Land Rental Markets: Experimental Evidence from Kenya
How to cite
Casaburi, Lorenzo; Willis, Jack. 2019. "Land Rental Markets: Experimental Evidence from Kenya." AEA RCT Registry. AEARCTR-0004530. https://doi.org/10.1257/rct.4530-1.0
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
In the experiment, we study the impact of renting out a plot on the plot and on the plot owner. A random sample of interested farmers receives a subsidy conditional on renting out a "target plot" (which is identified at baseline for all groups). Another group receives an unconditional subsidy. A final group is allocated to control. In a "first stage," we examine the impact of the rental subsidy on the likelihood that the target plot is rented out. In the main analysis, we plan to study the impact of renting out (instrumented with the subsidy) on plot-level and owner-level outcomes. The unconditional cash transfers will help shed light on the mechanisms behind this impact.
Intervention
We first collect names of farmers who own plots which: (i) they are not planning to rent out, and (ii) they would potentially be interested in renting out if they receive a subsidy to do so. Then we randomize these farmers into 3 groups:
1. Rental subsidy (payment conditional on renting the target plot)
2. Unconditional cash transfer
3. Control
Experimental design
We first collect names of farmers potentially interested in renting out an additional plot if they receive a subsidy to do so. Then we randomize these farmers into 3 groups:
1. Rental subsidy (conditional on renting the target plot)
2. Unconditional cash transfer
3. Control
For the analysis, we plan to instrument renting out with the rental subsidy (if the first stage is strong enough). If the unconditional cash transfer has little effect, we will report a specification with groups 2 and 3 pooled, so as to maximize power for the estimating the effect of renting out.
We will do multiple rounds of follow-up surveys: at least two, possibly more if funding allows.
Design details
Cited by
No citation data available for this registration yet.