RESULTS-BASED FINANCING RBF EDUCATION IMPACT DECEMBER 2019 TANZANIA A Simple Teacher Incentive System Can Improve Learning Introduction Tanzania devotes about one-fifth of government spending to education, focusing much of the funding on expanding school access. Primary school enrollment rates have surged, yet the quality of education services and learning outcomes remain poor, with only 38 percent of children aged 9–13 able to read or do arithmetic at the second grade level. Teachers play a critical role in helping children learn, but in Tanzania, many do not show up to teach. Poor motivation and lack of accountability have contributed to the high absenteeism and commensurate loss of instructional time. One way to strengthen teacher motivation and management is through performance pay. Teacher incentive schemes link bonuses or other rewards to specific targets, whether outputs (e.g., verified classroom presence) or outcomes (e.g., student test score improvement). Performance pay can help achieve learning results at low cost compared to teacher base salaries. In Tanzania, the Results in Education for All Children (REACH) Trust Fund supported a randomized control trial comparing two types of teacher performance pay systems and their effect on early grade learning. Intervention The two systems were developed by a civil society organization Twaweza East Africa, and built on an earlier experiment aimed at improving learning. Some 180 public primary schools participated, with 60 schools each testing one of two systems (Stadi or Mashindano) and 60 in the control group. Stadi, or “levels,” rewarded teachers based on the number of students The Results in Education for All Children who reached certain proficiency levels, for example, being (REACH) Trust Fund under the World Bank Group seeks to help countries strengthen their able to read certain sentences according to the Kiswahili education services by focusing initiatives on curriculum. Teachers received a bonus each time students results. The IMPACT series describes how achieved one of multiple thresholds, a more achievable goal results-based projects funded by REACH than under the earlier experiment, where bonuses were linked influence learning, education systems, and to students achieving just one threshold that comprised of a policy making. full range of curriculum skills. Mashindano, or “gains,” rewarded RESULTS-BASED FINANCING RBF EDUCATION IMPACT teachers based on students’ test score ranking relative to those Impact on Teachers of other children at the same starting level. This approach Participating in the Program recognized even small progress—helpful for teachers of students starting at low learning levels. For each intervention, the total prize was equivalent to about $3 per student, although TEACHERS’ ATTITUDES the actual amount teachers received depended on students’ performance. The final average bonus was equivalent to about 40 percent of a teacher’s monthly salary. Cover photo: “A teacher with her student in class” by GPE/Kelley Lynch, license: CC BY-NC-ND 2.0 Both interventions added three or four months of extra 9 in 10 learning on top of what students normally achieved. The support “levels” system was at least as effective as the “gains” system performance and resulted in a more equitable distribution of benefits, pay improving learning across all initial ability levels. It was also easier to communicate and implement. Impact 63% want The study was nationally representative, providing a basis to YES NO performance- assess whether the program could be scaled up nationally. based bonuses The results got the attention of the Ministry of Education and included in the President’s Office—Regional Administration and Local future salary Government, which together oversee Tanzania’s education reviews sector. They signed a memorandum of understanding with Twaweza in 2017 to formalize their interest in implementing a scale-up of the program. One consideration for scale-up was cost: The Twaweza intervention included one-on-one assessments of students to measure progress and in-person visits with teachers to ensure they understood the program. This required a lot of time and human resources. Learning from these implementation lessons, Twaweza funded a larger pilot of the “levels” model that will take place in 2019- RESULTS IN EDUCATION FOR ALL 2021. This pilot will reduce costs by assessing students in CHILDREN (REACH) RESOURCES groups rather than one-on-one, and by holding phone calls For more information on the REACH Tanzania intervention, please refer to the RBF Education with groups of teachers to explain the program instead of EVIDENCE note, “Can a Simple Teacher visiting schools. Incentive System Improve Learning?” on the REACH web page. Another positive impact was among teachers who participated in the Twaweza study. By keeping lines open and the message worldbank.org/reach simple, the program won the teachers’ trust. According to results of a survey of the teachers towards the end of the reach@worldbank.org program, nine in 10 expressed support for performance pay, and 63 percent said the government should include a performance-based bonus scheme in future salary reviews. Co-funders: the Governments of Germany, Norway, and the United States While teachers in some countries have resisted incentives programs, this kind of support for performance pay from teachers in Tanzania is noteworthy.