Custodes Futurifor the advancement of humanity

Teacher quality

Teacher quality is a barrier to human progress within the educational factors. The evidence for it, and the methods proposed to address it, follow.

Chetty, Friedman, and Rockoff (2014) linked U.S. school records to adult tax records.

Evidence

  • Long-run effects. Teachers who raise test scores more ("high value-added" teachers) also raise students' chances of attending college and their later earnings. These teachers also reduce teen pregnancy.
  • Replacing a weak teacher. Replacing a teacher in the bottom 5% with an average teacher would raise students' combined lifetime income by about US$250,000 per classroom. A published comment (Rothstein, 2017) found these long-run estimates sensitive to modeling choices. So this figure should be treated as uncertain.
  • Early grades, middle-income country. In Ecuador, more than 24,000 kindergarten children in 204 schools were assigned to teachers by a rule that was as good as random. A one standard deviation better classroom raised end-of-year scores by 0.11 standard deviations in language, 0.11 in math, and 0.07 in executive function (skills such as attention and self-control). Teacher behaviors seen in filmed lessons were associated with higher scores. Parents recognized better teachers but did not change how they helped their children (Araujo et al., 2016).
  • Pay and credentials miss quality. In Pakistan, standard methods produced reliable estimates of teacher value added (each teacher's contribution to students' test score gains). Effective teachers raised learning substantially. Yet observed teacher characteristics explained less than 5% of the differences in effectiveness. Public sector pay was unrelated to effectiveness. New teachers hired on temporary contracts at 35% lower wages were as effective as permanent teachers (Bau and Das, 2020).
  • Teaching time and knowledge in Africa. Researchers made unannounced visits, observed classrooms, and tested teachers in primary schools in Kenya, Nigeria, Mozambique, Senegal, Tanzania, Togo, and Uganda. They found that 44% of teachers were absent from class on average. They found that students received about 2 hours and 46 minutes of teaching a day, about half the scheduled time. Only 7% of language teachers and fewer than 70% of math teachers met a minimum standard of subject knowledge. Only 11% reached a minimum in general teaching knowledge. These are World Bank Service Delivery Indicators survey data, as published in the Journal of Economic Perspectives (Bold et al., 2017a). The earlier working paper version gave 2 hours 49 minutes and 10% (Bold et al., 2017b).
  • What helps. Identifying and supporting effective teachers. Long-run effects are contested (Rothstein, 2017).

Why it matters: Unequal access to effective teachers is itself a major barrier, independent of whether a school exists.

What can be done

Actions rated on the strength-rating scale (A strong and replicated, B solid but limited, C weak or debated, D contested or failed), applied to the specific claim made.

  • Verify teacher attendance and tie pay to it (B). A randomized trial in India checked teachers' attendance daily with cameras. Their salaries depended on days present. Teacher absence fell by 21 percentage points, and children's test scores rose by 0.17 standard deviations (Duflo, Hanna, and Ryan, 2012). This is one trial. A contract teacher trial in Kenya showed that results from nonprofit-run programs may not carry over to government delivery (see the last entry).
  • Individual teacher bonuses based on broad learning tests (B). A randomized trial used a representative sample of rural government primary schools in Andhra Pradesh, India. Students in schools with teacher performance pay scored 0.27 standard deviations higher in math and 0.17 in language after two years. The trial found no evidence of adverse effects. The program also beat schools given extra inputs of similar value (Muralidharan and Sundararaman, 2011). School-wide bonuses failed in New York City (see the proposed methods below).
  • Judge teachers by classroom results rather than credentials (B, C for long-run value-added effects). In New York City, certification status had at most small effects on student test scores. Teachers with the same certification differed greatly and persistently. So performance in the first two years predicted later effectiveness better (Kane, Rockoff, and Staiger, 2008). Another study used records for more than one million children. It found that students of teachers with high value-added (measured gains in their students' test scores) were more likely to attend college and earned more. It estimated that replacing a bottom 5% teacher with an average one would raise students' lifetime income by about $250,000 per classroom (Chetty, Friedman, and Rockoff, 2014). A reanalysis with North Carolina data found moderate bias in value-added and long-run results sensitive to controls (Rothstein, 2017). The original authors dispute this reanalysis.
  • Not effective: bonuses for short-run test scores without other change (B). In a randomized trial in rural Kenya, teachers were paid on students' test scores. Scores rose while the program ran, but through extra test preparation sessions. Teacher attendance, homework, and teaching methods did not change. The gains were not retained after the program ended (Glewwe, Ilias, and Kremer, 2010).
  • Not effective at scale: extra contract teachers hired through government (B). A randomized trial covered all eight Kenyan provinces. Contract teachers raised math and English scores by 0.19 standard deviations when a nonprofit hired them. But they had zero effect when the Ministry of Education provided them (Bold et al., 2018).

Proposed and experimental methods

Methods that are proposed, under trial, approved in some places, or tried and then failed. Each shows a stage label and an evidence rating. A stage label shows how far a method has progressed, not whether it works. The stage labels are explained on the educational factor page.

  • In-classroom teacher coaching (Large trial, B). A meta-analysis of 60 causal studies found that coaching improved teaching practice by 0.49 standard deviations and student achievement by 0.18. But the same meta-analysis found that effects from larger programs were only a fraction of those from small programs (Kraft et al., 2018). In South African public primary schools, both arms used the same lesson plans. Reading rose by 0.24 standard deviations with in-class coaching versus 0.12 with centralized training (Cilliers et al., 2020).
  • High-stakes teacher evaluation with dismissal threats and bonuses (Approved but not scaled, contested, B). Under the IMPACT system in Washington, DC, teachers just below the dismissal threshold left voluntarily about 11 percentage points more often. Those who stayed improved their evaluation scores by about 0.27 of a teacher-level standard deviation. Teachers near a large bonus threshold also improved (Dee and Wyckoff, 2015). The design compares teachers just on either side of rating cutoffs, so it shows effects near those cutoffs only.
  • AI assistants that coach tutors in real time (Early trial, C). Builds on the AI tutor entry in "Fixed beliefs about ability" in the psychological factors. Here, though, the AI guides the human tutor instead of the student. A randomized trial included 900 tutors and 1,800 K-12 students. Students whose tutors had the tool were 4 percentage points more likely to master topics, and 9 points more with lower-rated tutors. The tool cost about $20 per tutor a year (Wang et al., 2025). The study is not peer reviewed and was run by the tool's developers. Tutors reported some suggestions not suited to grade level.
  • Failed: school-wide teacher bonuses in New York City (Failed or reversed, B). Related to the loss-framed incentive entry in "Loss aversion and status quo bias" in the psychological factors, which tested a different bonus design. A randomized trial in more than 200 New York City schools found no effect on test scores, attendance, or graduation. The same trial found possibly lower achievement in larger schools (Fryer, 2013). A separate analysis found small math gains only where incentives to free ride on colleagues were weakest (Goodman and Turner, 2013).
  • Failed: doubling teacher pay without conditions (Failed or reversed, B). In a randomized trial in Indonesia, doubling the base pay of certified teachers improved their satisfaction with income and reduced outside jobs. But after two and three years, it had no impact on teacher effort or student learning (de Ree et al., 2018).

Sources cited on this page

  1. World Bank: Foundational learning brief (2025) B Moderate: agency estimate
  2. Chetty, R., Friedman, J. N., & Rockoff, J. E. (2014). Measuring the impacts of teachers II: teacher value-added and student outcomes in adulthood. American Economic Review, 104(9), 2633-2679. AEA C Limited: B for test-score effects
  3. Rothstein, J. (2017). Measuring the impacts of teachers: Comment. American Economic Review, 107(6), 1656-1684. AEA B Moderate
  4. Araujo, M. C., Carneiro, P., Cruz-Aguayo, Y., & Schady, N. (2016). Teacher quality and learning outcomes in kindergarten. The Quarterly Journal of Economics, 131(3), 1415-1453. DOI B Moderate
  5. Bau, N., & Das, J. (2020). Teacher value added in a low-income country. American Economic Journal: Economic Policy, 12(1), 62-96. DOI B Moderate
  6. Bold, T., Filmer, D., Martin, G., Molina, E., Stacy, B., Rockmore, C., Svensson, J., & Wane, W. (2017a). Enrollment without learning: Teacher effort, knowledge, and skill in primary schools in Africa. Journal of Economic Perspectives, 31(4), 185-204. DOI B Moderate
  7. Goodman, J., Gurantz, O., & Smith, J. (2020). Take two! SAT retaking and college enrollment gaps. American Economic Journal: Economic Policy, 12(2), 115-158. DOI B Moderate
  8. Roth, S., Robbert, T., & Straus, L. (2015). On the sunk-cost effect in economic decision-making: A meta-analytic review. Business Research, 8(1), 99-138. DOI B Moderate: for the effect in decision experiments, C for its size in real-world decisions
  9. Cilliers, J., Fleisch, B., Prinsloo, C., & Taylor, S. (2020). How to improve teaching practice? An experimental comparison of centralized training and in-classroom coaching. Journal of Human Resources, 55(3), 926-962. link B Moderate
  10. de Ree, J., Muralidharan, K., Pradhan, M., & Rogers, H. (2018). Double for nothing? Experimental evidence on an unconditional teacher salary increase in Indonesia. The Quarterly Journal of Economics, 133(2), 993-1039. link B Moderate
  11. Dee, T. S., & Wyckoff, J. (2015). Incentives, selection, and teacher performance: Evidence from IMPACT. Journal of Policy Analysis and Management, 34(2), 267-297. link B Moderate
  12. Fryer, R. G. (2013). Teacher incentives and student achievement: Evidence from New York City public schools. Journal of Labor Economics, 31(2), 373-407. link B Moderate
  13. Goodman, S. F., & Turner, L. J. (2013). The design of teacher incentive pay and educational outcomes: Evidence from the New York City bonus program. Journal of Labor Economics, 31(2), 409-420. link B Moderate
  14. Kraft, M. A., Blazar, D., & Hogan, D. (2018). The effect of teacher coaching on instruction and achievement: A meta-analysis of the causal evidence. Review of Educational Research, 88(4), 547-588. link B Moderate
  15. Wang, R. E., Ribeiro, A. T., Robinson, C. D., Loeb, S., & Demszky, D. (2025). Tutor CoPilot: A human-AI approach for scaling real-time expertise (arXiv:2410.03017v2) [Preprint]. arXiv. link C Limited
  16. Duflo, A., Kiessel, J., & Lucas, A. (2020). Experimental evidence on alternative policies to increase learning at scale (NBER Working Paper No. 27298, revised May 2022). National Bureau of Economic Research. link Published as Duflo, A., Kiessel, J., & Lucas, A. M. (2024). Experimental evidence on four policies to increase learning at scale. The Economic Journal, 134(661), 1985-2008. link B Moderate
  17. Duflo, E., Dupas, P., & Kremer, M. (2011). Peer effects, teacher incentives, and the impact of tracking: Evidence from a randomized evaluation in Kenya. American Economic Review, 101(5), 1739-1774. link B Moderate
  18. Duflo, E., Dupas, P., & Kremer, M. (2015). Education, HIV, and early fertility: Experimental evidence from Kenya. American Economic Review, 105(9), 2757-2797. link B Moderate
  19. Bold, T., Kimenyi, M., Mwabu, G., Ng'ang'a, A., & Sandefur, J. (2018). Experimental evidence on scaling up education reforms in Kenya. Journal of Public Economics, 168, 1-20. link B Moderate
  20. Duflo, E., Hanna, R., & Ryan, S. P. (2012). Incentives work: Getting teachers to come to school. American Economic Review, 102(4), 1241-1278. link B Moderate
  21. Glewwe, P., Ilias, N., & Kremer, M. (2010). Teacher incentives. American Economic Journal: Applied Economics, 2(3), 205-227. link B Moderate
  22. Kane, T. J., Rockoff, J. E., & Staiger, D. O. (2008). What does certification tell us about teacher effectiveness? Evidence from New York City. Economics of Education Review, 27(6), 615-631. link B Moderate

Every source for this factor is listed on the educational factor page.