambia Total Income AssetsLandholding size Fertilizer source: ‘000 kwacha per capita ha per capita Households not acquiring fertilizer: 266173.15 Cash purchases from private retailers: 774342.20 Government Fertilizer Support Program (50% subsidy) 804425.23 Source: Govereh et al, 2006

31. 31 Budget allocation to Agricultural Sector in Zambia: ZMK465 million in 2005

32. 32 IFPRI review of rate of return studies: Returns SubsidiesNegative – 12% Investments - research & extension35% to 70% - roads20% to 30% - education15% to 25% - communications10% to 15% - irrigation10% to 15% If we believe these findings, they have major implications

33. 33 If the decision has already been made to provide input subsidies: Four insights (from experience in Malawi and Zambia):

34. 34 Insight #1: 1.Targeted input voucher program less likely to undercut commercial input distribution system

35. 35 Insight #2: Ensure that input subsidies are pro-poor by targeting the poorest farmers: –Will generate greatest food security and poverty reduction impact –No evidence that fertilizer use is more efficient on large farms than small farms –Achieves more maize output per unit of subsidized fertilizer distributed (evidence from Malawi and Zambia)

36. 36 Insight #3: If subsidy programs are to be implemented, design them in ways that involve the full range of private importers, wholesalers, and retailers. Providing tenders to only 2-3 firms can: –entrench their position in the market –cause other firms to cease making investments in the system or drop out altogether –lead to a more concentrated input marketing system and restricted competition when the input subsidy program ends

37. 37 Insight #4: Recommend reduced rates of fertilizer application – 100kg per acre is certainly not optimal in most parts of Kenya.

38. 38 Summary of Main Findings 1.nationwide, the % of farmers using fertilizer on maize has increased from 56% in 1996 to 70% in 2007 2.Fertilizer dose rates on maize (maize fields receiving fertilizer) have increased only slightly, from 56kg/acre in 1997 to 59kg/acre in 2007 3.Fertilizer use has increased especially rapidly on the intercropped fields, and less so on monocropped fields 4.The dominant factor influencing smallholder households’ decisions to use fertilizer on maize is location: Over 90% of smallholders use fertilizer on maize in three of the zones surveyed: the High Potential Maize Zone; Western Highlands, and Central Highlands. Less than 30% use fertilizer on maize in Coastal Lowlands, Marginal Rain Shadow.

39. 39 Summary of Main Findings 5.Total area under maize has remained largely constant over the decade 6.maize yields increased by 20% between 1997- 2007 period, which is correlated with the rise in fertilizer use. 7.Paying attention to the different types of maize production technologies and maize cultivation techniques is important to carefully control for confounding factors when examining trends in maize yields in Kenya

40. Thank you http://www.aec.msu.edu/fs2/

41. 41 Table 1: Sampled districts in agro-ecological zones Agro-ecological zoneDistrictsNo. of households Coastal LowlandsKilifi, Kwale75 Eastern LowlandsMachakos, Mwingi, Makueni, Kitui, Taita-Taveta145 Western LowlandsKisumu, Siaya153 Western Transitional Bungoma (lower elevation), Kakamega (lower elevation) 148 Western HighlandsVihiga, Kisii331 Central HighlandsNyeri, Muranga, Meru129 High-Potential Maize Zone Kakamega (upper elevation), Bungoma (upper elevation) Trans Nzoia, Uasin Gishu, Bomet, Nakuru, Narok 242 Marginal Rain ShadowLaikipia37 Overall sample1260