NEW UPDATE: Completed trials may now upload and register supplementary documents (e.g. null results reports, populated pre-analysis plans, or post-trial results reports) in the Post Trial section under Reports, Papers, & Other Materials.
Land Markets Frictions, Technology Adoption and Farm Profits
Last registered on December 04, 2019


Trial Information
General Information
Land Markets Frictions, Technology Adoption and Farm Profits
Initial registration date
December 04, 2019
Last updated
December 04, 2019 1:29 PM EST

This section is unavailable to the public. Use the button below to request access to this information.

Request Information
Primary Investigator
World Bank
Other Primary Investigator(s)
PI Affiliation
World Bank
PI Affiliation
World Bank
PI Affiliation
University of California - Berkeley
Additional Trial Information
On going
Start date
End date
Secondary IDs
Land Market frictions may prevent efficient use of new technologies. We test two interventions meant to resolve land market frictions in the context of potentially transformative technological change through irrigation in Rwanda. Our study aims to provide evidence on the importance of land frictions in a context of property rights, examines whether frictions in the reallocation of land hinders rural transformations, and tests whether a scalable solution to information or contracting frictions can improve the efficiency of land investments. In this study, we test whether (a) land market frictions restrain adoption of a new, and highly profitable technology and (b) whether an intervention approach to resolve frictions related to information and contracting relaxes these constraints.
External Link(s)
Registration Citation
Karpe, Saahil et al. 2019. "Land Markets Frictions, Technology Adoption and Farm Profits." AEA RCT Registry. December 04. https://doi.org/10.1257/rct.4697-1.0.
Sponsors & Partners

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information
Experimental Details
Institutionally, land in and around these irrigation sites is titled to farmers, and rental markets are active, with some evidence of farmers engaging in these transactions ex-ante. Land rental contracts typically consist of informal verbal contracts which may be incompletely specified, and contracting disputes are adjudicated by village leaders. Focus group discussions suggest that farmers find incomplete contracting a limiting factor for rentals. Our proposal resolves this constraint by introducing at random a formal written contract, co-signed by the village leader and sponsored by the Rwandan Agricultural Board (RAB). A small pilot study found a high take-up of this contract (we conducted focus groups with 22 farmers in one of the existing irrigation sites, distributed 30 copies of the contracts, and 10 of these contracts were ultimately used for land rentals), and farmers reported that the greater security allowed by the complete contract allowed them to rent land to farmers more distant in the social network.

Focus group discussions also highlighted some information gaps: informal systems to learn about farmers interested in renting in or out land are restricted to a few village “gossips” who are not considered to be particularly reputable sources for information. This may also contribute to a pattern of land rentals which are systematically close in the village networks. In addition to introducing complete contracting, we would introduce a farmer broker to collect and disseminate information on the supply and demand for land rentals. In doing so, we would create a novel, randomized test for whether land market frictions prevent farmers from capable of capitalizing on transformational technological change, and whether contracts and/or brokerage institutions can resolve these frictions.
Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
Land rentals (in and out), rental prices, irrigation use, crop choice, labor use
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Rentals to commercial farmers, rentals to farmers who are not closely linked, rentals to farmers with large households, rentals to farmers with less land in CA, rentals to farmers with adjacent plots, short term rentals, any rental contract disputes, yields, input use, cash profits
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
We propose two treatments, which will be cross-randomized at both the village and individual level.

First, contracts: We suggest that formal, written contracts that specify the property, dates of land use rights, crops to be grown and rental prices to be paid (in cash or in kind) , co-signed by village leaders may encourage rental transactions, particularly transactions between farmers who may be less closely related and where trust of good-faith contracting may be lower. In 55 of these 110 villages, we would introduce contracts at a village meeting and provide copies to the village lead farmer (the lead farmer is both a local authority on agriculture and receives a stipend from RAB to disseminate information. That lead farmer would be provided a list of farmer names selected at random from village listings and encouraged to discuss contracts with as many people as possible, prioritizing the village list. We would then provide the lead farmer with a small incentive to successfully reach this listing, verified through audits a week after the meeting. We anticipate that (a) at the village level, the availability of formal contracts leads to additional rentals and (b) at the individual level, contract promotion leads to increased rentals. A copy of the contract is available in the annex.

Second, information. In 55 of these 110 villages, we will ask the lead farmer to serve as a “farmer broker.” At a village meeting enumerators will discuss land rentals, and introduce this new role of the lead farmer. That lead farmer will be given a list of names selected at random from the village listing and he or she will be asked prioritize that listing, but to speak to as many farmers as possible to identify which farmers are interested in renting in or out land, whether their interest is greater for plots in the irrigated site or outside of the irrigated site, and whether the farmer is willing to have this information shared with interested parties in the village. Once this information is elicited, the farmer broker will return to the priority list and share information about potential matches. Once again, the records of the farmer will be audited, with the potential to receive a small incentive. We hypothesize that solving this information friction will lead to additional rentals and rentals to individuals who are more distant in the social network. We additionally hypothesize that there may be complementary impacts of the two interventions as both information and contracting may be simultaneous constraints preventing rental contracts from being executed between distant parties.
Experimental Design Details
Not available
Randomization Method
Randomization in office by a computer
Randomization Unit
Randomization of interventions (information and contracts) at the village-level, combined with randomization of visits to individual farmers within treatment villages to discuss contracts in detail and suggest potential matches with others in the village. Randomization was stratified across 5 strata constructed related to proximity to the site's command area.
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
110 villages
Sample size: planned number of observations
2362 farmers
Sample size (or number of clusters) by treatment arms
28 villages control, 27 villages contracts, 27 villages information, 28 villages contracts and information
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
At these sample sizes, power at the village-level randomization in the pooled regression is sufficient to detect a moderate effect of 0.23 standard deviations assuming a moderate (0.15) intracluster correlation; the disaggregate treatment regression can detect a moderate to large effect of 0.33 standard deviations and the individual regression utilizing the contract and broker target interventions are adequately powered to detect very small effects (although of course village-level spillovers may mitigate the effectiveness of the individual level randomization, which is why we report the village level power calculations).
IRB Name
Human Subjects Committee for Innovations for Poverty Action IRB-USA
IRB Approval Date
IRB Approval Number