ANALYSISApril 13, 2026
Do teachers unions help or hurt student outcomes?
The question of whether teachers unions help or hurt student outcomes remains an actively debated topic in education policy, with a substantial and growing academic literature producing mixed findings. Decades of research suggest the effects are neither uniformly positive nor negative, but depend heavily on district context, contract provisions, socioeconomic factors, and the time horizon studied. The two dominant union organizations — the NEA and AFT — continue to wield significant political and financial influence over education policy at the state and national level.
Unions have won teachers higher pay, better conditions, and job security — but critics say those same protections make it nearly impossible to remove failing teachers and block reforms that could help struggling students. Who does the classroom actually belong to: the adults who work in it or the kids who need it?
Not familiar with this story? Get context →
Spending rises, achievement doesn't
C
Marianno's research establishes the core indictment cleanly: stronger collective bargaining agreements produce higher spending and higher salaries, but not measurable gains in student achievement. That is not a performance gap — it is the institution functioning exactly as designed, optimizing for the adults inside it rather than the students passing through it. The money flows; the kids don't move.
L
You're describing unions as the cause of flat achievement, but the same data shows a profession already paying 73 cents on the dollar compared to equivalent college graduates — a record low despite decades of advocacy. If union power were actually strangling the system, teacher compensation would look very different. What Marianno shows is that unions aren't powerful enough to fix the structural underfunding, not that they're the obstacle.
C
The 73-cent figure actually breaks the liberal argument — if decades of aggressive union advocacy produced a record wage penalty, then union contracts are clearly not the mechanism for improving teacher compensation. What they do is compress salaries on seniority schedules that protect the bottom of the distribution while talented newcomers are systematically underpaid.
L
Seniority schedules are a legitimate target — but that's an argument for redesigning what unions bargain for, not eliminating the institution keeping the profession from collapsing entirely in a labor market that already pays 27 percent below comparable work.
Wisconsin short-run harm to vulnerable students
C
The Wisconsin Act 10 evidence is genuinely complicated and intellectual honesty requires sitting with it: a 20% of a standard deviation test score decline, concentrated in the lower achievement distribution — exactly the students conservatives claim to care most about. We will not bury that. But the pre-Act 10 baseline was already trapping those same students in classrooms with teachers who couldn't be removed and schools spending more without learning more. Accepting ongoing compounding harm to avoid short-run disruption is not a neutral choice.
L
You acknowledge the 20% decline but reframe it as disruption costs from dismantling a bad system — yet the American Journal of Education study shows the largest declines fell in districts serving high concentrations of Black and Hispanic students. You're asking the most vulnerable kids to absorb the adjustment costs of an experiment whose long-run payoff isn't guaranteed.
C
The 2025 decertification paper shows test scores up 7% of a standard deviation and attendance improved five years out — that long-run payoff is not theoretical anymore. The question is whether short-run harm to a pre-existing underserving system should permanently foreclose the reform.
L
Five years of positive results is a real signal, but it matters enormously whether those gains came in districts that simultaneously invested in competitive compensation to recruit replacements — because if they did, the lesson is 'flexible contracts plus real pay,' not 'weaken bargaining and wait.'
Teacher quality as first-order equity issue
C
Eric Hanushek has spent a career documenting that the gap between a highly effective and persistently ineffective teacher is equivalent to one to two years of learning. Every contractual mechanism that makes it nearly impossible to dismiss ineffective teachers is a direct redistribution of educational opportunity from low-income students — who cannot escape their assigned school — to the adults employed there. That is not a technocratic preference; it is a question of basic justice.
L
Hanushek's research is real and we shouldn't dismiss it — but the Ohio State finding you're citing describes districts that produced gains by *hiring more teachers*, not simply by removing union contracts. The mechanism driving achievement was increased staffing, which requires the candidates to be there in the first place. Weaken collective bargaining in a profession paying 27 percent below market and the pipeline you need to replace ineffective teachers shrinks.
C
You're assuming the union contract is what holds the pipeline together — but the same contracts that make dismissal nearly impossible are systematically underpaying the talented early-career teachers most likely to fill that pipeline. You cannot protect the institution as the workforce solution when the institution's own structure is driving those candidates out.
L
That's the strongest version of the conservative case, and it actually points to the same conclusion: the problem is lockstep salary schedules, not collective bargaining itself — which means the fix is renegotiating contract terms, not decertification.
Reform strategy versus demolition strategy
C
The liberal framing — 'redesign what unions bargain for' — assumes unions will voluntarily bargain away the tenure protections and seniority rules that are their primary source of institutional power. That has not happened in the decades available to try it. The Ohio State evidence is instructive: districts that actually produced achievement gains did so by operating outside collective bargaining pressure entirely, not by negotiating better terms within it.
L
You're treating decertification as the revealed solution, but the conservative rebuttal itself just acknowledged that flexible contracts need to be paired with competitive salaries to produce the long-run gains. That's not an argument against unions — that's an argument for unions that bargain for different things. The demolition strategy gets you the Wisconsin short-run harm without the investment that makes the long-run story work.
C
The burden of proof runs both ways here: liberals have had the institutional architecture they're defending in place for decades, and the outcome is a record wage penalty alongside flat achievement gains for the most vulnerable students. At some point 'reform from within' becomes the argument that protects the institution regardless of what it produces.
L
Fair — but 'reform from within has failed, so eliminate the institution' is only a valid move if the replacement produces better outcomes for the students currently being harmed, and the evidence that decertification alone does that, without compensating investment, is still thinner than reformers are willing to admit.
Turnover costs falling on minority students
C
The American Journal of Education finding that union reform produced larger test score declines in districts serving Black and Hispanic students is a moral cost the conservative argument must acknowledge, not minimize. But the counterfactual matters: those same students, in a system optimized to protect incumbent teachers, were already experiencing compounding harm invisible because it had no dramatic policy event attached to it.
L
The 'invisible compounding harm' framing is speculative — you're asking us to weigh a documented, measurable test score decline against a hypothetical baseline that was already bad. The documented harm fell on specific children in specific classrooms when experienced teachers left following Act 10. That is not an abstraction; it is a policy consequence with names and zip codes.
C
The five-year data shows those same districts recovering and improving — which means the disruption was real but not permanent, while the pre-reform harm was both real and permanent by design, because the contracts made it so.
L
Recovery in aggregate five-year data still obscures which students bore the transition costs — and if the answer is disproportionately Black and Hispanic kids in high-turnover districts, then 'it got better eventually' is cold comfort to the cohort that graduated during the dip.
Conservative's hardest question
The short-run Wisconsin evidence — a 20% of a standard deviation test score decline concentrated among students in the lower achievement distribution following Act 10 — is genuinely difficult to dismiss and cuts directly against the claim that reducing union power benefits vulnerable students in the near term. A policy framework that accepts real, measurable harm to disadvantaged children today in exchange for theoretical long-run gains demands a higher burden of proof than the current evidence fully satisfies.
Liberal's hardest question
The 2025 Wisconsin working paper finding test score increases of 7% of a standard deviation and improved attendance five years after local union decertification is genuinely difficult to dismiss — if the long-run mechanism of flexible hiring and compensation attracts better teachers, it suggests the liberal position may be protecting institutional arrangements that eventually harm the students we claim to defend. The honest answer is that the evidence on long-run effects is still thin, but it is not zero.
Both sides agree: Both sides accept that teacher quality is the single most important in-school determinant of student achievement, making the institutional rules governing how teachers are hired, retained, and removed a genuine first-order equity question rather than a procedural one.
The real conflict: The core factual dispute is causal: conservatives read the Ohio State hiring data and Marianno's salary-without-gains finding as evidence that union contracts structurally misdirect resources away from student outcomes, while liberals read the same data as evidence that the specific mechanism matters — staffing levels, not union absence, drove achievement — making the policy implication fundamentally different.
What nobody has answered: If the 2025 Wisconsin decertification paper's positive long-run findings required districts to simultaneously invest in competitive compensation to recruit replacement teachers, then the actual policy prescription is 'flexible contracts plus higher pay' — and neither side has explained how to finance that combination without the bargaining power unions provide, or without the accountability reforms unions resist.
Sources
- Search results summarizing academic literature on teacher unions and student outcomes, including studies by Bradley Marianno on collective bargaining and school spending
- Ohio State University study on union bargaining, salary allocation, and teacher hiring effects on student achievement
- Research on Wisconsin Act 10 effects on test scores, teacher turnover, and student achievement distribution
- 2025 working paper on long-term effects of local union decertification in Wisconsin on test scores and attendance
- American Journal of Education study on statewide union reforms, test scores, and racial disparities
- Economic Policy Institute data on the teacher wage penalty (2024)
- Research on NEA membership figures and founding history
- Eric Hanushek scholarship on teacher quality as the primary in-school determinant of student achievement