Big Picture
Pay better teachers more. As campaign promises go, it practically sells itself.
It sounds fair. It sounds demanding. It offers an appealing answer to a difficult problem: If Minnesota wants better results from its schools, reward the people who deliver them.
But who counts as a better teacher? Who decides? What would that teacher receive, and what would happen to someone who falls short? Most importantly, what would students gain?
Those questions separate a catchy slogan from a workable policy. Lisa Demuth has offered the former without explaining enough of the latter.
At her September 10 education announcement, Demuth proposed teacher performance pay because “better teachers deserve better pay,” unfortunately however, she could not identify a single metric that would determine the additional compensation.
While her plan offers a broad ambition to return Minnesota to the national top ten. It does not lay out a performance-pay formula, a funding amount or a measurable improvement attributable to changing teacher compensation. A statewide aspiration is not an explanation of what this particular policy would accomplish.
That omission matters because performance pay has a mixed record at best. The evidence includes failures and successes, depending on the intervention and setting. That mixed bag gives voters no reason to accept the pay-for-performance label as a guarantee for better results.
Minnesota Has Already Tried Performance Pay
Minnesota also has something more useful than a slogan to start with: experience.
The state’s Q Comp program dates to 2005. It allows districts and teachers’ bargaining representatives to develop voluntary plans combining performance pay with professional development, evaluation and career advancement. As a result, Minnesota has already spent years working through the relationship between compensation and teaching.
Demuth therefore owes voters an explanation of what she would change. Expand participation? Replace the existing approach? Change its measures? Make a voluntary arrangement mandatory? The available proposal does not answer those questions.
That is not proof Q Comp is sufficient. An existing program can need improvement. But a candidate promising reform should identify the shortcomings she intends to fix.
What the Research Shows
The research makes that obligation more pressing.
For example, in Nashville’s Project on Incentives in Teaching, middle-school math teachers could earn bonuses of up to $15,000 for student achievement gains. The experiment did not produce consistent, lasting improvements. Offering a substantial reward did not, by itself, reliably produce the hoped-for result. RAND’s account of the experiment
It would be convenient for critics to stop there but that would also be somewhat misleading.
A federal evaluation of the Teacher Incentive Fund found small positive reading and math effects from performance bonuses within a broader compensation system. The comparison schools had similar systems but paid a uniform bonus instead. That is evidence that incentives can make a difference under some conditions.
More recent research also provides a favorable case for compensation targeted at attracting and retaining effective teachers in disadvantaged schools. Of course, that is a different mechanism from expecting the same teachers to produce better results after announcing a bonus.
Minnesota’s own research further complicates the conflicting findings above. Researchers studying Q Comp adoption found modest reading gains, with less precise math findings. Perhaps most notably – they could not isolate which elements of the program produced the improvement as compensation was bundled with changes to teachers’ professional support and working practices.
In reality, the conclusion is more demanding than either side’s preferred talking point. Performance pay is not one policy with one predictable result. Its design determines what claim a candidate can reasonably make for it.
The Measure Is the Policy
Start with the measurement problem.
Imagine one teacher whose students arrive well prepared and mostly finish the year above a proficiency threshold. Another teacher’s students start much further behind, improve substantially, but remain below that line. A system rewarding proficiency could favor the first teacher. A system rewarding growth could reach a different conclusion.
Neither example tells us which teacher deserves a bonus. It shows why choosing the measure is the policy itself.
Then consider special education, kindergarten, music or career instruction. A compensation system needs a defensible way to recognize those teachers’ contributions, too. Classroom observations introduce their own questions about evaluator training and consistency. Statistical estimates do not eliminate uncertainty: the American Statistical Association has warned about the limitations of using value-added measures in consequential judgments about teachers.
The consequences require equal clarity. Would a low rating mean no bonus, slower salary growth or a reduction in pay? Demuth has not established those answers in the materials released to date. It would be wrong to claim she has announced pay cuts but it is entirely reasonable to demand that she explain the stakes before asking teachers and voters to embrace the idea.
Better Teaching Takes More Than a Bonus
And what would help a teacher improve? A reward can identify a destination without providing a route.
Time to plan, useful feedback, instructional materials and opportunities to learn from colleagues belong in the discussion. Q Comp’s results cannot establish that those supports always make incentives succeed. They do establish why a candidate cannot simply borrow a favorable finding about a comprehensive program and treat it as validation of an unspecified bonus.
Demuth’s broader plan does discuss instruction and tutoring. The missing explanation is how her compensation proposal would connect to that work—and what she would fund to make the connection real.
Someone also has to pay for the rewards and the system administering them. Would districts receive additional recurring money? Would evaluation take resources from other instructional work? Would bonuses compete with base salaries?
A modest benefit might justify its cost. A positive result alone does not establish that a proposal is the best use of the money.
Bottom Line
No candidate needs to arrive at a press conference carrying finished legislation. But the essential choices cannot all be deferred.
What changes? How is performance measured? What are the consequences? What improvement should families expect, by when, and at what cost?
Those are reasonable demands of anyone who wants to run Minnesota’s schools. Democrats defending existing programs or offering reforms of their own should meet them, too.
Demuth’s pitch asks voters to associate her with rewarding excellence. The policy she has explained does not yet earn that association. Teachers deserve fair compensation, students deserve effective instruction, and taxpayers deserve evidence that a reform can deliver.
“Better teachers deserve better pay” is an attractive line. Minnesota deserves to know whether there is a better policy behind it.