On a certain level, it makes intuitive sense that your child’s public school principal does everything within her power to improve the standardized test scores of her students. Principals are personally invested in your child’s education, after all. Or at least they should be. But exactly how invested they are became a question that gripped Kellogg School of Management professor Michael Mazzeo after he attended a seminar by Julie Berry Cullen of the University of California, San Diego.

Cullen, an expert on the economics of education, presented research that documented different ways school administrators try to pull their students’ test scores up a few percentage points. They may offer more nutritious lunches on days the tests are given or take disciplinary actions, such as suspensions, to ensure that poor test-takers are not present on the Big Day.

But why on earth would they do this, Mazzeo wondered. What’s in it for them?

“I had a sneaking suspicion that money was involved,” recalls Mazzeo, who is a professor of management and strategy. The economics of education was an entirely new field for him; he typically studies industrial organization. But he was so intrigued, he dove into the literature.

Eventually, Mazzeo found that his sneaking suspicion about money was right. While principals do not receive performance-based pay increases when students improve their standardized test scores, Mazzeo and Cullen—who teamed up for the research—found that they do get an implicit reward.

Taking the Long View
“It’s not a direct reward, like a bonus paid to your salary, but a long-term career reward,” Mazzeo explains, “because the principals who have had successful schools have a better chance of getting higher-paying principal jobs in other districts, or jobs as superintendent of schools, that are much higher paid.”

The pair worked to get this result by picking apart a treasure trove of statewide public school data from Texas that spanned the years from 1987 to 2006. “Texas happens to keep really organized, well-detailed records,” Mazzeo says, “which made it really easy for us to track campus-level testing scores and to couple that with employment and wage information of principals from the same schools.” Texas also dictated that standardized tests be given throughout the state’s public school system, which provided a built-in uniformity to campus scores within the dataset. Mazzeo and Cullen were even able to track individual principals as they moved from school to school or district to district.

Mazzeo and Cullen searched for links between improved campus test scores and implicit payoffs to principals over time. They linked what Mazzeo calls two disparate sets of literature: one on the economics of education and the other on the economics of C.E.O. compensation.

“When you think about it, principals are really analogous to C.E.O.s,” Mazzeo says. “They set curriculum, they hire teachers.”

Principals set the tone for how to implement state- or district-wide curriculum, they motivate the teachers, and they lead the schools. Mazzeo studied the C.E.O. literature and realized that one weakness was that performance measures typically varied across firms, making it difficult to link performance and wage data across the board. But he found that by treating Texas’s principals as pseudo-C.E.O.s, he had standardized performance data in the form of the campus-level test scores as well as solid data on the principals’ wages. This meant that he and Cullen could look at what happened to principals’ compensation levels one, two, three, and five years after they ratcheted up the test scores at their campuses.

They found that, while principals’ wages did not increase if they stayed at the same schools, principals from improved schools often got higher-paying jobs when they became principals in different districts or advanced into district-level administrative jobs—including assistant or superintendent of schools—which paid significantly more. In 2006 the median wage of a school principal in Texas was $69,872, while the median wage of a district-level administrative staffer was $89,916 and a superintendent of schools was $91,340 (90th percentile = $133,910). So for principals who were willing to move around, the educational career ladder could be lucrative—if they had a record of improving campus-level test scores.

“Sticky” Wages
This might sound like a Catch-22 when viewed from the parents’ perspective because you want a high-performing principal to stick around. Mazzeo says it is a common finding in the labor market literature that wages are “sticky” when a worker stays with an employer, but often increase when they move to a new employer.

Cullen and Mazzeo also found that the reverse was true. Principals who had a record of overseeing schools whose test scores dropped were often “demoted” to become principals of schools in lesser-paying districts or to administrative jobs that paid less.

Prior to this study, Cullen says that the labor markets of principals were poorly studied mainly because of a lack of longitudinal data, or information that tracks back through time. Some of the results surprised her.

“I was surprised that there is as much cross-district mobility as there is,” she says. “Given that it is such a flexible market, I find it less surprising that effective principals are rewarded.” She had assumed that the labor markets would also be more internal to each district, so the fact that competition stretched across districts was also intriguing.

“Our findings show that the principal labor market is one mechanism that disciplines public schools, providing administrators with more incentive to put forth effort than they otherwise would have,” Cullen says.

Cullen and Mazzeo used a method that compared a school’s predicted test scores (based on individual students’ past performances, sociodemographic data, and other factors) to their actual test scores to judge whether principals were performing well or poorly. Their method winnowed down the variables that suggested certain schools were just good or bad, independent of which principals were leading them at the time of the data capture.

Mazzeo says it is not clear from their results that offering higher wages would cause a principal to perform better. “Rather, it’s the opportunity to do better somewhere else that motivates in itself,” he explains. “This profit motive creates a very competitive labor market internally and externally within schools and districts.”

Related reading on Kellogg Insight

On the Origin of Schools: The diversity of Arizona’s charter schools

Matriculation Matters: Refining the college admissions guessing game