Improving sanitation subsidy targeting: Comparing methods for identifying vulnerable households in rural Ethiopia

Executive Summary

Introduction

Targeted sanitation subsidies can play a transformative role in increasing access to improved latrines and reducing open defecation among the poorest households. However, a persistent challenge is identifying eligible households in a way that is fair, accurate, and cost-effective.

This research brief, developed by IDinsight in partnership with iDE, compares three targeting approaches: 1) Community-Based Health Insurance (CBHI) exemption lists, 2) the Poverty Probability Index (PPI), and 3) food insecurity measures to evaluate their effectiveness in identifying vulnerable households in rural Ethiopia. The analysis draws on data from more than 3,200 households across 104 villages in the Wolaita Zone and assesses each method’s accuracy, overlap, and operational feasibility.

Key Findings

• The three approaches capture different segments of the poor population with limited overlap. Only 22% of CBHI-exempt households were classified as poor by PPI, and 71% of PPI-poor households were not on CBHI lists.
• CBHI appears to capture the most visibly vulnerable households, such as those experiencing acute hardship, social isolation, and food insecurity. The PPI, which relies on asset-based indicators, may miss households facing short-term shocks or less visible forms of vulnerability.
• Importantly, this does not suggest that either method is ineffective in targeting poor households. Instead, CBHI and PPI reflect different dimensions of poverty. CBHI, grounded in community knowledge, often prioritizes households experiencing immediate or observable hardship, such as illness, disability, or lack of social support, that may not be captured by standardized poverty metrics. PPI offers more consistency and scalability but may overlook marginalized groups.
• Food insecurity indicators, especially whether households went a day without eating or experienced hunger due to a lack of money, were helpful in identifying ultra-poor households that may fall outside standard poverty lists.
• Operational challenges with CBHI, including outdated or incomplete records, were mitigated through local verification processes.

Recommendations

To improve the accuracy and equity of subsidy targeting:

1. Use CBHI exemption lists as the primary targeting mechanism. CBHI is low-cost, widely recognized, and aligned with existing government systems. It captures critical forms of deprivation often missed by statistical poverty tools.
2. Supplement CBHI with two simple food insecurity questions to reduce exclusion errors. Specifically, identify households that report either (a) going a whole day without eating or (b) experiencing hunger due to lack of money in the past four weeks. These indicators highlight extreme deprivation and can help reach those who may be left out of CBHI.
3. Incorporate latrine ownership status to ensure subsidies are directed toward households lacking access to improved sanitation.
4. Local leaders and health workers should be involved in verifying CBHI lists before rollout to minimize inclusion errors and build community trust in the process. 
5. Where resources are limited, prioritize CBHI-exempt households who are also food insecure or have larger household sizes. This ensures the most vulnerable benefit first when subsidy pools are constrained.

Conclusion 

A combined targeting approach grounded in existing administrative systems, strengthened by simple household-level indicators, and validated through community engagement offers a practical, scalable, and context-appropriate strategy for identifying poor households in sanitation programs. The findings and recommendations presented here are relevant for Ethiopia and other low-resource settings aiming to design more equitable and effective WASH subsidy schemes.