x

Please fill out this short user survey of only 3 questions in order to help us improve the site. We appreciate your feedback!
Impact Evaluation of the Londö Public Works Project in Central African Republic.
Last registered on June 28, 2019

Pre-Trial

Trial Information
General Information
Title
Impact Evaluation of the Londö Public Works Project in Central African Republic.
RCT ID
AEARCTR-0003404
Initial registration date
October 16, 2018
Last updated
June 28, 2019 4:54 PM EDT
Location(s)
Region
Primary Investigator
Affiliation
World-Bank
Other Primary Investigator(s)
PI Affiliation
World Bank
PI Affiliation
World Bank
Additional Trial Information
Status
On going
Start date
2017-02-01
End date
2019-12-31
Secondary IDs
Abstract
The Londö project provides temporary employment to vulnerable people through participation in a public works scheme. The phasing of Londö follows an adaptive schedule that takes into account geographic characteristics, existing infrastructure, and issues related to security and logistics. Work is being carried out by teams of 25 beneficiaries headed by a team leader recruited locally. Each sous-préfecture benefits from the manpower of 20 teams for 40 days in two successive campaigns with cohorts of 250 beneficiaries each.

Beneficiaries are selected through lotteries open to all individuals willing to participate. Each sous-prefecture runs two lotteries (one for each Londö campaign per sous-prefecture) during which 250 beneficiaries are selected for a main list and 50 for a waitlist. Work teams of 25 workers are formed according to the lottery ranking (individuals ranked 1 to 25 are in team 1, those ranked from 26 to 50 are in team 2, etc.). Each worker receives a daily wage of about $3 for the 40-day contract period and a bicycle. The low daily stipend promotes self-selection towards the poorest and most vulnerable. In addition, each worker receives a bicycle, which he or she can keep after successful completion of the 40-day contract period (during the 40-day period, the bike formally remains the property of the project).

The random selection of Londö project beneficiaries is exploited to design a rigorous impact evaluation methodology and answer some of the following research questions:
• How do Londö household-level impacts differ depending on whether a woman or a man participates in a public works project?
• Do female participants benefit in terms of increasing bargaining power or decision-making in response to participating in the Londö project?
• How does participating in a public works program change attitudes towards the state in the context of low public service delivery, poverty, and marginalized populations?
• How does the impact of a cash-for-work program depend on participants’ individual characteristics (e.g. by gender, age, household background)?
• Does women’s participation in a public works program have any negative consequences such as increased GBV?
• Do the asset transfers alter the relationship between spouses, in particular when women are the recipients, given the possibility that a bicycle may impact the economic opportunities available, access to services, etc.?
• Does access to a bike reduce the gender gap in mobility?
External Link(s)
Registration Citation
Citation
Alik-Lagrange, Arthur, Niklas Buehren and Markus Goldstein. 2019. "Impact Evaluation of the Londö Public Works Project in Central African Republic.." AEA RCT Registry. June 28. https://doi.org/10.1257/rct.3404-3.0.
Former Citation
Alik-Lagrange, Arthur et al. 2019. "Impact Evaluation of the Londö Public Works Project in Central African Republic.." AEA RCT Registry. June 28. http://www.socialscienceregistry.org/trials/3404/history/48885.
Experimental Details
Interventions
Intervention(s)
The Londö project (Central African Republic) provides temporary employment to vulnerable people through participation in a public works scheme. This work program consists of road maintenance, including vegetation clearing, debris removal, side drain and culvert cleaning, and minor surface repairs. These tasks are performed using a labor-intense technology that maximizes opportunities for the use of local labor and resources rather than machines. No work related to development, expansion, or new construction is carried out.

The phasing of Londö follows an adaptive schedule that takes into account geographic characteristics, existing infrastructure, and issues related to security and logistics. Beneficiaries are selected through lotteries open to all individuals willing to participate. Each sous-prefecture runs two lotteries (one for each Londö campaign per sous-prefecture) during which 250 beneficiaries are selected for a main list and 50 for a waitlist. Work teams of 25 workers are formed according to the lottery ranking (individuals ranked 1 to 25 are in team 1, those ranked from 26 to 50 are in team 2, etc.).

Each worker receives a daily wage of about $3 for the 40-day contract period. The low daily stipend promotes self-selection towards the poorest and most vulnerable. In addition, each worker receives a bicycle, which he or she can keep after successful completion of the 40-day contract period (during the 40-day period, the bike formally remains the property of the project). The bike serves two key objectives: (i) it represents an extra form of compensation and (ii) it enables workers to conveniently reach work sites, which is essential in the absence of a reliable and affordable public transport system. These bicycles are valued roughly at $60 each. Therefore, this asset transfer is an important element of the compensation package and it is assumed that bicycle ownership will impact socio-economic activities of beneficiaries well beyond the duration of the public works contract. The program’s monetary compensation and asset transfer are offered as the same bundle of benefits in all project areas.

The type of work involved is not gender-discriminatory and the project’s communication strategy ensures women are encouraged to apply to the public lottery. So far, the project has made great progress toward this target, as currently 34% of all beneficiaries are women.
Intervention Start Date
2017-02-01
Intervention End Date
2018-03-31
Primary Outcomes
Primary Outcomes (end points)
- asset
- primary and secondary economic activity
- income
- food consumption
- health and school expenditure
- trust in government
- exposure to and perpetration of violence
- mobility
- use of bicycle
- gender norms related to labor
- gender norms related to mobility
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
Given the timing of the intervention, we propose to collect follow-up data from those individuals who have already participated in - and been paid by - the project. The random assignment of the contracts will ensure that there are no systematic differences in observable and unobservable characteristics between individuals in the treatment and control group. Using the Londö project monitoring and evaluation database, we sample randomly winners and losers from the Londö lotteries lists stratifying by the gender of the lotteries' participants. We collect survey data for a large number of respondents with the objectives to be able to produce results dis-aggregated by the gender of the lottery participants and other observed characteristics. The survey includes recall modules for the main outcomes with the objectives to produce balance checks at the date of the lottery.
Experimental Design Details
Randomization Method
public lottery
Randomization Unit
individuals (and their household)
Was the treatment clustered?
No
Experiment Characteristics
Sample size: planned number of clusters
no cluster
Sample size: planned number of observations
targeted sample size 8000 sample among participants to the Londö lotteries as per the Londö project Monitoring and Evaluation data base (50/50 split between winners and losers at the Londö public lotteries)
Sample size (or number of clusters) by treatment arms
2 treatment arms, no cluster, targeted sample sizes are 4000 winners / 4000 losers
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
We are not able to access recent household living condition survey data for rural CAR. We thus have to rely on data from other countries to assess minimal detectable effects (MDEs) with the planned sample size. We checked estimated MDEs for sample sizes varying from 750 to 1500 observations by group, for outcome such as consumption per capita and food security, anchoring means and standard deviation on the 2012 Democratic Republic of the Congo 1-2-3 survey. This approach, albeit an imperfect substitute to power calculations based on CAR data, shows low MDEs and lets us expect sufficient statistical power with the planned number of respondents (assuming conservatively an heterogeneity analysis of winners/losers groups of size 750, we are expecting to be able to detect ~16% effect on expenditure per adult equivalent, going to ~12% for groups of size 1500).
IRB
INSTITUTIONAL REVIEW BOARDS (IRBs)
IRB Name
HML IRB Research and Ethics
IRB Approval Date
2018-12-13
IRB Approval Number
(IRB #1211, IORG #850, FWA #1102)
Analysis Plan

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

Request Information
Post-Trial
Post Trial Information
Study Withdrawal
Intervention
Is the intervention completed?
No
Is data collection complete?
Data Publication
Data Publication
Is public data available?
No
Program Files
Program Files
Reports, Papers & Other Materials
Relevant Paper(s)
REPORTS & OTHER MATERIALS