Mozambique - Land Tenure Regularization - Rural and Urban

Institutional Strengthening: - Land administration officials at national (DINAT headquarters), provincial (SPGCs), as well as district and municipal land offices - Local elected officials and leaders across provinces, and the matched targeted districts (19) and municipalities (16) - DINAT officials and Municipal/District/SPGC Office Directors - Local implementing partners, Registo Prediales and notaries, bank and micro-finance institutions, investors, APIEX, and donors with relevant land programming Rural Site-Specific: - Household members in Malema and Mecufi districts - Wives in Malemna and Mecufi districts - Community leaders in all matched treatment and control aldeias for Malema and Mecufi - SDAE in the districts and treatment and control aldeias in Mecufi and Malema, as well as SPGC and district office directors - NGOs and stakeholders in Maputo, as well as with investors in Malema and Mecufi districts - Land administration officials - Subset of the matched treatment and control aldeias in Malema and Mecufi, including adult men and adult women, elders/leaders responsible for dispute resolution, small and medium land size holders, households that have engaged in credit-taking, as well as households engaged in land transfers - Displacers persons in Mecufi - Individuals living in two non-project aldeias bordering each treatment aldeia Urban Site-Specific: - Household members in Nampula City and Monapo Vila - Wives in Nampula City and Monapo Vila - Traditional authorities/elders in five treatment and control bairros in Nampula City, as well as eight treatment bairros in Monapo Vila - Individuals affiliated with Nampula City and Monapo Vila municipal land offices, including municipal office directors - Mayors in Nampula City and Monapo Vila - Investors and employees working at bank or micro-finance institutiions, as well as notaries and the Registo Predial - SIGIT technicians in Nampula City and database technicians in Monapo Vila, in addition to any staff tasked with amanging the conflict office or registry - Individuals living in four bairros in Nampula City (two treatment, two control) and two of the original treatment bairros from Monapo municipality, including women above 35, women below 30, men above 35, men below 30, elders and leaders responsible for dispute resolution, residents in informal settlements, local business owners, households engaged in credit-taking, as well as households engaged in rental markets - Individuals living in five non-project bairro bordering five treatment bairros
For the quantitative urban and rural tools (incl. household and wives surveys), SI recalculated the minimum detectable effect size (MDES) through a power analysis to determine whether there are programmatically significant impacts that the evaluation will not be able to detect. The MDES is defined as the smallest impact the study could identify with a significance of .05 and 80 percent power. We estimate the MDES for outcomes aggregated to the household level using Stata 15. We use a multi-level cluster design to account for the fact that treatment was administered at the neighborhood level.
Overall, we find that the study is sufficiently powered to detect moderate program effects, in line with the expected impact of the Land Project. SI used the following parameters for power calculations:
· a = 0.05 - probability of a false positive (Type I) error
· Power (1-ß) = 0.8 - power to detect an effect if one truly exists
· ? - intra-cluster correlation (ICC); calculated for each variable
· j - number of clusters,
· m - average cluster size,
· µ - mu, baseline mean value; calculated for each variable
· s - sigma, standard deviation; calculated for each continuous variable
Beyond MSU's original design, we also expanded the set of outcome indicators examined by the power analysis and show the percentage change detectable by the evaluation above the baseline mean. This additional information on relative percent change will reduce the chance that policymakers erroneously conclude the intervention failed on the basis of statistical significance alone.
For Nampula City, Malema district, and Mecufi district baseline household samples. The ICCs were re-calculated for each of the outcome variables assessed baseline means. This was to ensure the study had sufficient power to detect policy-relevant program impacts where they existed, given variability around responses and actual ICCs obtained.
For Nampula City, depending on the indicator, we find that the study will be able to detect a wide range of effects from 5-40%, with most effects detectable around ten percent. The study is generally powered to see changes at a policy relevant level at or below MCC's expected economic rate of return of 24.8% for the Land Project; in terms of knowledge (8%), gender (7%), conflict (8-13%), credit (5%) and investments (9-13%). Service delivery outcomes are only powered to detect larger effects closer to 20%. The study is underpowered to detect anything beyond substantial changes in income (42%) and expenditures, but we will be able to determine whether the Land Project had an impact on important livelihood proxies, such as improvements in nutrition and increases in non-farm employment.
For the Malema and Mecufi rural site-specific impact evaluation, the study is powered to detect effects in the 10-43% range for Malema and from 6-36% for the Mecufi study area. The Malema study has less power than the Nampula urban hotspot impact analysis. In Malema, the study is generally powered to see changes at a policy relevant level for MCC, including knowledge (15%), gender (15-22%), conflict and perception of tenure (11-22%), and investments (10-24%).The evaluation is underpowered to detect anything in Malema beyond substantial changes in income (43%) and expenditures (26%).
Table 8 for Mecufi district shows that the evaluation is generally powered to see changes at a policy relevant level for MCC, including knowledge (9%), gender (8%), conflict (12-15%), and investments (10-19%). Beyond very large program effects, the evaluation is underpowered to detect changes in income (32%), expenditures (34%) and service delivery (36%).

If a baseline respondent is not identified/ needs to be replaced, the household living on the land that was originally in the baseline will be interviewed. If there is no longer a household on this land, then a replacement household selected via random walk will be selected.

For the urban community leader survey, we propose 3 to 5 structured surveys with traditional authorities/elders in five treatment and control bairros in Nampula City, as well as eight treatment bairros in Monapo Vila. To address questions regarding spillover, Community Leader surveys will also be fielded in one non-project bairro bordering treatment bairro (totaling 5 bairros). For each bairro, we plan to survey one to two of each of the following types of representatives: secretario do bairro, chefe de communidade, and chefe do quarteirão. The secretario do bairro is elected at the bairro level and the other two leaders correspond to subsequently smaller administrative areas. Across all of the relevant bairros, the sampling frames with the names of all secretario do bairros, chefe de communidades, and chefe do quarteirãos should be recorded, along with leader contact information and position title.
In Nampula, all five of the secretario do bairros corresponding to the five treatment and control bairros should be surveyed. In the two control bairros (Muatala and Mutaunha), a chefe de communidade should be randomly selected to survey from one of the study area unidad communales (i.e. Cossole, Minicane, Muralene, Namavo, or Napala in Muatala) and (i.e. 25 de Setembro, 7 de Setembro, Eduardo Mondlane, Muthita, or Piloto in Mutaunha). The third leader should be a chefe do quarteirão randomly selected from within the selected unidad communale. For the other three bairros (Muahivire, Muhala-Sede, and Namutequeliua), some unidad communales are included in the treatment group and some are included in the control group. For these three bairros, we will survey the secretario do bairro, a randomly selected chefe de communidade from one of the treatment UCs, and a randomly selected chefe do quarteirão from that selected UC, as well as a randomly selected chefe de communidade from one of the control UCs, and a randomly selected chefe do quarteirão from that selected UC. The sampling procedure in Monapo Vila will follow that of the procedure in Muatala and Mutaunha, where the chefe de communidade and chefe do quarteirão are selected from the study treatment areas.
For the rural community leader surveys, we propose in that in each neighborhood, to speak to 2 chefes (including the Regulo) from bairros that are part of the household sample.
For qualitative respondents:

Selection Requirements
Tool Stakeholder Tool Number Selection Notes
District Land Office- Directors 1 - Director of the Department who works the most with land administration - If the office uses SIGIT, interview this person
District Land Office- Technicians 2 - Any technician who is available - Prioritise technicians who have been in the institution for the longest period of time and who are involved with more aspects of land administration - If there is a SIGIT technician in the office, please interview him/her - If this person is the same as the Director, please inform Forcier staff
SPGC 3 - Director if possible - If not, please interview the SIGIT technician
Rural Notaries 4 - Any consenting notary
Investors 5 - Size of the investment does not matter. - Largest land holding investors are preferred - Active investors are preferred
Banks/MFIs (Rural and Urban) 6 - Loan officers who participate in lending in designated locations (i.e. in the District/Municipality of Malema, etc.) - Priority is for lenders in Mecufi, Malema, Monapo, Moma, and Nicoadala districts - If rural lending is not taking place, please inform Forcier staff (but interview different banks in city centres if necessary)
Municipal Land Office- Directors 7 - Director of the Department who works the most with land
Municipal Land Office- Technicians 8 - Any technician who is available - Prioritise technicians who have been in the institution for the longest period of time and who are involved with more aspects of land administration - If there is a SIGIT technician in the office, please interview him/her - If this person is the same as the Director, please inform Forcier staff
President of the Municipality or Mayor 9 - If the President or Mayor is not available, someone from his/her office who knows about land administration will be acceptable
Urban Notaries 10 - Any consenting notary
Predial 11 - Any consenting predial


Removal of Cabo Delgado locations due to growing insecurity:
1. Mueda
2. Mocoimba da Praia
3. Palma
4. Montepuez

The data collection firm, Forcier consulting, will attempt to interview respondents for KIIs from these areas via the phone or meetings in areas outside of insecure locations (i.e. Pemba). Quantitative data collection is not affected.