Decentralizing Forest Governance: A Cluster Randomized Experiment in the Himalayas

How to cite

Fleischman, Forrest; Rana, Pushpendra; Choksi, Pooja; Fischer, Harry; Coleman, Eric; Khanyari, Munib; Suryawanshi, Kulbhushan; Sills, Erin; Madhok, Raahil. 2025. "Decentralizing Forest Governance: A Cluster Randomized Experiment in the Himalayas." AEA RCT Registry. AEARCTR-0016090. https://doi.org/10.1257/rct.16090-1.2

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

Forrest Fleischman* , Pushpendra Rana , Pooja Choksi , Harry Fischer , Eric Coleman , Munib Khanyari , Kulbhushan Suryawanshi , Erin Sills , Raahil Madhok

* Primary investigator

Registry ID
AEARCTR-0016090
DOI
https://doi.org/10.1257/rct.16090-1.2
First registered
Last updated
Status
In development
Trial dates
2025-06-02 – 2028-06-01
Countries
Private
JEL codes
Q01, Q23, Q28
Registry record
http://www.socialscienceregistry.org/trials/16090

Abstract

This project will evaluate two governance regimes for community-oriented forest restoration. We will conduct a cluster-randomized trial in Himachal Pradesh, India, where Mahila Mandals (women's groups) are assigned to either (i) a standard participatory model where the Forest Department retains control but consults communities, or (ii) a local autonomy model where Mahila Mandals receive funds and planting stock and make independent decisions. Key outcomes include forest cover (high-resolution satellite imagery), livelihoods (household surveys), institutional engagement, and ecological indicators (ecological surveys). The study provided actionable evidence on the best way to engage communities in forest restoration.

Intervention

About 300 women's groups (Mahila Mandals) will be receiving tree plantation sites under the The Rajiv Gandhi Van Samvardhan Yojana (2025) program, a scheme designed by the Forest Department ‘to strengthen conservation efforts and enhance afforestation by engaging Mahila Mandals...’. A treatment group of 150 Mahila Mandals will receive full decision-making autonomy and planting stock to manage forest restoration activities independently, while a control group of 150 Mandals will follow the status quo regime where the Forest Department retains primary decision-making authority but may consult Mahila Mandas in forest management.

Experimental design

The study uses a cluster-randomized controlled trial design, with 300 Mahila Mandals randomly assigned to either a “local autonomy” governance treatment (n = 150) or the status quo participatory model (n = 150).

Design details

Randomization method
Randomization was done using Stata. Mahila Mandals were grouped by their forest type identifiers before random assignment. Within each stratum, Mandals were randomly assigned to treatment or control using a reproducible random number generator to enable replicability. For strata with uneven number of clusters, "misfits" were randomly assigned to a treatment or control group. Out of the total 300 Mahila Mandals, 148 were allocated to the treatment group through this procedure.
Randomization unit
Mahila Mandal (women's group)
Clusters
300 Mahila Mandals
Observations
9000 households (approximately 30 households in a Mahila Mandal)
Arms
150 treatment Mahila Mandals, and 150 control Mandals. With approximately 30 households per Mandal, this makes about 3500 treatment and control observations (households).
Primary outcomes
Remotely-sensed forest cover, household livelihood outcomes
Secondary outcomes
institutional outcomes, ecological (plantation site level) outcomes.

Cited by

No citation data available for this registration yet.

Keywords