Input-Based Teacher Incentives: Chile’s Carrera Docente
Introduction
In 2016, Chile implemented Carrera Docente, a nationwide teacher pay-for-performance reform that—unlike most performance-pay systems, which reward gains in student test scores—ties compensation to input-based measures of teaching quality: a standardized subject-and-pedagogy exam (STEI) and a portfolio of classroom practice (Portfolio). Teachers are assigned to one of five performance categories, each associated with different salary premiums, ranging from 1–7% for entry-level teachers to 10–16% for those in advanced categories. Using administrative data on more than 13,000 teachers and 400,000 students, Araya-Córdova et al. (forthcoming) exploits the staggered timing of teacher evaluation cohorts in a difference-in-differences design. The results show small and statistically insignificant average effects on teaching quality and student test scores, but positive and significant effects among teachers facing the strongest marginal incentives to exert effort. To assess the welfare implications of Carrera Docente and alternative policy designs, the paper develops and estimate a structural model of teacher effort choices and use it to compute the Marginal Value of Public Funds (MVPF) under both the observed reform and a range of counterfactual policies.
MVPF = 0.5
Net Cost
The net fiscal cost equals the increase in teacher compensation under Carrera Docente, net of the additional tax revenue generated by higher teacher and student earnings:
- Provision cost (PC): The increase in teacher compensation relative to the salary schedule that would prevail in the absence of performance incentives. Under the baseline Carrera Docente policy, the provision cost is $1,774 per teacher.
- Additional revenue (AR): The tax revenue generated by the combined increase in teachers’ earnings and students’ predicted lifetime earnings, calculated using a marginal tax rate of (35%), following Kline and Walters (2016). Under the baseline policy, this additional revenue is $823 per teacher.
- Net fiscal cost: (PC-AR=$1774-$823=$951) per teacher under the baseline policy.
The paper also computes net fiscal costs for two counterfactual reforms designed to equalize marginal incentives across teachers. Policy 1, which removes experience and progression requirements, has a net fiscal cost of $2,295 per teacher. Policy 2, which replaces the existing system with a linear pay-for-performance schedule, has a negative net fiscal cost of −$771 per teacher. Thus, Policy 2 generates a fiscal surplus: the additional tax revenue attributable to its larger effects on student achievement exceeds the increase in teacher compensation.
Willingness to Pay
Willingness to pay (WTP) is the sum of students’ and teachers’ valuations of the policy, each measured relative to a status quo with no monetary incentives to improve Portfolio or STEI scores:
- Students’ WTP: This component is approximated by the predicted increase in after-tax lifetime earnings. The calculation assumes a 40-year earnings horizon, a 3% discount rate, and—following Kline (2014)—that a one-standard-deviation increase in test scores raises earnings by 10%. Earnings gains are measured net of a 35% marginal tax rate. Under the baseline Carrera Docente policy, students’ WTP is $376.
- Teachers’ WTP: This component is the compensating variation implied by the estimated structural model—that is, the amount of income a teacher would be willing to forgo while remaining indifferent between the incentive policy and the no-incentive baseline. It therefore incorporates both the additional compensation received and the disutility associated with the extra effort required to obtain it. Under the baseline policy, teachers’ WTP is $112, considerably less than the associated $1,774 increase in compensation because much of the income gain is offset by the utility cost of additional effort. Excluding teachers’ WTP would reduce the overall MVPF by roughly 20%.
- Overall WTP: Using the underlying unrounded estimates, total WTP is approximately $487.
Teachers’ WTP varies substantially across the counterfactual reforms. It increases to $258 under Policy 1 but becomes negative (−$159) under Policy 2. The linear compensation schedule in Policy 2 induces large increases in effort whose utility costs exceed the value teachers assign to the resulting income gains.
MVPF Summary
An MVPF of 0.51 implies that Carrera Docente, as implemented, generates approximately 51 cents of welfare for each dollar of net fiscal expenditure. This value lies below the benchmark MVPF of one associated with a lump-sum transfer, indicating that an equal-cost cash transfer to teachers would have generated greater welfare under the model’s assumptions. The main reason is the mismatch between the policy’s average and marginal incentives. Although the reform produces sizable increases in teacher compensation, its discrete performance categories leave many teachers with little marginal incentive to exert additional effort. Consequently, a substantial share of the compensation is inframarginal and generates limited improvements in teaching quality and student learning. Moreover, among teachers who increase their effort, the associated utility cost offsets part of the value of the additional compensation.
For comparison, the paper evaluates two counterfactual reforms relative to the same no-incentive baseline. Removing experience and progression requirements increases the MVPF to 0.63. Replacing the categorical system with a linear pay-for-performance schedule produces negative net fiscal costs, implying that the policy pays for itself and therefore has an unbounded MVPF under the conventional definition.
References
Araya-Córdova, Patricio, Dante Contreras, Jorge Rodríguez, and Paulina Sepúlveda (forthcoming). “Getting Teachers Back to Study: Input-Based Teacher Incentives and Student Achievement.” Forthcoming in the Journal of Labor Economics. https://www.dropbox.com/scl/fi/v1kurchom9tjn3rmsql7r/Teachers-JOLE-accepted.pdf
Kline, Patrick and Christopher R Walters (2016). “Evaluating Public Programs with Close Substitutes: The Case of Head Start.” The Quarterly Journal of Economics, 131(4), 1795-1848. DOI: https://doi.org/10.1093/qje/qjw027
Policy Features
- Category
- Education
- Sub-Category
- Child Education
- Beneficiary Type(s)
- Adults, Primary school enrollees
- Country of Implementation
- Chile
- Empirical Method
- Difference in Differences
- Research Type
- Primary
- Peer Reviewed
- Yes
- MVPF Publication Link
- www.dropbox.com/scl/fi/v1kurchom9tjn3rmsql7r/Teachers-JOLE-accepted.pdf