DDI_BFA_2010_MCC-RLGP_v01_M Millennium Challenge Corporation IMPAQ International 2014-09-05 NADA Version 1.0 (September 2014) Version 2.0 (April 2015). Edited version based on Version 01 (DDI-MCC-BFA-IMPAQ-RGL1-2014-v01) that was done by Millennium Challenge Corporation. Rural Land Governance Phases I and II 2010-2013 MCC-RLGP 2010-13 BFA_2010_MCC-RLGP_v01_M IMPAQ International NADA Millennium Challenge Corporation Monitoring & Evaluation Division of the Millennium Challenge Corporation Independent Performance Evaluation Anonymized dataset for public distribution A quasi-experimental evaluation design was used to assess the impacts of the RLG Project. Program impacts will be estimated by comparing treatment group outcomes with outcomes from a comparison group. To enhance the accuracy of quasi-experimental evaluation results, comparison groups should be chosen to be as similar as possible to the treatment group on all characteristics that might affect the outcomes and program participation. To select comparison communes in Phase I, MCA-BF selected similar nearby communes. Unfortunately, MCA-BF does not have clear documentation of the selection criteria. Given the uncertainty about the process used in selecting the treatment and comparison communes, households located in the treatment areas may be systematically different from those located in comparison areas. In Phase II, IMPAQ participated in the comparison group selection and has clearly documented rationale and procedures for this selection. As a result, we are able to develop a rigorous evaluation design for the Phase II implementation of RLG. We therefore propose to use a difference-in-differences (DID) with (matched) comparison group analytic design to evaluate the impacts of the RLG Project. The DID design compares the before and after changes in outcomes between households in intervention communes (treatment group) and households in (matched) comparison communes (comparison group). It is important to note that DID methodology does not require baseline (pre-intervention) conditions to be the same in treatment and comparison groups. But for the DID design to be valid, the comparison group must accurately represent the change in outcomes that would have been experienced by the treatment group in absence of the intervention. In other words, the key identifying assumption behind the DID is that trends (changes) in outcomes between the treatment and comparison groups should be similar. Burkina Faso National Coverage Households Households from rural Burkina Faso. Sample survey data [ssd] The RLG Project was launched in September 2009. It was designed for implementation to take place in two phases. In the pilot phase, 17 treatment communes and 17 comparison communes were selected by MCA-BF. The pilot phase was recently completed and MCA-BF is implementing the expansion of the RLG Project to an additional 30 communes. In the RLG Project Phase II, the 30 treatment communes were selected non-randomly by MCA-BF. The 17 Phase I and the 30 Phase II treatment communes proposed by MCA-BF were agreed by MCC during Compact development. IMPAQ, together with MCA-BF, selected 29 comparison communes for Phase II (See Exhibit 4). These communes were selected through a purposive sampling rather than a probabilistic sampling procedure. Given the finalized study zones for Phase II (30 treatment and 29 comparison communes in 20 provinces and 12 regions), IMPAQ performed statistical power analysis to determine the necessary sample size - the total number of villages within each commune and the total number of households within each village to be sampled. We used two-stage sampling. First-stage is sampling at the village level and second-stage is sampling at the household level. Millennium Challenge Corporation Millennium Challenge Corporation http://data.mcc.gov/evaluations/index.php/catalog/104 Cost: None Meza, Jaime et al (2014). “Burkina Faso Rural Land Governance Performance Evaluation.” IMPAQ International: Columbia, MD.