What happens when we pay public high school teachers based on performance

Original Article ↗ Hacker News Discussion ↗

Overview of the Study and Initial Reactions

  • Paper finds Wisconsin districts that switched from seniority-based to performance-based pay:
    • Attracted “high-quality” teachers and pushed out weaker ones.
    • Increased teacher effort and improved standardized test scores.
  • Some commenters find this intuitively appealing and consistent with “pay for performance.”
  • Others say the results “confirm my biases,” and may be too neat or zero-sum.

Test Scores, “Teaching to the Test,” and Goodhart’s Law

  • Many worry test scores are a poor proxy for true learning:
    • Incentivize drill, rote prep, and narrow curricula.
    • Risk turning schools into test-prep factories; some students learn to hate school.
  • Goodhart’s law is repeatedly invoked: once pay depends on scores, the metric is gamed.
  • Examples cited:
    • Teachers focusing on easier students and neglecting the hardest cases.
    • Selection of better-behaved, more advantaged classes.
    • Cheating on standardized tests.

Inequality and Talent Redistribution Across Districts

  • Concern that early adopters of flexible pay might just poach good teachers from other districts, not increase total quality.
  • Debate over whether rich or poor districts are more likely/able to implement such schemes:
    • Poor districts may want change but lack funds.
    • High per-pupil funding can be eaten by bureaucracy, leaving little for salaries.
  • Several note this could reinforce existing SES-based inequalities unless overall teacher pay rises.

Incentives, Motivation, and Ethics

  • Disagreement over whether monetary rewards improve cognitive work:
    • Some cite experience (and research) that large extrinsic rewards can backfire.
    • Others argue incentives “work” but may distort behavior.
  • Anecdotes:
    • Performance pay tied to tests led to intense test prep and “widespread cheating.”
    • Underpaid teachers facing financial stress may be more susceptible to gaming incentives.

Measurement Problems and Alternatives

  • Strong skepticism that standardized tests capture teacher “quality,” especially in challenging classrooms.
  • Suggestions and questions:
    • Longitudinal outcomes rather than single-year scores.
    • Recognition that even “bad” teachers may be better than no teacher.
    • Some argue focusing on reducing poverty and class sizes, and raising baseline pay, would do more than complex pay schemes.