1. IFPRI Social Protection
Evaluation Work with the
Ministry of Social Solidarity
Sikandra Kurdi
International Food Policy Research Institute
May 7, 2023
2. Takaful Impact Evaluations
Regression Discontinuity
measures the impact even without a
baseline or a control group
Ideal for rigorous evaluation of large
national program without dedicated
randomized pilot
o Two rounds in 2017 and 2022
Disadvantage that measures “local
average treatment effect”: the
impact on hhs near the threshold
Supplemented by qualitative
evaluation in 2018 and
heterogeneity analysis (on-going)
3. Validation of Regression Discontinuity Methodology
2017 Round 2022 Round
Sample of 5326 households
• Applied for Takaful 2015- June 2017
• PMT score between 3900 and 5100
Second round sample of 6475 households
Applied for Takaful in May 2016 – Dec. 2016
•PMT score between 4437 and 4563
New set of households and narrower regression
discontinuity window in order to maximize
statistical power
4. 2017 Survey Results
70
13
101
0 500 1000 1500
Meat and Poultry
Fruits
Total Food
Household Spending per Month (EGP)
Total consumption increased by 8.4%; food spending increased by 8.9%
Increased spending on fruits and poultry
Increased spending on school costs
0
2
4
6
8
10
12
Egypt
(Takaful)
Brazil Columbia Honduras Mexico
Impact
on
Consumption
(%
increase)
123
211
0 500 1000
Household with primary school age
children
Household with secondary school
age children
Household Spending per Year (EGP)
Average Spending Takaful Impact
5. 2017 Survey Results
No impact on school attendance or healthcare attendance as conditionalities not yet
Negative impact on women’s reported decision-making
o Further explored in qualitative evaluation and 2022 round
o Suggests this was a short-term result concentrated in women with less education and
possibly related to decreased in employment for women
Targeting is good overall, but urban poor less likely to be included than rural poor and there
was a lack of understanding of program conditions
Urban Households in Poorest 40%
Rural Households in Poorest
40%
Heard of Takaful 78% 86%
Applied to Takaful 37% 50%
Takaful Beneficiary 9% 18%
Share of Applicants Accepted 18% 31%
6. Continue to fund the program and expand coverage
Improve targeting through better outreach particularly in urban areas and
updating PMT formula
Improve communication regarding program design on conditionality and
program length and recertification
Work towards a comprehensive social protection strategy
Increase transparency and communication regarding status of applicants
Explore additional measures to promote women’s decision-making as
giving transfers alone does not change women’s role in the household
Continue to survey and evaluate the program
2017 Policy Recommendations
7. 2022 Survey
After 5 years, financial impacts of cash have shifted from consumption to
investment and debt repayment
oReduction in debt by 4000 EGP
oIncrease in ownership of drip irrigation, plows, tractors, and livestock
Significant impacts on school enrollment:
0 20 40 60 80 100 120
Primary
Preparatory
Secondary
Takaful impacts on school enrollment
(percent enrollment)
Non-Takaful Enrollment Estimated Impact
8. Ongoing Work: Heterogeneity Analysis
Use administrative data to measure changes in asset ownership between
original registration and recertification
Test whether program impacts are stronger for poor households:
𝐴𝑠𝑠𝑒𝑡𝑆𝑐𝑜𝑟𝑒1 − 𝐴𝑠𝑠𝑒𝑡𝑆𝑐𝑜𝑟𝑒0 = 𝛼𝐴𝑠𝑠𝑒𝑡𝑆𝑐𝑜𝑟𝑒0 + 𝛽𝑍
9. Ongoing Work: Demographic Targeting Analysis
In 2017 survey results, impacts on consumption are higher for households
with older heads
ELMPS panel data analysis shows households accumulate assets as they
age so they have high PMT score relative to their consumption
=> For given PMT score, older householders are poorer
Policy recommendation supports change already made by Takaful to have
separate cut-offs by household head age
10. Ongoing Work: Forsa evaluation
New economic inclusion program to graduate beneficiaries of Takaful & Karama
to economic self-reliance
Two modalities: asset transfer and wage employment
Pilot program with 50,000 participants in 8 governorates
o 70 % current Takaful beneficiaries
o 30% rejected Takaful applicants
Randomized Control Trial:
o 160 Treatment sub-villages
o 163 Control sub-villages
o Goals: to measure impacts of Forsa on income and consumption for different
types of households (beneficiaries vs. rejected applicants; female vs. male
participants; high-skilled vs. low-skilled, etc.)
11. Ongoing Work: Forsa evaluation
Household survey in January-
February 2022
Sample 24 eligible households per
village
oExpect most but not all eligible
households in treatment villages to
end up participating in Forsa
83% of sampled households reported
willingness to enroll in Forsa
Of which 77% preferred self-
employment track
Current Takaful
beneficiaries
HH types
Rejected Takaful
applicants
Treatment Control
12. Ongoing Work : Forsa evaluation
Baseline survey in data provides a detailed description of work skills,
current employment, and time use in targeted households which has been
shared as an input to program design decisions
Midline telephone survey tentatively planned for late 2023 to measure
program uptake and impacts on employment and household production
13. Publications
4 Policy briefs
Full-reports on Takaful first round, Takaful second round, and Forsa
baseline
Academic journal article accepted at Economic Development and Cultural
Change on impacts of Takaful on women’s decision-making
In-progress:
oPolicy brief on Forsa employers survey
oAcademic journal article on Takaful medium-term impacts
oAcademic journal article on demographic heterogeneity in consumption
impacts