Up-take of Contract Farming and Refugee-Host interactions in Uganda

How to cite

Belay, Dagim; Adong, Annet; Tellez, Juan F.. 2026. "Up-take of Contract Farming and Refugee-Host interactions in Uganda." AEA RCT Registry. AEARCTR-0017798. https://doi.org/10.1257/rct.17798-1.0

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

Dagim Belay* , Annet Adong , Juan F. Tellez

* Primary investigator

Registry ID
AEARCTR-0017798
DOI
https://doi.org/10.1257/rct.17798-1.0
First registered
Last updated
Status
In development
Trial dates
2026-02-25 – 2027-12-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/17798

Abstract

Refugees face significant barriers to agricultural livelihoods, including limited access to land, inputs, credit, and markets, as well as exposure to climatic risks and host community hostilities. These constraints hinder refugees’ ability to build resilient and self-sustaining livelihoods. This study examines whether a contract farming model can help alleviate these barriers by reducing upfront liquidity constraints, facilitating access to land and inputs, and easing intergroup tensions with host communities. We experimentally test the uptake of contracts that vary in the share of production costs deferred until harvest. The study assesses contract acceptability and default, identifies key implementation frictions, examines host landlords’ beliefs toward refugee livelihood support, and estimates delivery costs to inform the design and scalability of a larger randomized controlled trial.

Intervention

The intervention offers refugee households a one-season agricultural land rental contract with host community landlords. The contract aims to reduce barriers to refugees' agricultural participation by easing liquidity constraints and varying interaction with host landowners. All participants receive an offer to rent land under one of several contract variants. Two contract features are randomized at the household level. First, payment timing varies the proportion of total costs paid upfront: 100% (control), 80%, or 50%, with remaining payments deferred until after harvest. Second, landowner contact varies whether refugees farm independently (control) or cultivate alongside the host landowner.

These features generate a 2 × 3 factorial design with six treatment conditions. Contracts are randomly assigned to 1,000 refugee households within village agents.

Experimental design

The study evaluates an agricultural land rental contract offered to refugee households using an individually randomized factorial design. From a universe of eligible refugee households in the settlement, a sample of 1,000 households is selected to participate. All sampled households are offered a standardized one-season land rental contract with host community landlords.

Two contract features are randomized at the household level. First, payment timing varies the share of total contract costs paid upfront versus deferred until after harvest (100%, 80%, or 50% upfront). Second, the degree of interaction with the host landowner varies between farming independently and farming alongside the landowner. These features generate six experimental conditions in a 2 × 3 design. Randomization is conducted at the household level within local recruitment units to ensure balance.

Design details

Randomization method
Randomization done in office by a computer
Randomization unit
household
Clusters
No clustering
Observations
1000 refugee households
Arms
The study includes 1,000 refugee households assigned under a 2 × 3 factorial design with equal allocation across treatment arms. Approximately one-third of households (about 334 each) are assigned to the 100% upfront payment and 80% upfront payment conditions, with the remaining 332 assigned to the 50% upfront payment condition. Independently, half of the sample (about 500 households) is assigned to farm with no contact with the host landowner and half to farm alongside the landowner. This yields approximately 166–167 households in each of the six experimental cells, with minor variation due to integer constraints.
Primary outcomes
Contract uptake and use: willingness to accept/pay, actual upfront payment for contracted inputs; size of land cultivated under the contract
Secondary outcomes
- Contract compliance and default: post-harvest payment completion, arrears amount - Agricultural productivity: yield, sales - Market participation - Intergroup interaction: refugee-host attitudes and trust, interaction frequency, disputes/conflict - Psychological well-being

Cited by

No citation data available for this registration yet.

Keywords