KIAT Guru made remote teacher pay depend on parents' evidence

Remote Indonesian schools had a monitoring problem: district supervisors were far away, one in five teachers was absent, and student learning was two grade levels below target in KIAT Guru's baseline.

The pilot did not just ask teachers to try harder. It helped communities form user committees, define five to eight service indicators, monitor teacher presence and score performance. Some schools linked those checks to teachers' remote-area allowances.

The World Bank evaluation found the model tying social accountability to pay based on presence was the most effective and sustainable. Control schools needed 12 months to complete a one-year curriculum; the strongest KIAT Guru model took 8.5 months.

Between 2017 and 2019, community presence scores rose from 65 to 95, service-performance scores from 56 to 97, and classrooms with teachers present from 75 percent to 87 percent.

Let parents verify the payroll signal

It used local observation to solve an information problem central supervisors could not cheaply solve.

The allowance was large enough to matter, but the scoring was anchored in agreed service indicators.

Public meetings made the signal hard to hide and gave principals a second source of authority.

The experiment compared payment designs instead of assuming accountability alone was enough.

When supervisors are far away, move verification to the users of the service. The incentive works better when the people who see the work can affect the payment.

After phase one, the government requested support to expand KIAT Guru to 410 schools. Phase two simplified the model and digitized accountability and pay processes; Indonesia later adopted parts of the innovation into national programs.