Community-Based Forestry and the Modernization of Smallholder Farming: The case of Ethiopia

How to cite

Bulte, Erwin; Kahsay, Goytom. 2023. "Community-Based Forestry and the Modernization of Smallholder Farming: The case of Ethiopia." AEA RCT Registry. AEARCTR-0012014. https://doi.org/10.1257/rct.12014-1.0

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

Erwin Bulte* , Goytom Kahsay

* Primary investigator

Registry ID
AEARCTR-0012014
DOI
https://doi.org/10.1257/rct.12014-1.0
First registered
Last updated
Status
In development
Trial dates
2023-09-15 – 2027-07-01
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/12014

Abstract

This proposal focuses on the nexus of two challenges for Ethiopian policy makers: the unproductive nature of smallholder farming and the rapid deterioration of natural forest stocks. We will evaluate the impacts of a community-based forestry intervention (devolving extraction rights to local communities) on (i) the modernization of farming, (ii) rural livelihoods and (iii) forest conservation. Community-based forestry is expected to have a transformative effect on smallholder farming by relaxing liquidity and risk constraints, providing key inputs for high-value agriculture, and facilitating collective action (enabling groups of farmers to benefit from economies of scale in trading, storage and transacting).

Intervention

The main intervention is the creation of so-called forest user groups (FUGs). The right to manage a specific forest block and extract forest resources there is delegated to this newly created communtity-based organization.

The RCT has 4 experimental arms: one control arm and 3 arms with FUGs but different governance modalities..

• Business-as-usual forest management: This is our control group villages in which forest user groups will not be established and forest resources are under government management.

The following three treatment arms consists of villages in which FUGs will be established based on alternative approaches to select FUG leadership.

• Status quo leadership: according to the OFWE sample bylaw, members of the executive committee (EC) are “elected” by group members during a public voting event. Kahsay and Bulte (2021) show that EC members are often representatives of the local elite—higher incomes, greater wealth, better educated and connected, and involved in positions of (formal) authority. Business-as-usual selection of leadership implies that village elites “volunteer” for positions and are subsequently “supported” by other members.

• Anonymized voting: a small recent literature suggests that “local knowledge within local communities” may be leveraged to improve allocative outcomes (e.g., Casey et al., 2022, Hussam et al., 2022). We will explore possibilities to create conditions under which the most suitable and capable local candidates are identified and selected (details to be decided after follow-up pilot sessions, jointly with OFWE).

• Gender quota: in a random sub-sample of FUGs, OFWE will introduce the requirement that the EC’s vice chairperson and at least one other member are female. Currently, FUG leadership is dominated by men. Correlational evidence based on observational data suggests that local governance improves, and that better livelihood and conservation outcomes emerge when a woman is represented on the EC (Kahsay et al., 2021).

Experimental design

Villages are assigned to treatment through phase-in randomization in target expansion zones. Conditional on assignment to treatment, villages are assigned to leadership sub-treatments through a lottery. The unit of randomization is the village, as the currently-favored approach for CBFM is based on creating one FUG per village. Our total sample consists of 250 villages, spread across the 4 experimental arms (50 villages per treatment arms and 100 villages in the control group). The main unit of analysis is households and individuals within households (men, women).

Design details

Randomization method
Randomization done in office by computer
Randomization unit
village
Clusters
250 villages
Observations
5000 respondents: 2500 households (2500 males, 2500 females)
Arms
50 villages per each of the three treatments arms and 100 villages in the control group
Primary outcomes
Modernization of agriculture (adoption of modern inputs: fertilizer and modern seed) Agricultural sales (revenues) Household income Forest conservation (forest cover density of primary crop trees, species diversity, evidence of transgressions such as charcoal burning sites)
Secondary outcomes
Women empowerment

Cited by

No citation data available for this registration yet.

Keywords