Targeting Humanitarian Aid in Crisis: Evidence from the reduction of WFP Assistance in Lebanon

The evaluation compares three targeting methods based on proxy means tests (PMT) which predict household vulnerability, and one on categorical scoring which selects households based on whether they match pre-identified vulnerability criteria.

The evaluation found that combining targeting methods leads to the selection of households that are more vulnerable. Although different targeting methods lead to similar average wellbeing overall, categorical targeting underperforms against PMT-based methods. Households excluded under categorical targeting had higher appeal rates, while the method itself was less effective at selecting the most vulnerable households.
In the short term (six months), stopping food assistance leads to lower food consumption, greater reliance on harmful coping strategies (including borrowing food or buying food on credit), lower life satisfaction and mental well-being, and reduced trust in aid agencies. 
The negative effects on economic wellbeing were significantly stronger for the most vulnerable households. They were less able to increase food spending to compensate for the end of assistance and more likely to rely on harmful coping strategies and have lower food security.