Pay for locally monitored performance? A welfare analysis for teacher attendance in Ugandan primary schools

To achieve the twin objectives of incentivizing agent performance and providing information for planning purposes, public sector organizations often rely on reports by local monitors that are costly to verify. Received wisdom has it that attaching financial incentives to these reports will result in collusion, and undermine both objectives. Simple bargaining logic, however, suggests the reverse: pay for locally monitored performance could incentivize desired behavior and improve information. To investigate this issue, we conducted a randomized controlled trial in Ugandan primary schools that explored how incentives for teachers could be designed when based on local monitoring by head teachers. Our experiment randomly varied whether head teachers' reports of teacher attendance were tied to teacher bonus payments or not. We find that local monitoring on its own is ineffective at improving teacher attendance. However, combining local monitoring with financial incentives leads to both an increase in teacher attendance (by 8 percentage points) and an improvement in the quality of information. We also observe substantial gains in pupil attainment, driven primarily by a reduction in dropouts. By placing a financial value on these enrollment gains, we demonstrate that pay for locally monitored performance passes both welfare and fiscal sustainability tests.