Effects of financial compensation structures on community health worker performance, motivation, and retention: evidence from a multi-arm quasi-experiment in Uganda.

Community Health Workers (CHWs) are central to extending primary health care in low-resource settings, yet their compensation remains a policy challenge. This paper reports findings from a five-arm quasi-experimental study conducted by Living Goods Uganda to test how different mixes of fixed stipends and performance-based incentives (PBIs) affect CHW performance, motivation, and retention. Over a nine-month period, 1,104 CHWs were assigned to five compensation models-70:30, 50:50, 30:70 PBI-to-stipend ratios, a 100% stipend arm, and a control-implemented across five districts. Quantitative data were analyzed using a difference-in-differences model with cluster-robust standard errors and wild-bootstrap inference, complemented by qualitative interviews and focus groups exploring experiences and perceptions. Results showed that moderate performance-based incentives (30%-50%) achieved the most balanced outcomes: improved household coverage, immunization follow-up, and referrals, alongside higher motivation and satisfaction. The 70% PBI arm generated stronger performance gains but increased stress and reduced retention, while the stipend-only arm offered stability but lower service coverage. Overall retention exceeded 95%, though sustained motivation depended heavily on supervision quality, recognition, and fairness of pay. Findings highlight that hybrid pay structures combining predictable stipends with moderate PBIs can enhance CHW productivity while safeguarding motivation and sustainability. The study offers practical guidance for Uganda's National Community Health Strategy and similar programs seeking equitable, gender-sensitive, and financially feasible CHW compensation models.