STARS
Supporting Teacher Achievement in Rwandan Schools
About the Program
The STARS program is an innovative initiative working to improve education quality in Rwanda through performance-based compensation. Building upon a previous IPA evaluation that demonstrated pay-for-performance contracts had no negative impact on teacher recruitment while positively influencing teacher performance—particularly in classroom presence and pedagogy—and enhancing student learning outcomes, we are now partnering with researchers, gui2de, and the Rwandan government to identify the most effective designs for integrating performance-based compensation into teacher contracts nationwide.
The Challenge
Rwanda's education system faces significant challenges across multiple dimensions:
- Nearly 20 percent of primary school teachers leave the profession annually
- Only 37 percent of teacher graduates actually pursue teaching jobs after graduation
- Fewer than 40 percent of Primary 3 students meet expected English proficiency benchmarks
- Less than 60 percent of students achieve math proficiency standards
- Performance gaps are widening in economically disadvantaged districts
In recent years, Rwanda has implemented education reforms including a 2022 salary increase of 40-80% of net income for teachers and the development of the Comprehensive Assessment Management Information System (CAMIS) launched in 2019-2020. CAMIS serves as Rwanda's national digital platform where teachers regularly record student learning outcomes.
Research Foundation
From 2015 to 2018, researchers partnered with IPA to rigorously evaluate whether a pay-for-performance model could improve teacher performance. The study found that this approach led to better teaching quality and improved student learning, without discouraging teacher recruitment—a common concern with performance-based pay.
Building on these findings, the STARS program is now working with IPA, researchers, gui2de, and the Rwandan government to explore how performance-based bonuses can be integrated into national teacher contracts.
The Challenge with Traditional Contracts
Teacher contracts in Rwanda have traditionally relied on the imihigo performance management system, where teachers set annual targets and receive bonuses of 0–5 percent of their salary based on their performance. However, the system is hindered by:
- Subjective evaluation criteria
- Fixed bonuses rather than meaningful performance incentives
- Lack of clear links between teacher practices and student learning
- No standardized method for measuring classroom effectiveness
Variations Tested
Model B.1: Headteacher Assessment
This variation adds school leadership perspective by incorporating headteacher evaluations of teachers alongside the standard measurements. Headteachers assess teachers on the same 5Ps but from a school-level viewpoint based on their more frequent interactions with teachers. These assessments count for 10 percent of a teacher's overall performance score.
Model B.2: Performance Feedback
This model adds regular feedback throughout the school year. Teachers receive termly updates about their relative performance and projected year-end ranking compared to peers. This approach tests whether awareness of one's standing motivates improved performance, even without changing the measurement or bonus structure.
Model B.3: Threshold Bonuses
Initially designed to test a three-tier bonus structure (0 percent for bottom 25 percent, 3 percent for middle 50 percent, and 5 percent for top 25 percent of teachers), this model became effectively identical to Model A when all models had to adopt the same threshold structure due to payment system limitations.
Model B.4: Enhanced Assessment Verification
This variation strengthens accountability by adding re-testing of an additional 5 percent of pupils beyond the standard assessment audits. This provides stronger verification of student learning outcomes, ensuring that the pupil learning component of teacher evaluation is based on reliable data.
Model B.5: School Leadership Incentives
Added in Year 2, this model creates aligned incentives throughout the school by introducing performance contracts for headteachers. Half of a headteacher's evaluation is based on the collective performance of teachers in their school across the 5Ps, potentially encouraging more supportive leadership and coaching.
Results
- 248 schools with STARS contracts across 10 Rwandan districts
- 95 comparison schools with standard contracts
- 4,762 primary school teachers engaged
- Over 207,000 students impacted
- Teacher presence increased slightly from an already high 92 percent in standard contract schools
- STARS teachers were 17-19 percentage points more likely to be prepared
- Pedagogy scores rose by 4.5-7 percent compared to standard contract schools
- CAMIS participation increased by 5-9 percentage points across contract variations
- Improvements in teacher inputs were observed across all contract variations
- Year 1 impacts on student learning outcomes were modest and statistically insignificant
- This aligned with previous RCT findings where larger learning gains emerged only in the second year
- Added 100 additional schools with STARS contracts
- Added 50 new standard contract schools
- Total: 489 schools
- Over 6,600 teachers
- Approximately 300,000 students
- Introduction of Model B.5 with headteacher performance incentives
- STARS contracts continued to strengthen teacher preparation, classroom practices, and CAMIS participation
- Teacher presence remained high across all schools
- CAMIS participation rates improved from 30.3 percent to 58.9 percent nationwide
- STARS districts continued to lead in implementation
- Emerging evidence that some contract variations (models B.1, B.4, and B.5) positively influenced student learning outcomes
- The Ministry decided to drop lower-performing models for Year 3 and reallocate schools to new variations
- Testing the most promising contract designs identified from previous years
- Preparing for potential nationwide expansion
- Developing a system that can support all 66,000 teachers
- Benefiting nearly 3 million primary school students across Rwanda
Year 1 (2022-2023)
Scale:
Key Findings:
Year 2 (2023-2024)
Expanded Scale:
Key Findings:
Year 3 (2024-2025)
Current Focus: