Farmer soils information for decision making

How to cite

Harou, Aurelie; Palm, Cheryl; Madajewicz, Malgosia; Michelson, Hope. 2020. "Farmer soils information for decision making." AEA RCT Registry. AEARCTR-0005470. https://doi.org/10.1257/rct.5470-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Aurelie Harou* , Cheryl Palm , Malgosia Madajewicz , Hope Michelson

* Primary investigator

Registry ID
AEARCTR-0005470
DOI
https://doi.org/10.1257/rct.5470-1.0
First registered
Last updated
Status
Completed
Trial dates
2014-08-01 – 2019-09-30
Countries
Tanzania, United Republic of (Morogoro)
Registry record
http://www.socialscienceregistry.org/trials/5470

Abstract

Low fertilizer use in sub-Saharan Africa contributes to low crop yields and poverty. Extension services advise the use of fertilizers but may not recommend fertilizer types that address prevailing soil deficiencies, potentially resulting in the perception that fertilizer is not profitable. This randomized control trial tests how fertilizer recommendations based on farm-level soil tests interact with the ability to finance fertilizer to influence fertilizer application and crop yields.

Experimental design

We test how fertilizer recommendations based on farm-level soil tests interact with the ability to finance fertilizer to influence fertilizer application and crop yields. We randomly select 50 villages in Morogoro Rural and we assign 20 randomly chosen villages to the treatment group and 30 to the control group. In each of the 20 treatment villages, we randomly select 40 farmers to participate in the study, and we assign 10 farmers randomly selected from these 40 to each of the three treatment arms and to the control group. In each of the 30 control villages, we randomly select 10 farmers. The treatment group includes 600 farmers in 20 treatment villages. The control group includes 500 farmers: 200 control farmers in treatment villages and 300 farmers in control villages. The treatments include 3 arms: plot-specific recommendations, an agricultural input voucher, plot-specific recommendations and an agricultural input voucher.

Design details

Randomization method
Randomization done in office by a random number generator on computer.
Randomization unit
Farming household
Clusters
50 villages
Observations
1,100 farming households
Arms
Recommendation treatment: 20 villages, 200 farmers Voucher treatment: 20 villages, 200 farmers Recommendation and voucher treatment: 20 villages, 200 farmers Control: 50 villages, 500 farmers
Primary outcomes
Fertilizer use (kg/acre) (farmer reported), fertilizer purchases (kg) (agro-input dealer reported), and maize yields (kg/acre) (farmer reported).

Cited by

No citation data available for this registration yet.

Keywords