Updated May 21, 2025

School spending is up. So why aren’t teacher salaries? 

For example, from 2002 to 2022, per-pupil spending in the Los Angeles Unified School District rose 108%. That’s in real terms, after accounting for inflation. 

But that money didn’t lead to higher pay. In fact, the average salary earned by district employees rose just 5%. 

These trends are not unique to Los Angeles and are, in fact, playing out in most schools and districts across the country. 

We set out to document the disconnect between rising school spending and stagnant salary levels. 

Nationally, average teacher salaries have been remarkably flat for a very long time. In inflation-adjusted terms, they’ve been around $70,000 for decades.  

Meanwhile, pay for other college-educated workers has risen steadily, leaving teachers behind. For example, an analysis last year found the average nurse made less than the average teacher in the 1970s, but nursing pay has since then while teacher salaries have not. Compensation for other professionals, including accountants, engineers, college professors, doctors, health technicians, managers, officials, proprietors, lawyers, judges and scientists, has pulled even farther ahead.

The most obvious explanation for stagnant teacher salaries would be if total education spending were flat. But that’s not it, because America is investing more in schools. 

The graph below compares the growth in school spending versus employee salaries for the last two decades. Both figures are adjusted for inflation. But, while spending rose 31% per student, the average salary paid to district employees fell by 2.5%. 

If district salaries had merely kept up with total education spending, they would have been 34% higher. That would have worked out to nearly a $22,000 raise for the average employee. 

I’m not the first person to document this. Last year, the libertarian Reason Foundation published looking at state-level trends. It found that inflation-adjusted, per-pupil spending had risen across the country and in every state except North Carolina. And yet, there was not a single state where teacher salaries kept up with the pace of overall spending. 

In my home state of , per-pupil school spending rose by 15% from 2002 to 2020 while teacher salaries fell by 4%. In , spending rose 36% while teacher salaries increased by just 8%. In , school spending skyrocketed a whopping 70% while average teacher salaries rose by a more modest 16%. In , school spending rose by 16% while teacher salaries crept up just 1%. Again, these figures are all adjusted for inflation. 

What about individual school districts? I worked with Eamonn Fitzmaurice, The 74’s art and technology director, to look at local trends. 

Unfortunately, there’s no national database of average teacher salaries by district, but we got pretty close. The NCES Common Core of Data collects the total salary expenditures per district and the total number of staff employed. By dividing these figures, we calculated an average salary across all employees in the district. These are not “teacher” salaries, but we think they’re a reasonable approximation. Particularly large changes in the growth of school spending versus employee salaries could be due to a variety of factors, including rapid increases in revenues or decreases in student enrollments, governance changes or data input errors.

We looked at revenue and salary expenditures from 2001-02 to 2021-22, the last year for which the data were available. We adjusted everything for inflation and took out districts with fewer than 500 students or missing data. That left 8,877 districts that educate about 90% of students nationwide. Use the interface below to see how per-student revenue and salary trends are changing in your community.  

INTERACTIVE

School District Revenue vs Salary Change 2002-2022

For some districts, about 10% of our sample, salaries rose at a rate commensurate with district revenue increases. Washington state districts were disproportionately represented on this list, thanks in part to in state funding explicitly tied to teacher salaries. 

But that means 90% of districts did not raise salaries in proportion to their revenue increases. This disconnect may help explain why some teachers feel their salaries aren't keeping up with their expenses. 

Take housing costs, for example. Imagine someone living on a teacher’s salary in Santa Monica, California, where housing costs are some of the nation's highest. While policymakers have delivered on the budget side, and the total amount of money allocated to the Santa Monica-Malibu Unified School District has mirrored in housing prices, that money isn’t translating into higher salaries. The district average actually fell by 19% over the last two decades, making it more and more difficult for employees to live in the city in which they work. 

These trends are playing out across the country. In Billings, Montana, per-student revenue rose 51% while average salaries declined 32%. In Philadelphia, revenue per-student climbed by 155% while average salaries fell by 8%. In Buffalo, New York, revenue rose 114% while salaries fell 14%. In Jefferson County, Kentucky, revenue per-student climbed 62% while salaries rose a more modest 12%. 

Where is the money going? The answers vary state by state and district by district, but the national trends provide at least some answers. 

The biggest factor is the number and type of staff. Schools employ a lot more people than they used to, meaning they have to divide their budgets across more workers. While enrollment rose 4% from 2002 to 2022, the number of full-time equivalent rose three times as fast, led by particularly large increases in instructional coordinators, classroom aides and district administrative staff. 

But these figures actually mask another trend. According to Census Bureau , schools employed about 200,000 fewer part-time workers in 2022 than they had in 2002. (Part of that was due to COVID.) Meanwhile, schools added 540,000 full-time workers. This shift carries real costs because two half-time employees don’t earn as much as one full-time employee, and part-time workers also don’t qualify for benefits like health insurance or retirement. 

And the cost of those benefits has risen rapidly. Since 2004, the Bureau of Labor Statistics has broken out the components of the costs for a school district to a teacher. In 2004, base salaries and wages represented 74% of a teacher’s total compensation package, the rest being a combination of health care benefits, retirement plans and Social Security contributions. Over time, total teacher compensation has grown faster than inflation, even though the salary component has not. Instead, benefits costs, especially , have increased rapidly, eating up a larger and larger share of district budgets. 

These changes have been slow and gradual over many years, but they have all played a role in stagnant salaries. State, district and school leaders should dig into these trends if they want to boost take-home pay in their communities.

Clarification: This story has been updated to note that particularly large changes in growth of school spending versus employee salaries could be due to a number of factors. In addition, we noted that salary calculations do not include additional compensation such as retirement or health benefit contributions. 

Few of them would think of Neshoba County in Mississippi.

But Neshoba County schools are doing something that those other places are not: They serve a high-poverty community, yet their students’ math scores are competitive with those in wealthier areas.

Back in September, I worked with Eamonn Fitzmaurice, The 74’s art and technology director, to find districts around the country that were doing the best job of helping kids learn to read proficiently by third grade. Today, we’re taking the same approach to eighth-grade math. We calculated each district’s expected math proficiency rate, based on its local poverty level, and compared that to its actual scores. This methodology helped us identify districts that are beating the odds in math. 

Select from the menu below to find the high fliers in your state.

INTERACTIVE

Eighth Grade Math Proficiency

At the national level, eighth-grade math scores peaked in 2013, were slipping leading up to the pandemic and then fell dramatically. The declines were particularly large for students who were already among the lowest-performing.

Mississippi weathered these declines better than most states. As a result, the found that Mississippi climbed the state rankings in both math and reading over the last decade. After controlling for student demographics, Mississippi was ahead of 40 other states by 2019, and its scores quicker than other states’ after COVID.

Neshoba County helped lead that rise. According to data from the at Stanford University, Neshoba’s students went from scoring more than half of a grade level below the national average in 2016 to nearly 1.5 grade levels above the national average last year. Their students made gains even in the midst of the COVID-19 pandemic.

When we started looking for districts that were beating the odds, we aimed to find and celebrate districts like Neshoba. We ultimately identified nearly 600 districts that are getting exceptional results in math, which we defined as significantly outscoring their expected eighth-grade proficiency rate.

Some districts are showing strong performance in both third-grade reading and eighth-grade math. For example, in the reading project we highlighted Steubenville City, Ohio, at the top of our rankings. Despite its relatively high poverty, 81% of its eighth graders score as proficient in math, which puts it on par with districts that have many fewer disadvantaged students. 

States set their proficiency cut points at different levels, and Maryland has one of the highest bars. And yet, students in Worcester, a community that is neither high- nor low-poverty, stands out for having eighth-grade math proficiency rates 20 points higher than kids in any other district in the state.

In Michigan, Dearborn City is getting the same results as other districts with much lower poverty rates.

Other strong outliers include places like Genoa Central in Arkansas, Lake Washington in Washington state, the Fossil School District in Oregon and the Murray Independent district in Kentucky. 

In northern Virginia, where I live, people often say they move here for the schools. But if they were really looking for the best school system in the state, they would move to Wise County, on the Kentucky border. Wise County has much higher poverty rates than the more well-known D.C. suburbs, yet it topped our Virginia rankings in both reading and math.

Looking at the scores this way helps identify the places with great school systems, where learning gains are driven by what students learn in the classroom. This is especially true in math, because unlike reading proficiency — which is closely tied to language skills and background knowledge that children acquire at home — math scores are more directly linked to school-based instruction.

This gets at the heart of the issue at hand. Parents and policymakers should not be content with answering the simple question of, “Where do students do the best?” Wealthy communities are likely to look good by that standard, just by the nature of the students they serve.

Instead, policymakers should be trying to find schools and districts that help all students learn, regardless of their income levels. Poverty is certainly predictive of school performance, but it need not be determinative.

Note: For more details about the data sources and methodology for this project, see our earlier reading analysis.