The Impact of Religion on Conservation: Evidence from Burundi

Last registered on October 03, 2026

View Trial History

Pre-Trial

Trial Information

General Information

Title

The Impact of Religion on Conservation: Evidence from Burundi

RCT ID

AEARCTR-0016548

Initial registration date

August 09, 2025

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

August 11, 2025, 10:09 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Last updated

October 03, 2026, 11:10 AM EDT

Last updated is the most recent time when changes to the trial's registration were published.

Locations

Country

Burundi

Region

Muramvya, Gitega and Mwaro.

Primary Investigator

Name

Pedro Naso

Affiliation

Swedish University of Agricultural Sciences

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Jean Budurha

PI Affiliation

Universite Evangelique en Afrique

Contact Investigator

PI Name

Franck Irakoze

PI Affiliation

University of Burundi

Contact Investigator

Additional Trial Information

Status

Completed

Start date

2025-09-01

End date

2025-11-26

Keywords

Behavior, Environment & Energy, Lab

Additional Keywords

Religion, Conservation

JEL code(s)

C93, D91, Q54, Q56, Z12

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

We study whether religion can promote conservation of natural resources. We run a field experiment in rural Burundi with forest users—forest owners and workers such as charcoal producers. Participants are randomly assigned to hear a religious stewardship message, a secular civic-duty message from Burundi’s environmental law, or a neutral message. They then play a public-goods game and decide how much to contribute to forest conservation. Contributions are real: 80% of the conservation account is donated to the office for the protection of forests in Burundi. We compare contributions across groups to test whether religious framing increases cooperation. We also examine heterogeneity by religiosity, affiliation, forest income, and pro-environmental perceptions. The study may inform outreach in places facing deforestation and other environmental problems. If the religious message performs best, this would indicate scope for environmental protection agencies and NGOs to work with clergy to promote conservation through co-designed, voluntary messaging. Beyond Burundi, the results can inform other resource-dependent, highly religious settings on when faith-based framing can complement secular outreach.

External Link(s)

Registration Citation

Citation

Budurha, Jean, Franck Irakoze and Pedro Naso. 2026. "The Impact of Religion on Conservation: Evidence from Burundi." AEA RCT Registry. October 03. https://doi.org/10.1257/rct.16548-1.2

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

We will prime our participants with three types of messages. Participants are randomly assigned to hear a religious stewardship message, a secular civic-duty message from Burundi’s environmental law, or a neutral message. They then play a public-goods game and decide how much to contribute to forest conservation.

Intervention (Hidden)

Intervention Start Date

2025-09-01

Intervention End Date

2025-11-25

Primary Outcomes

Primary Outcomes (end points)

The primary outcome is participants' contribution to a conservation account, to protect forests in Burundi.

Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)

We will heterogeneity analsys to examine whether the primes are more or less effective according to participants' religiosity, religious affiliation, forest income, and pro-environmental perceptions.

Secondary Outcomes (explanation)

Experimental Design

We will run a simple randomized controlled trial with three groups. Participants are randomly assigned to: (1) a religious prime, (2) a secular prime, or (3) a neutral prime. All primes are short messages of similar length, read aloud in a neutral tone by enumerators. After the message, participants play a standard public-goods game. They receive an endowment and choose how much to keep in a private account and how much to place in a conservation account. If they contribute to the conservation account, they have a 20% chance of receiving the collective payoffs from that account, which simulate the positive impact of conservation for forest users. Eighty percent of their raw contributions (without the multiplier) is donated to OBPE, the forest agency of Burundi, to support forest protection. Participants make decisions individually and in private, and groups are formed randomly within treatment arms afterward.

Participants will be adult forest users in Muramvya, Mwaro, and Gitega. We include (i) forest owners who earn income from their forests and (ii) workers providing forest-related services (e.g., charcoal producers, transporters, sellers). We will request from OBPE an up-to-date list of forest users in the three provinces. From this list, we will draw a simple random sample, sampling proportionally by commune (the administrative unit below the province). If the OBPE list is incomplete, we will supplement it with a short listing exercise with local authorities and producer groups to identify additional eligible forest users, then randomly sample from the supplemented list.

Experimental Design Details

Randomization Method

Randomization done in office by a computer.

Randomization Unit

Individuals

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

We aim for a sample of about 760 participants. We will randomly assign them in a 1:1:1 ratio to T1 (religious prime), T2 (secular prime), or Control, targeting about 253 per group. With equal group sizes, this yields 80% power (two-sided α=0.05) to detect an average treatment effect of 0.25 SD in pairwise comparisons.

Sample size: planned number of observations

760

Sample size (or number of clusters) by treatment arms

253 individuals.

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

Supporting Documents and Materials