Assessing Regional Heterogeneities in the Uptake of a Pilot Community-based Health Insurance Scheme in Ethiopia

Abstract

In 2011, Ethiopia launched a pilot Community-Based Health Insurance (CBHI) scheme to expand healthcare access for underserved populations and diversify healthcare financing. Initially introduced across four major regions - Tigray, Amhara, Oromia, and the Southern Nations, Nationalities, and Peoples' Region (SNNP) - the program’s design and implementation varied significantly in terms of contribution levels, renewal periods, and service delivery. Recognizing these disparities and the limited literature on sub-national performance, this study investigates the determinants of enrolment and membership renewal across these four regions. By taking into account inherent regional heterogeneities, the study seeks to generate empirical evidence to inform the ongoing national rollout of the CBHI scheme. The study utilizes a longitudinal household survey spanning five years (2011–2015). The baseline survey in 2011 sampled 102 households from each of the 16 pilot districts, totaling 1,632 households (9,455 individuals). Follow-up surveys were conducted in 2012, 2013, and 2015, maintaining a high retention rate with only 4.1% attrition over the study period. Logit models were employed to estimate marginal coefficients and identify regional drivers of scheme performance. Results reveal stark regional disparities: Amhara emerged as the highest-performing region, recording a 69% coverage rate and a low dropout rate of 10% by 2015. In contrast, Tigray exhibited the lowest performance, with a 49% uptake rate and a 32% dropout rate. Multivariate analysis indicates that the poor are disproportionately excluded from the scheme in Amhara and Oromia, while membership renewal patterns show a "pro-rich" bias in Oromia and SNNP. Adverse selection was identified as a notable concern uniquely in Amhara, likely linked to a 2014 policy shift that made contribution rates for households of up to five members the same, which was not the case before (it was based on household size). Demographic drivers also varied: female and elderly populations were positively correlated with insurance uptake in Tigray, whereas male adulthood was the primary predictor in SNNP. On the supply side, the availability of blood-testing equipment significantly enhanced enrolment in Tigray and Oromia. To improve scheme retention and coverage, interventions must target region-specific bottlenecks, such as enhancing the quality of care in SNNP and reducing clinical consultation waiting times in Tigray and Amhara. Ultimately, the evidence underscores that ‘one-size-fits-all’ strategies are insufficient; expanding CBHI coverage must be tailored to address local socioeconomic and supply-side heterogeneities.