Start of Mini-Series: Randomized Controlled Trials as Tools to Understand Public Policy Decisions
(I) “Encouraging the Adoption of Agroforestry: A Case Study in Eastern Province, Zambia” (Kelsey Jack, Tufts; Paulina Oliva, UCSB; Elizabeth Walker, Harvard; Samuel Bell, Cornell)
In a quest to understand whether cashflow constraints or the absence of short-term financial benefits impede the uptake of agroforestry practices of farmers in Zambia, the research team surrounding Kelsey Jack from Tufts University in collaboration with Trees on Farms Programme conducted a randomized controlled trial in which they analyzed the effect of different subsidy and incentive schemes on farmer adoption. The subsidies took the form of different input-price reductions (from free hand-out of tree seedlings to having farmers pay their full market price) while the incentive payments were outcome-based and rewarded farmers, whose tree survival rate surpassed 70% after one season (Nov ’11-Dec ’12).
All farmers underwent training on planting and tree care before they could decide to opt into the programme for one year and receive the seedlings or not.Randomisation was implemented through a lottery (scratch cards that listed the individual rewards – randomization on the individual level) and a computer-orchestrated algorithm, that allocated input-prices at the farmer group-level (in total, 1,300 farmers are trained in 125 farm groups). Performance-incentives were revealed alternatingly either before or after people opted into the program. Data were collected at baseline, at periodic intervals at which every fifth participant was visited to observe planting progress, for a ‘planting survey’ to register the number of trees planted and at end-line to track the tree survival rate.
According to the researchers their data reveal that:
a) Increases in input-costs result in lower programme take-up, but have no impact on tree survival once farmers decided to join the programme. (1 USD decrease in input-cost lead on average to a 13% increase in programme uptake.)
b) Programme adoption and tree survival rate are positively influenced by higher cash incentives. (For very low to low input-prices incentives seem to have no effect on uptake, whereas for higher input prices incentives drive participation. A 1 USD incentive raise yields a 2% increase in the one-year tree survival rate. Interestingly, the effect of incentives on tree survivals is almost identical for those who learned of the incentive-scheme before and after opting in.)
c) When group-level controls are introduced incentives continue to matter within groups with respect to programme uptake and tree survival, but group dynamics impact these outcomes as well.
d) Overall higher incentive payments within one group positively influence the two key outcomes if the individual farmer’s incentive level is controlled for.
e) If monitored, around ten trees more are planted when the individual reward is controlled for.
f) Individual characteristics matter: households with more members, those with a female head and those which are more risk-loving have a higher tendency to join the programme; “larger households” and “having bought fertilizer the previous year” are statistically significant determinants of earning the cash reward after one season.
The researchers derive from their results that subsidies of input-prices in effect enhance programme uptake without having the feared adverse effect of leading to irresponsible behaviour and slacking on the side of those benefitting from the program. Beyond, the success of outcome-based financial incentives allows agroforestry project managers to weigh the additional fixed cost of one participating famer against higher incentive payments – both of which are effective in having more trees planted.
References (recommended for expanded account and graphical depiction of the results):
Innovations for Poverty Action 2013, Encouraging the Adoption of Agroforestry: A Case Study in Eastern Province, Zambia, viewed 21 September 2014, <http://www.poverty-action.org/sites/default/files/in-depth_research_results_2013.pdf>
Jack, K 2012, Encouraging the adoption of agroforestry, viewed 21 September 2014, <http://www.theigc.org/wp-content/uploads/2014/08/Kelsey-Jack-Zambia-GW2012.pdf>