Impact Evaluation of the Cash Transfer for Orphan and Vulnerable Children, 2007-2011
Other Household Survey [hh/oth]
This impact evaluation study composed of a three-wave panel dataset that was created to analyze the impact of Kenya’s CT-OVC cash transfer program. They include:
1. W1, Baseline 2007
2. W2, Follow-up 2009
3. W3, Follow-up 2011
This dataset is released by The Transfer Project, housed at the Carolina Population Center at the University of North Carolina - Chapel Hill. The third wave of the project, conducted by the Transfer Project, was funded by the National Institutes for Mental Health (NIMH). Additional information about the project not found here or without a direct link can be found on The Transfer Project’s Website: https://transfer.cpc.unc.edu/. The data package contains six primary datasets (from three individual/household and three community surveys) and several auxiliary datasets that provide additional information on tracking and attrition. A sensitive section from the third wave of the survey is excluded from this package and can be obtained through special requests (see Sensitive Data section). The survey interviewed households, individuals, and community members at three time points, in 2007, 2009, and 2011.
Kenya has been seriously affected by the HIV/AIDS epidemic. It has increased the number of orphans in the country and also the vulnerability of affected households, both through the loss of productive adults and through the impact of chronic illness. In response, the Department of Children's Services (DCS) in the Ministry of Gender, Children and Social Development, with assistance from UNICEF, developed the Cash Transfer Programme for Orphans and Vulnerable Children (CT-OVC). After a small pre-pilot phase, a second larger pilot phase was initiated in seven districts in 2006. At the same time, the Government of Kenya expanded the Programme in other districts to an additional 30. The Programme expanded further in 2008/09, with a total of 30,315 households having received financial support by mid-2009. Additional expansion is planned, the eventual target being to support 100,000 households by 2012.
The objectives of the Programme were clarified as Phase 2 progressed. Its overall objective is to provide a social protection system through regular and predictable cash transfers to families living with OVCs [orphans or vulnerable children] in order to encourage fostering and retention of OVCs within their families and communities, and to promote their human capital development. The latter includes, specifically, to increase enrolment and attendance in basic school; to reduce the rates of mortality and morbidity in children aged five years and under, particularly through increasing the uptake of immunization, growth control and vitamin A supplements; to promote household nutrition and food security; to increase civil registration of children and caregivers; and to improve household knowledge and appropriate case management for individuals with HIV/AIDS through coordination with other service providers.
Kind of Data
Sample survey data [ssd]
Unit of Analysis
Households, Individuals, Communities
The scope of the household survey includes:
A. Baseline survey, 2007: Preliminary questions; Household roster; Food consumption and expenditure; Non-food consumption and expenditure; Sources of income and ownership of productive assets; Housing characteristics and ownership of durable assets; Education; Child health.
B. Follow-up survey, 2009: Confirmation of household consumption; Household roster; Operational performance; Food consumption and expenditure; Non-food consumption and expenditure; Sources of income and ownership of productive assets; Subjective poverty assessment, savings and shocks; Housing characteristics and ownership of durable assets; Education; Child health.
C. Follow-up survey, 2011: Confirmation of household consumption; New household members listing; Household roster; Education; Health (All persons age 3 and above); Child health (Children of age 0-60 months old); Fertility (All female members ages 12-49); Food and non-food consumed; Housing and ownership of durable goods; Income generating activities; Income, transfers, and program participation; Preferences, expectations and taste; Anthropometrics (Children age 0-10 years).
The scope of the community survey includes:
A. Baseline survey, 2007: Preliminary details; Interviewee roster and community characteristics; Civil infrastructure and seasonality; Wages and economic activity; Local prices; Roster of orphans; Service provision.
B. Follow-up survey, 2009: Preliminary details; Interviewee roster and community characteristics; Civil infrastructure and seasonality; Wages and economic activity; Local prices; Roster of orphans; Service provision.
B. Follow-up survey, 2011: Preliminary details; Community characteristics; Civil infrastructure and seasonality; Wages and economic activity; Local prices; Service provision; Social norms; Savings.
Producers and sponsors
Research Solutions Africa
Government of Kenya
United Nations International Children's Fund
Oxford Policy Management
University of North Carolina at Chapel Hill
Research Solutions Africa
Food and Agriculture Organization
United Nations International Children's Fund
Department For International Development, UK
National Institute of Health
The evaluation took place in the seven districts that had already been identified by the Programme (Kisumu, Migori, Homa Bay, Suba, Nairobi, Garissa and Kwale). In each district, two locations were randomly selected to benefit from the Programme intervention, and two acted as controls. In three districts and one sub-location of Nairobi, the Programme decided to impose conditions with penalties; in the remainder, there was no systematic monitoring of compliance with conditions and no penalties were imposed. For the evaluation, recipient households were sampled from a list supplied by the Programme. Other households were sampled from a household listing undertaken in a random sample of census enumeration areas. A total of 2,759 households were included in the baseline sample; of these, 2,255 were interviewed again at follow-up.
The sample for the quantitative survey consists of four groups:
- Group A Households with OVCs in the Programme areas selected for inclusion in the Programme - divided into two groups; areas with conditions with penalties, and those without;
- Group B Households with OVCs in control areas that were expected to have the met Programme criteria and would therefore (in theory) have been selected by the Programme if the Programme had operated there;
- Group C Households with OVCs in Programme areas that were not selected for inclusion in the Programme; and
- Group D Households with OVCs in control areas that were expected not to have met Programme criteria and would not (in theory) have been selected had the Programme operated there.
Samples were drawn for these four groups of households. Programme recipient households were sampled from a list supplied by the Programme. Households in groups B, C and D (i.e. all except Programme recipients) were sampled from a frame developed through undertaking household listing in a random sample of census enumeration areas (EAs). Census enumeration areas were sampled with probability proportional to population size (PPS). The household listing collected information used to identify OVC households and to classify households as likely to be poor, based on socio-economic information provided by the households. This was used to distinguish the group of poor OVC households that acted as controls (group B). Households from groups C and D provided information on non-beneficiary households; group C households were used to assess Programme targeting.
More detailed information on the sampling process is given in Annex A of the final report of the impact evaluation.
Deviations from the Sample Design
The intended initial total sample size was 3,161 households. Due to the unavailability of the right respondents, a total of 2,759 households were interviewed and included in the baseline sample for analysis (87 per cent). The households were paneled and, when it was possible to trace them, survey teams revisited and interviewed the same households for the follow-up. Some 2,255 of the baseline households were interviewed at follow-up (82 per cent of those interviewed at baseline). The proportion of households that could not be re-interviewed at follow-up was higher in control households. The sample at baseline included a total of 15,464 individuals, of whom 9,231 were children. At follow-up, the sample included 12,959 individuals, of whom 7,532 were children, although not all of these individuals were necessarily included in the baseline survey. The loss of households between the baseline and follow-up survey was higher than had been hoped and was, in part, due to the post-election violence. It could potentially affect the results of the analysis and is discussed in Annex F of the final report of the impact evaluation.
Data was analysed using sampling weights calculated as the inverse of the relevant sampling fractions within the locations that has been selected for inclusion in the study, based on the baseline sample. The study does not provide information about the OVC population in the country as a whole, but only for the particular population included in the evaluation: the weights reflect this.
Dates of Data Collection
Wave 1; Baseline
Wave 2; Follow-up
Wave 3; Follow-up
Data Collection Mode
Face-to-face paper [f2f]
Non-sensitive data are available for download upon approval of a restricted use application, which comprises a data request form, a data use agreement and a security pledge. Sensitive data are likewise available to qualified researchers through an additional process of obtaining an IRB approval and providing a satisfactory data security plan. Available datasets vary by country and wave of the data collection, and may be designated “sensitive” due to inclusion of birth day and month, sexual history, or other medical information.
The dataset is accompanied by a Data Use Instructions document (available below) that should be read prior to requesting any data, as it provides further information on the project and dataset contents.
See https://transfer.cpc.unc.edu/tools/data-2/ for more details.