Afghanistan Targeting the Ultra-Poor Impact Evaluation

The Intervention to be evaluated is the “Targeting the Ultra-poor” Program (TUP), a holistic package combining the transfer of a productive asset with structured training, mentoring, a basic stipend, and other complementary services for a defined period aiming to help households break their poverty cycle. The original (TUP) graduation program was designed and implemented by BRAC in Bangladesh and has subsequently been promoted by CGAP and the Ford Foundation.

The program focuses particularly on ultra-poor women, who are able to manage an enterprise but have no productive assets in the household and are not connected to a microfinance institution. The Afghanistan TUP program will take place in 6 provinces and the impact evaluation will take place in the Balkh province only. The intervention is implemented by the Microfinance Investment Support Facility for Afghanistan (MISFA) through different NGOs in each province where the program operates. In the study villages, the implementation was conducted by the Coordination for Humanitarian Assistance (CHA).

To identify ultra-poor households, the program included a village- and household-level selection process. Program staff first qualitatively identified the poorest villages in the province subject to having availability of veterinary services, financial institutions and social services, and high population density. Once the villages were selected, a Participatory Rural Appraisal (PRA) was conducted to identify poor households. This included a combination of a community poverty wealth ranking, followed by a verification of the poorest 20% based on this ranking. The verification was conducted by CHA through a short survey, and this was followed by a final verification by MISFA of the eligible households submitted by CHA. The final selection criteria for TUP beneficiaries was based on meeting at least three of the following six criteria checked during the verification:

(1) Household is dependent on female domestic work or begging;
(2) Household owns less than 20 decimals (1 biswa) of land or is living in a cave;
(3) Targeted female is 50 years old or younger;
(4) There are no active adult male income earners in the household;
(5) Children of school age are working for pay; and
(6) Household does not own any productive assets.

Once identified, the ultra-poor households received the various components of the program, which can be broadly categorized into the following:

(1) The transfer of a productive asset in the form of livestock (e.g., cows, goats);
(2) A monthly cash transfers/stiped (AFN 1,000 per month for 12 months);
(3) A training focused on basic concepts of livestock rearing and entrepreneurship;
(4) Fortnightly “mentoring visits” by social organizers, and vet services to:
a. evaluate the asset and related outputs as well as to recommend follow-up actions. Depending on the evaluation of the asset, additional support in the form of food supplement or asset replacement were options for the program participants; and
b. promote activities encouraging improved behavior across a range of dimensions (health, education, female empowerment, financial inclusion, and social cohesion/community support).

A sequenced approach is applied, which is informed by the original BRAC graduation model. The recipient receives support for their livelihood selection, which includes an intensive and repeated consultation between MISFA, the partner field staff and the ultra-poor women so that participants can make an informed choice from different enterprise options. The productive asset transfer is worth AFN 21,000 – 27,500 (approximately 330 - 450 USD). The consumption stipend aims to support the household with basic food needs, initially to replace the potential forgone income or productive time that the TUP recipient spends learning about and initiating their business rather than attending to their usual duties.

2016-05-02

2017-07-31

We examine the impact of the intervention on 7 main outcome groups that measure household well-being across diverse dimensions.

The primary outcomes include:
1. Consumption
2. Assets
3. Financial Inclusion
4. Time spent working
5. Psychological well-being/Mental health
6. Political Involvement
7. Female Empowerment

1. Consumption: Total per capita consumption (food consumption, non-food expenditure, durable good expenditures)
2. Assets: Total value of productive and household assets
3. Financial Inclusion: Access to formal and informal borrowing, total value of formal borrowing, total value of informal borrowing; total amount deposited into savings, total savings balance
4. Time spent working: Total time spent working including on agriculture, livestock, business and paid and unpaid work (primary male and female). Switch from lower- to higher-return activities.
5. Psychological well-being/Mental health: Index of subjective and objective psychological well-being self-perception of life satisfaction, happiness and self-esteem and lack of stress. Subjective components are: self-reported life satisfaction (WVS), self-reported happiness (WVS), CES-D-SF depression symptoms, Rosenberg self-esteem Scale (RSES), and Cohen Perceived Stress Scale (PPS-4). The objective component is salivary cortisol level, a biomarker of stress. (Primary male and female).
6. Political Involvement: Voted in the last presidential election, voted in the last provincial council election, has a (Tazkira Afghan national identification card), is a member of a political party, attended village meetings, approached village leaders about village issues.
7. Female Empowerment: Female decision was taken into consideration and/or followed in (1) household expenditure decisions (food, household repairs, clothing, land, property and other high-value expenditures); (2) household financial (credit and savings) decisions; (3) children education decisions; (4) children’s marriage decisions; (5) household’s business decisions

1. Food Security
2. Income from productive activities
3. Labor Participation
4. Child Health
5. Education

1. Food Security: Everyone eats at least two meals every day, no adult skips meal or cut the size of meals, no child skips meal or cut the size of meals
2. Income from productive activities: Livestock, agricultural, business and labor income
3. Labor Participation: Employed (work as paid employees, work in their own business/farm, or work as unpaid worker in the family-operated enterprise) or looking for job (try to find work outside the home in the past 2 weeks) (primary male and female)
4. Child Health: Lack of diarrhea in under-5 children
5. Education: Children in school-age (are enrolled in school), number of days missed of school in the last four weeks.

We use a household-level randomized experimental design to estimate the causal effect of the intervention on socioeconomic outcomes of ultra-poor households. The evaluation sample comes from four districts and 80 villages in Balkh province. The participatory wealth ranking was conducted in all villages of the Balkh province to identify the eligible population and collect baseline data on them before the public lottery took place. Larger villages were split into multiple PRA groups (133 in total), The randomization was stratified by PRA by holding separate lotteries for each PRA group. 1,219 ultra-poor households were randomly assigned into one treatment group (491 households) and one control group (728 households). The treatment group received the intervention package as described in the intervention and the control group does not receive any of the components. We will estimate the impact of the program on ultra-poor households, comparing the treatment and control group.

Public lottery. The names of the ultra-poor household that received the intervention were drawn from a bin.

Household

No

0

1,219 ultra-poor households

1,219 ultra-poor households
491 ultra-poor households in the treatment group
728 ultra-poor households in the control group

Using baseline data, for our main outcome of interest, monthly per capita consumption, and some other outcomes we estimate a minimum detectable size of around 16% SD for this sample, with power = 0.8 and alpha = 0.05. 1) Consumption: Mean AFN 1,111 (SD AFG 671) MDE AFG 110 (10% of the mean in the untreated state). 2) Saved in the last 4 weeks: Mean 1% (SD 10%) MDE 1.64 pp (164% of the mean in the untreated state). 3) Food security (Household members ate at least 2 meals) Mean 80% (SD 40%) MDE 6.55 pp (8% of the mean in the untreated state). The additional stratification at the PRA level and multiple measures will give an additional cushion to the MDE.