Cash Transfer Evaluation Design in Northern Kenya
Présenté par Marta Marzi (Oxford Policy management) - Discutant: Jérémie Gignoux (Paris School of Economics - INRA)
The Hunger safety net programme (HSNP) is an unconditional cash
transfer programme that aims to reduce poverty in northern Kenya by
delivering regular cash transfers to beneficiary households or
individuals in four counties in the arid and semi-arid lands
(ASALs) of northern Kenya: Mandera, Marsabit, Turkana and Wajir.
During the pilot phase approximately 300,000 beneficiaries (60,000
households) were targeted under three different targeting
mechanisms. A consortium led by Oxford Policy Management (OPM) has
been contracted by DFID to undertake a rigorous evaluation of the
programme’s targeting effectiveness and impact based on panel data
(around 5000 households) collected over three years between August
2009 and November 2012.
The impact of the HSNP has been assessed using rigorous scientific
methods. Using a randomised controlled trial approach, a number of
geographical areas (sub-locations) were randomly chosen to be
assessed by the evaluation. These sub-locations were then randomly
assigned to be either ‘treatment’ areas, where the programme began
to operate straight away, or ‘control’ areas, in which the
programme did not start for two years (the duration of the impact
evaluation). The programme’s selection process was implemented
(prior to the baseline survey) identically in both treatment and
control areas. Impact was then assessed by comparing HSNP
beneficiary households in treatment areas with households in
control areas that were selected for the programme but who would
not come into the programme for two years.
In the presentation I will focus on the design of the evaluation
and only briefly touch upon the impact evaluation
results.