Merit pay

Based on Wikipedia: Merit pay

In 2009, Nashville tried something radical. The city offered middle school math teachers a chance to earn bonuses of up to fifteen thousand dollars—real money, enough to change someone's year—if their students showed significant gains on standardized tests. The experiment ran for three years, carefully designed by researchers at Vanderbilt University. And when the results came in?

Nothing.

Students whose teachers were eligible for the bonuses performed no better than students whose teachers weren't. The hypothesis that dangling money in front of teachers would cause test scores to rise simply wasn't confirmed by the data.

This is the paradox at the heart of merit pay in education: an idea that sounds intuitively obvious—pay people more for doing better work, and they'll do better work—keeps running into a reality that refuses to cooperate.

What Merit Pay Actually Is

Merit pay goes by several names: pay for performance, performance-based compensation, incentive pay. The core idea is simple enough. Instead of paying everyone the same salary for the same job title, you measure how well people perform and reward the top performers with bonuses or higher salaries.

In the private sector, this often means salespeople earning commissions based on revenue they generate, or executives receiving bonuses tied to stock performance. The connection between individual effort and measurable outcome feels relatively clear, even if it's not perfect.

But when you try to apply this logic to teachers, things get complicated fast. What exactly are you measuring? Test scores? But a single test on a single day can't capture a year of learning. Student growth? But students arrive with vastly different starting points, home situations, and challenges that have nothing to do with their teacher. Principal evaluations? Those can be subjective, inconsistent, and influenced by politics.

The theoretical foundation of merit pay draws from behavioral psychology and incentive theory—the belief that humans are fundamentally rational creatures who respond predictably to rewards and punishments. Give someone a reason to work harder, and they'll work harder. The right incentive becomes a catalyst for improved performance.

It's an elegant theory. The trouble is that teaching might not be the kind of work where this theory applies cleanly.

The Teacher Advancement Program

One of the most ambitious attempts to make merit pay work in education is the Teacher Advancement Program, known as TAP, created by the Milken Family Foundation in 1999. By the mid-2000s, TAP had spread to more than one hundred eighty schools across the United States.

TAP works differently from simple bonus schemes. Yes, salary increases are tied to performance—measured through a combination of classroom observations and student test scores. But the program also creates new career paths for teachers who don't want to leave the classroom entirely.

Under TAP, teachers can advance in three ways. They can remain in the classroom and become mentors, guiding newer teachers. They can become master teachers, stepping out of daily instruction to support and coach their colleagues. Or they can pursue traditional administrative positions like assistant principal or principal.

But perhaps the most distinctive feature of TAP is its focus on collaborative learning. Teachers are organized into small groups that meet for a few hours each week to discuss what's working in their classrooms. The goal isn't just to reward individual performance—it's to help teachers improve by learning from each other.

Bell Street Middle School in South Carolina adopted TAP in 2001 and became something of a poster child for the program. The percentage of students scoring at advanced levels in math and reading doubled. The number of students scoring below basic level in math dropped by forty-six percent. And teacher turnover—a constant drain on school resources and consistency—fell from thirty-two percent to just ten percent.

These are impressive numbers. But there's a catch.

TAP is expensive. Schools pay between two hundred fifty and four hundred dollars per student per year to participate. Many TAP schools rely on grants to cover these costs—which raises obvious questions about sustainability. What happens when the grant money runs out?

And when researchers at Vanderbilt University compared TAP schools to non-TAP schools, the results were mixed at best. Elementary schools using TAP seemed to show some benefits. But in grades six through ten—middle school and early high school—TAP schools actually performed worse than comparable non-TAP schools.

The same program that doubled advanced scores at Bell Street was producing negative results elsewhere. Education, it turns out, resists simple formulas.

Denver's Experiment

In 2004 and 2005, Denver tried a different approach. The city's Professional Compensation System for Teachers, called ProComp, was approved by teachers themselves and financially backed by Denver voters—a notable achievement in a field where teachers' unions have often resisted performance-based pay.

ProComp gave teachers nine different ways to increase their earnings. They could earn more by working in high-needs schools that struggled to attract staff. They could earn bonuses for exceeding expectations on state exams. They could set professional development objectives at the beginning of the year and receive additional pay for meeting them. Positive evaluations from principals counted. So did helping their school achieve "distinguished" status through various metrics including parent satisfaction.

This multi-faceted approach was designed to avoid some of the obvious pitfalls of pure test-score-based pay. Teachers weren't just being judged on whether their students passed a standardized test—they were being evaluated across a range of professional behaviors and outcomes.

A 2010 evaluation of ProComp found encouraging signs. District-wide mathematics and reading achievement increased substantially in the years after ProComp's implementation compared to the period immediately before. Teachers hired after ProComp went into effect showed higher first-year achievement than teachers hired before the program.

But this raises a question that haunts all merit pay research: was the improvement caused by the pay incentives themselves, or by the broader culture change that accompanied them? Denver had made a citywide commitment to improving education. Voters had agreed to fund it. Teachers had bought in. The money was just one piece of a larger transformation.

The Federal Push

In 2006, the United States Congress created the Teacher Incentive Fund, allocating six hundred million dollars in federal grants to support performance-based compensation systems. The program, known by its acronym TIF, was expanded in 2009 with additional funding from the American Recovery and Reinvestment Act—the massive stimulus package passed in response to the 2008 financial crisis.

According to the Department of Education, TIF was designed to support four main goals. First, improve student achievement by making teachers and principals more effective. Second, reform compensation systems so that educators are rewarded for increases in student performance. Third, increase the number of effective teachers serving poor, minority, and disadvantaged students, particularly in subjects where schools struggle to find qualified staff. Fourth, create sustainable performance-based compensation systems that could continue without federal support.

That fourth goal—sustainability—has proven elusive. Many of the pilot programs funded by TIF and similar initiatives have struggled to continue once the grant money disappeared. Building a new compensation system is one thing; maintaining it with local funds is quite another.

Preliminary research on TIF-funded programs has shown the same mixed results that characterize the broader merit pay literature. Some programs showed positive effects on student achievement. Others didn't. The Nashville study, which found no impact despite substantial financial incentives, was itself partially supported by TIF-related funding.

The Effort Problem

Here's a question that gets to the heart of the merit pay debate: when teachers are offered bonuses for improved student performance, and performance doesn't improve, what exactly does that tell us?

The most obvious interpretation is that teachers were already working as hard as they could. They weren't holding back effort, waiting for someone to offer them extra money. When the money arrived, there was no additional effort to unlock.

This interpretation challenges a core assumption of incentive theory—the idea that people need external motivation to perform at their best. Perhaps teachers, at least many of them, are already intrinsically motivated. They became teachers because they wanted to help children learn. Adding financial incentives doesn't create new motivation; it just changes the framing of motivation that already exists.

But there's another interpretation worth considering. Maybe the problem isn't effort at all. Maybe it's selection.

When economists talk about selection effects in labor markets, they're asking: who chooses to enter a profession, and who chooses to stay? The current teaching profession developed under a compensation system that pays everyone roughly the same based on years of experience and credentials. This system might attract people who value stability and predictability over the chance at higher earnings.

A different compensation system—one where top performers could earn significantly more—might attract a different pool of candidates. People who believe strongly in their own abilities might be more likely to enter teaching if they knew exceptional performance would be recognized and rewarded. And the best performers, currently earning the same as mediocre colleagues, might be more likely to stay in the profession rather than leaving for fields where their talents command higher pay.

These selection effects wouldn't show up in short-term studies of existing teachers. They would only become visible over time, as the composition of the teaching workforce gradually shifted.

Economist Ludger Woessmann has examined international data on countries that do and don't use performance pay for teachers. His research suggests that performance pay does improve educational outcomes when viewed across countries—a finding that could reflect long-term selection effects rather than short-term effort effects.

Similarly, research by Eric Hanushek on the impact of dismissing ineffective teachers can be understood as a special case of performance pay. When pay is reduced to zero—meaning you lose your job—for teachers who don't perform well, the selection effect is immediate and dramatic. Teachers who can't meet standards leave the profession, replaced by others who might do better.

Why We Hate Losing More Than We Love Winning

In 1979, two psychologists named Daniel Kahneman and Amos Tversky proposed an idea that would eventually help earn Kahneman a Nobel Prize in Economics. They called it loss aversion.

The insight is counterintuitive at first. Most economic theory assumes that a dollar is a dollar—gaining ten dollars should feel exactly as good as losing ten dollars feels bad. But Kahneman and Tversky found that people don't work this way. Losses hurt more than equivalent gains feel good. We hate losing what we have more than we enjoy acquiring something new.

This psychological quirk shows up everywhere once you start looking for it. In one famous experiment, researchers gave some participants coffee mugs at random. Then they asked the mug owners how much money they would need to give up their mug, and asked non-owners how much they would pay to acquire one.

If a mug is just a mug, these numbers should be roughly the same. But they weren't. Mug owners demanded about seven dollars to part with their possession. Non-owners were willing to pay about three dollars to acquire one. Simply owning the mug made it more valuable—or rather, made losing it more painful.

Terrance Odean, an economist at Berkeley, examined the trading records of ten thousand investors at a large discount brokerage. He found that investors consistently held onto losing stocks too long and sold winning stocks too quickly. This behavior made no financial sense—the winning stocks they sold continued to outperform the losing stocks they kept. But it made perfect psychological sense. Selling a losing stock meant admitting you had lost money. Holding onto it meant you could still tell yourself the loss wasn't real.

What does any of this have to do with merit pay?

Consider how performance-based compensation is typically structured. Teachers are offered the chance to earn bonuses—extra money on top of their base salary. This frames the incentive as a potential gain. But loss aversion suggests that potential gains don't motivate people as strongly as potential losses.

Some researchers have experimented with flipping the script. Instead of promising teachers a bonus at the end of the year if their students perform well, what if you gave teachers the bonus at the beginning of the year and told them they'd have to give it back if students didn't improve?

Economically, these two scenarios are identical. In both cases, teachers end up with extra money if students do well and without it if students don't. But psychologically, they're completely different. The fear of losing something you already have is a more powerful motivator than the hope of gaining something you don't.

The Case For

Despite the mixed evidence, advocates for merit pay in education make several arguments worth considering.

The most straightforward is that merit pay leads to better educational outcomes. Proponents point to success stories like Bell Street Middle School or the positive trends in Denver's ProComp evaluation. Even if results are inconsistent, they argue, the programs that work demonstrate that performance-based compensation can improve teaching and learning.

A second argument focuses on teacher recruitment and retention. Low-performing schools in poor neighborhoods often struggle to attract and keep good teachers. Why would an excellent teacher choose to work in a challenging environment when they could earn the same salary at a well-resourced suburban school? Merit pay, especially when combined with bonuses for working in high-need schools, might help level this playing field.

Related to this is the argument about retention. Teaching has notoriously high turnover in the early years—many new teachers leave the profession within their first five years. If exceptional teachers saw their excellence reflected in their paychecks, they might be more likely to stay.

Advocates also argue that merit pay shows teachers their work is valued. In most current salary systems, a transformative teacher who changes students' lives earns the same as a mediocre teacher who does the minimum required. Performance-based pay sends a message that excellence matters and will be recognized.

Finally, there's the straightforward behavioral argument: people work harder when they see a connection between their effort and their reward. This principle seems to work in many other fields. Why should teaching be different?

The Case Against

Critics of merit pay offer equally compelling counterarguments.

Perhaps the most frequently cited concern is the impact on teacher morale and collaboration. Teaching at its best involves cooperation—sharing lesson plans, discussing difficult students, mentoring newcomers. If teachers are competing with each other for a limited pool of bonus money, this collaborative culture could erode. Why help a colleague succeed if their success comes at your expense?

There's also the question of cost—not just the bonuses themselves, but the administrative infrastructure required to measure performance, calculate rewards, and handle appeals from teachers who feel they were evaluated unfairly. This bureaucratic overhead might consume resources that could otherwise go directly to students.

Critics point to the research evidence, particularly studies like the Nashville experiment, as proof that the intuitive logic of incentives doesn't apply cleanly to education. After decades of experimentation, there's still no consensus that merit pay actually improves student outcomes.

The collaborative concern deserves elaboration. In many schools, experienced teachers informally mentor new ones, share materials they've developed, and help colleagues troubleshoot problems. This culture of mutual support benefits everyone, especially students. But if teachers feel they're competing for bonuses, they might hoard their best ideas rather than sharing them. The school becomes less than the sum of its parts.

Finally, there's a darker worry: that tying teacher pay to test scores creates incentives for cheating. The most notorious example is the Atlanta Public Schools scandal, where dozens of educators were convicted of falsifying test results under pressure from a high-stakes accountability system. When enough money and reputation ride on test scores, some people will find ways to game the system rather than doing the hard work of actually improving instruction.

The Political Divide

Merit pay for teachers has become a predictable point of partisan disagreement in American politics, though the positions don't always line up as you might expect.

In 2004, Florida Governor Jeb Bush enacted a program offering five percent bonuses to teachers ranked in the top twenty-five percent statewide, with selection based heavily on student score improvement. Houston implemented a similar program around the same time, without a cap on how many teachers could qualify.

Teachers in both states pushed back, arguing that a year of instruction couldn't be fairly measured by a single test on a single day. Their concerns proved prophetic in several ways. One Florida county ran out of money before all eligible teachers received their bonuses—a failure of basic program design. Accusations of discrimination emerged when predominantly Black schools, serving students with greater challenges, received fewer bonuses than predominantly white schools serving more advantaged populations.

President Barack Obama complicated the partisan narrative by expressing support for performance-based teacher pay while criticizing what he called "arbitrary tests." He sought to work with the National Education Association—the nation's largest teachers' union—to develop better metrics.

The NEA itself has historically opposed many forms of merit pay. In 2003, NEA President Reg Weaver said teachers understood that "politically motivated panaceas such as merit pay and eliminating tenure do nothing to improve teacher quality." The union expressed openness to alternatives but remained skeptical of "quick fixes designed to weaken the voice of teachers."

In 2011, Obama took executive action to reform aspects of No Child Left Behind, the federal education law that had created strong pressure on schools to raise test scores. Most notably, his administration offered waivers from the law's requirement that all students reach "proficient" levels in reading and math by 2014—a deadline that most observers agreed was impossible for many schools to meet.

To qualify for these waivers, states had to demonstrate adoption of "college and career ready" academic standards and present plans for transforming their lowest-performing schools. States like Georgia, Kentucky, Wisconsin, and Colorado were early candidates. Others held back, either unwilling to disrupt existing accountability systems or hoping Congress would eventually reauthorize the law with more realistic requirements.

Beyond Education: Government Reform

Merit pay debates extend beyond schools into the broader question of how to compensate government workers. In early 1993, newly inaugurated President Bill Clinton—facing recession and record federal debt—assigned Vice President Al Gore to lead a National Performance Review panel aimed at "reinventing government."

Working with David Osborne, co-author of the popular book Reinventing Government, Gore's panel developed recommendations for creating what they called Performance-Based Organizations, or PBOs. These would be discrete management units within the federal government committed to clear objectives, specific measurable goals, customer service standards, and targets for improved performance.

Under the PBO model, agency heads would be hired competitively rather than politically appointed, sign annual performance agreements with their supervising cabinet secretary, and have a portion of their pay tied to organizational results. The model drew inspiration from British reforms that, according to the panel, had improved performance and cut administrative costs.

But the panel was careful to note that not all government functions were suited to this approach. "Operations that do not have clear, measurable results should be excluded," the Blair House Papers explained. Foreign policy, for example, resists simple metrics. Success in diplomacy might mean preventing wars that never happen—how do you measure that?

This caveat gets at something important about merit pay in any context. It works best when you can clearly define and measure success. In sales, that's revenue. In manufacturing, that's units produced with acceptable quality. But in fields where outcomes are complex, multidimensional, and partly dependent on factors outside anyone's control, measuring merit becomes much harder.

What We Still Don't Know

After decades of experimentation, debate, and research, the honest answer about merit pay in education is that we still don't have definitive answers.

We know that short-term experiments offering bonuses to existing teachers haven't consistently improved student outcomes. The Nashville study is perhaps the cleanest demonstration of this, but it's not alone.

We don't know as much about long-term selection effects—whether different compensation structures would attract different people to teaching in the first place, or retain the best teachers who currently leave for higher-paying fields.

We know that program design matters enormously. TAP's combination of performance incentives with collaborative professional development looks quite different from a simple bonus-for-test-scores scheme, and the outcomes seem to differ too.

We know that context matters. The same program might succeed in one school and fail in another, depending on leadership, culture, community support, and dozens of other variables that are hard to control or even identify.

We know that measurement is hard. Student test scores capture something important, but not everything important. Principal evaluations bring different information but introduce subjectivity. Multi-metric systems like Denver's ProComp try to balance these concerns but add complexity.

Perhaps most importantly, we know that education reform is rarely about finding the one right answer. Schools are complex systems embedded in communities, shaped by politics, constrained by funding, and populated by humans with all their irrationality and variation. Approaches that work somewhere won't work everywhere. Promising ideas will have unintended consequences.

Merit pay might be part of improving education in America. It probably isn't a silver bullet. The work of teaching and learning is too complicated for any single policy to transform.

And that Nashville study, with its null results after three years of careful experimentation? It didn't prove that merit pay can never work. It proved something perhaps more valuable: that our intuitions about incentives and human behavior, however logical they seem, need to be tested against reality. Sometimes reality doesn't cooperate. When that happens, the question isn't whether to abandon the idea entirely or push forward regardless. It's whether we're willing to keep learning.