Testing Agricultural Technology in Northern Ghana Introduce department and team -We have 2 mandates is to raise the profile of IPA and protect it’s image and reputation. We want to be know as the best organization in our field. Secondly, we raise non-project funds. Money that can be used to strengthen IPA as an organization. -Small team of 5 Federica Di Battista IPA, Research Coordinator Joanna Van Asselt IFPRI, Senior Research Assistant

Discover and promote effective solutions to global poverty problems Who are we? Innovation for Poverty Action Discover and promote effective solutions to global poverty problems

The problem

The solution

Three projects for agricultural development: IPA Ghana Three projects for agricultural development: Disseminating Innovative Resources and Technologies for Smallholders (DIRTS) Testing Agricultural Technology (TAT) Conservation Agriculture Evaluation (CAGE) Established in 2005 Two offices: Accra, Tamale Around 50 long term staff and 100 short term staff Education, health, micro-finance and agriculture.

Testing Agricultural Technologies in Northern Ghana (TAT) STUDY AREA: 3 northern regions of Ghana: Upper East Upper West Northern RESEARCHERS: Christopher Udry (North Western University) Dean Karlan (North Western University) Shashidhara Kolavalli (IFPRI) TIME LINE: Phase I: 2015 Phase II: 2016 PARTNERS: Savanna Agricultural Research Institute (SARI) International Food Policy Research Institute (IFPRI)

Low adoption of improved inputs: TAT – Introduction Sub-Saharan Africa in the last 40 yrs (Akram-Lodhi, 2008) : Growth of the Agriculture sector: 20% Increase in the amount of land cultivated: 80% Increase in productivity: 20% Low adoption of improved inputs: In Northern Ghana only 20% farmers adopt improved seed varieties The share of harvested cereals from improved seeds varieties: Sub-Saharan Africa: 24% South Asia: 77% East Asia: 85%

TAT – Needs assessment Inadequate availability Heterogeneity in the localized returns to technologies Improving information about the performance of new seeds in a variety of contexts in Northern Ghana Studying the means by which farmers learn about and test new technologies Inadequate availability of improved inputs: The choice of available certified seeds of varieties that have been proven to outperform farmers’ traditional choice is often limited in the local retailers; Heterogeneity in the localized returns to various technologies: Even when proven varieties become available, farmers may not have sufficient evidence to believe that a particular technology will prove profitable in their environment; this can also reinforce the supply-side issue, as dealers could be hesitant to supply products that do not have demonstrated demand on the part of farmer.

We tested the performance of 5 seed varieties- TAT – Phase I We tested the performance of 5 seed varieties- Pioneer PAN53 Mamaba Sanzal Sima Obaatanpa

TAT – Phase I In 10 districts in the 3 northern regions of Ghana. Two types of trial plots per district: 1 Mother trial 4 Baby trial on the fields of a lead farmer (i.e. on-station trial) included four reps of each of the five varieties, closely supervised by either a SARI scientist or an Agricultural Extension Agent (AEA) Additionally, four farmers were each chosen to oversee a baby (on-farm) trial, testing their own variety, as well as one foreign hybrid and one other variety. These trials were meant to more closely mimic the level of attention a plot would receive on a farmer’s own farm, and SARI Scientists and AEAs were instructed to visit the plots every week, and to also assist farmers for key stages in the production process.

TAT – Phase I

Within-district correlation TAT – Phase I N Mean Yield (t/ha) Standard deviation Within-district correlation Pioneer 57 4.99 1.16 0.20 PAN 53 61 3.97 1.43 0.48 Obaatanpa 78 3.19 1.27 0.21 Sanzal Sima 60 3.13 1.34 0.50 Mamaba 58 2.99 1.49 0.73

TAT – Phase I

TAT – Phase II Community sensitization in all communities that took part to the 2015 trials; Interested farmers were given the opportunity to buy; Pioneer subsidized starter pack; Delivery of the starter pack; Data collection: 302 farmers interviewed With help of the AEA in each area, a meeting was convened at the community level and on average 20-30 farmers participated. The meeting began with a group discussion on the results of the 2015 trials, after which the team explained that farmers could purchase starter packs of Pioneer and Obatampa, if they were interested in testing these on their own land. The starter packs included: - 1 kg of certified seed (sufficient to set up a 20x20m demo field) - 11 kg of NPK fertilizer - 5.5 kg of Sulfate of Ammonia fertilizer - 20m rope (to measure the demo plot) - A simple instruction guid

Pioneer yields from farmers’ plots TAT – Phase II Freq. Yield (t/ha) - farmer Yield (t/ha) - trial Bawku Municipal 28 2.11 4.91 Bolga Municipal 2.13 5.12 Kassena-Nankana 33 1.83 3.85 Nadowli-Kaleo 27 1.70 4.69 Sagnerigu 52 1.80 - Savelugu-Nanton 31 1.50 Sissala East 29 1.68 6.09 Tolon 1.37 5.11 Wa Municipal 53 1.31 6.07 West Gonja 62 1.38 4.49 West Mamprusi 2.45 4.68 Yendi Municipal 18 1.16 4.72 Total 441 1.66 5.01 Pioneer yields from farmers’ plots

Other hybrids include Mamaba, Pan 53 and, non-IPA sourced Pioneer TAT – Phase II Other hybrids include Mamaba, Pan 53 and, non-IPA sourced Pioneer

TAT – Phase II Freq Seeding rate Bawku Municipal 28 3.88 Bolga Municipal 2.58 Kassena-Nankana 33 2.96 Nadowli-Kaleo 27 2.22 Sagnerigu 52 3.23 Savelugu-Nanton 31 3.45 Sissala East 29 2.94 Tolon 4.15 Wa Municipal 53 3.67 West Gonja 62 4.04 West Mamprusi 2.25 Yendi Municipal 18 3.07 Total 441 3.34 The ideal seeding rate for Pioneer should be 10 kg of seeds per 1 acre of land cultivated

Are farmers over-reporting land? TAT – Phase II Are farmers over-reporting land? Freq Mean Std. Dev. Land cultivated - reported 6,130 2.22 1.64 Land cultivated - measured 2.25 1.69 Difference -0.02** 0.77 Source: DIRTS – Plot Measurement Survey, preliminary results

TAT – Conclusions Pioneer and PAN53 are high-yield profitable technologies; Neither of the other varieties tested, Mamaba or Sanzal Sima, performed consistently higher than the farmers’ variety, Obaatanpa; Due to inappropriate seeding rate, the average yield that farmers are obtaining from Pioneer is lower than that from demo trials; Nonetheless, Pioneer still outperformed all other varieties, especially in the Northern region, where Pioneer yields were more than double the yields of other varieties. -Despite its high price, the results suggest that Pioneer would have been a profitable investment on all of the plots studied. The varieties performed consistently well in all districts. -Because it showed significant variation from district to district, Mamaba may be a viable technology for increasing yields in the districts where it performed especially well

TAT – Conclusions Farmers expressed significant interest in purchasing Pioneer next season; This suggests that Pioneer seeds should be made available to farmers in the North. At GHS 32 farmers stated that they would plant 48% of their maize land to Pioneer. At GHS 16, they would increase this area to 62% of their land. If the seeds were free, they would plant pioneer on 78% of their land. This differed little between regions.

Questions? comms@poverty-action.org