How much say should the public have about public education? Gaps between public preferences and the practice of public education have been around for decades.
In my experience, for example, most non-educators believe teacher pay ought to vary according to demonstrable performance in the classroom. Most public-school teachers I know dislike this idea. Recent polling confirms the gap: in a 2019 survey, 72% of the general public supported “basing part of the salaries of teachers on how much their students learn.” Only 42% of teachers agreed.
This does not mean, strictly speaking, that teachers oppose merit pay. It depends on how the term is defined. One might assume that more-experienced teachers tend to be more effective than less-experienced ones, or that teachers with more formal education, as signified by graduate degrees, tend to produce better results.
Most teachers do, indeed, assume these propositions are true, and overwhelmingly support pay boosts based on years of experience and academic credentials.
In truth, the empirical link between experience and effectiveness is not perfectly linear (you get bigger gains from experience earlier in a teacher’s career), and the link between graduate degrees and effectiveness is largely a fiction (except for certain high-level math and science classes). But if you assume the propositions to be true, then you think paying more for experience and graduate degrees constitutes merit pay.
Now, is the general public right to assume that teacher quality can be directly measured, by value-added test scores or supervisors or some combination? If so, we need not use roundabout means (years on the jobs or degrees earned) to identify and reward high performance. We can use performance-based pay to increase average teacher quality in at least three ways, by 1) incentivizing teachers to perform at their highest capacity, 2) encouraging high-performing teachers to stay in the profession, and 3) nudging low-performing teachers to exit the profession.
My guess is that the second and third mechanisms are more important than the first. You probably don’t care about my guesswork, however. Is there any hard evidence for merit pay?
Yes. Independent researchers have studied these questions for decades. Some answers remain tentative and hard to interpret. Other results are more immediately useful. In general, they find that merit pay improves teacher quality, although it depends on the specific program studied and the specific variables assessed.
In early 2020, the American Educational Research Journal published a meta-analysis that brought together findings from 37 different studies, including 26 papers that examined teacher-pay programs in the United States. The researchers found that the use of merit pay was associated with statistically significant gains in student test scores.
That doesn’t mean that every model produced positive gains. Nor were those positive gains gigantic. That’s okay — moderate gains in student learning are hard enough to accomplish in education reform, and can accumulate into substantial benefits over time.
Delving more deeply into the meta-analysis, the researchers found that student gains were larger in math than in language, larger in elementary school than in higher grades, and tended to be larger in the early years of implementing performance pay. When the pay boost teachers could earn was bigger, the benefits for students were bigger. And when pay boosts were predicated on the performance of groups rather than of individual teachers, the benefits for students disappeared.
While this paper combines findings from dozens of prior studies, it shouldn’t be treated as the final word. Indeed, there is no final word when it comes to research, whether in the social sciences or any other discipline. We should always be open to the possibility that future studies will refine our understanding of an inherently complicated issue.
At the moment, however, the North Carolina General Assembly and other policymakers should continue to enact and implement performance pay for teachers. The public has this one right.