Filipski, M, Taylor, JE, Abegaz, GA, Ferede, T, Taffesse, AS and Diao, X, 2017. General equilibrium impact assessment of the Productive Safety Net Program in Ethiopia. 3ie Impact Evaluation Report 66. New Delhi: International Initiative for Impact Evaluation (3ie)
Link to Source
This study examines the impact of Ethiopia’s large-scale social protection programme Productive Safety Net Programme (PSNP) on local and national economy-wide productivity. PSNP combines public works employment for cash or cereals and direct transfers and is targeted at the poor and vulnerable. It is anticipated,however, that the programme will likely have impacts on non-beneficiaries too. To estimate these impacts, the study uses local economy-wide impact evaluation (LEWIE) model to simulate PSNP and evaluate impacts. It also uses computerised general equilibrium (CGE) modelling to estimate nationwide spillovers.
Results suggest substantial local economy-wide impacts of PSNP. The programme’s multiplier effect vary considerably across kebeles, and they include both positive and negative spillovers for non-recipient households. Nationwide impacts are sizeable. While PSNP areas see the largest income benefits (with 6% increase in household income), the rest of the country also experiences income benefits of up to 2 - 4%, as a result of the PSNP’s impacts on supply and demand, wages and prices.
PSNP was started in 2005 by the Government of Ethiopia and a consortium of donors as a response to the chronic and severe food insecurity in rural Ethiopia. The main objective of the PSNP was ‘ to provide transfers to the food-insecure population in chronically food-insecure woredas in a way that prevents asset depletion at the household level and creates assets at the community level’. Most beneficiary households do public works which lead to the creation of public goods, such as irrigation and land conservation structures. Studies have shown that PSNP is a well-targeted programme (Coll-Black et al., 2012). While there is some evidence on the positive impact of PSNP on beneficiaries, spillovers to non-beneficiaries have not been studied rigorously. This study attempts to address this gap in knowledge.
What is the impact of PSNP on local and national economy?
Theory of change
The PSNP is a targeted programme that aims to provide cash and in-kind income to the poorest rural households by providing them employment in public works. While there are direct impacts on beneficiaries, there could be indirect spillover effects on the local and national economy. By providing additional income to the poor, the programme can stimulate the local economy by increasing consumption of local goods and services. The creation of public goods through the public works undertaken under PSNP could also increase local productivity. Both these mechanisms could potentially set into motion local income and productivity multiplier effects. These local impacts may also translate into impacts on the national economy as the programme is scaled up to cover the most food-insecure regions.
The main data set used for the estimation of PSNP impacts is the Ethiopian Food Security Survey (EFSS), a four-wave longitudinal quantitative survey, conducted from 2006 to 2012, which collected information at the household level on a range of variables, such as household characteristics, land and crop production, assets, non-agricultural income and credit, access to the PSNP, consumption expenditure, health, shocks, perceptions, and anthropometry. A community-level survey was also conducted on the types of public works undertaken as part of PSNP, wages and production. This study covered 63 woredas in the first round, and 77 woredas in the remaining rounds. Overall, around 2000 households were covered during this survey.
For CGE modelling at the national level, this study used the social accounting matrix (SAM) for 2005-2006 published by the Ethiopian Development Research Institute (EDRI 2009). The SAM consolidates all value flows in the economy during the 2005-2006 fiscal year, including production, consumption, investment, trade and government balances.
This study also uses qualitative findings from eight case studies.
To estimate the first level of impacts, this study uses a Generalized Method of Moments- Instrumental Variable estimation method. Once causal effects are established, a LEWIE model is used to estimate the impacts on the local economy. Then a CGE model is used to estimate impacts on the national economy.
The study finds differential impact on agricultural yield by the type of public works undertaken under PSNP. While soil and water conservation works led to an increase in agricultural yield by 2.8%, road construction had no impact on this outcome. A PSNP irrigation project led to an increase in the area under irrigation by 1.5 hectares. LEWIE estimates simulated at the kebele level too showed that the mean increase in total agricultural output was 6.4%. However, there was a large variation in these impacts across kebeles. The study finds significant income and production multiplier effects due to the cash transfers. For example, every US$1 transfer had an income multiplier effect of US$1.4 on average. Not surprisingly, this multiplier effects were higher in kebeles that were less integrated with outside markets. At the national level, the increase in real income due to PSNP was estimated to be 2-4%. Thus, both PSNP beneficiaries and non-beneficiaries were impacted by the programme.
About this impact evaluation
This project will evaluate one of the largest and oldest safety net programmes in Africa - the Productive Safety Net Program (PSNP) of Ethiopia - which covers 7-8 million recipients. The PSNP has provided cash and/or food support to eligible people primarily through participation in large-scale public works and through unconditional direct support to labour-constrained poor households.
While PNSP has been rigorously evaluated in the past, this evaluation will fill two important knowledge gaps. First, where existing assessments have mainly focused on the direct impact of transfers of the programme, this study will evaluate the impact of community assets built via the public works component. Second, while the safety net component has been evaluated in terms of the impact on beneficiary households, this study will test the hypothesis that significant income gains in rural areas can go beyond direct beneficiary households through consumption and other local linkages. Given that both community assets and multiplier effects could result in substantial benefits for non-beneficiaries, it is quite possible that the impacts of the PSNP on communities as a whole could be larger than the direct impact originating from transfers. The study will also disaggregate and measure the relative roles of multiplier effects and the accumulation of community assets.
This evaluation has the potential to serve policymakers as a powerful tool for running further experiments of immediate policy relevance, such as varying the levels, composition and targeting of transfers, or the composition of public works schemes.