States are Bad at Giving Teachers Raises

Only pennies of new public-school dollars make it to teacher salaries

A woman holds open a wallet with no money

Teacher shortages have been a chief concern of public school leaders since the Covid-19 pandemic began in 2020. According to an August 2023 survey by the National Center for Education Statistics, 79 percent of public schools with at least one vacant teaching position reported difficulty filling slots for the 2023–24 school year.

The most commonly discussed solution to mitigating K–12 teacher vacancies is spending more on K–12 education so that teachers receive more competitive pay and the profession becomes more attractive to prospective hires. By this logic, the teacher shortage represents a failure of state legislatures to invest adequate dollars in education over the long run.

But for legislators hoping to improve teacher pay and attract more people to the profession, increasing education spending is surprisingly ineffective. Data published in Reason Foundation’s new study Public Education at a Crossroads indicates that states do a poor job of translating additional education investments into higher teacher salaries. Many states have lost ground on average teacher salaries over the past two decades despite spending more. Those states that do manage to raise teacher salaries largely do so only after investing exorbitant amounts in public education, with most of those dollars still being diverted to expenditures besides take-home pay for current teachers.


More money, lower salaries

Figure 1 displays national trends in inflation-adjusted total revenues, instructional salary spending, and instructional benefit spending per student from 2002 to 2020, as reported by the U.S. Census Bureau. Total K–12 revenues per student grew by 25 percent from $12,852 to $16,065 over that period. But at the same time, inflation-adjusted spending on instructional salaries hardly budged, increasing from $4,920 to $5,145 per student.

The very small share of new education revenues making it into instructional salaries looks even worse when considering that the Census Bureau’s instructional salary expenditures include spending on newly created positions and instructional aides. Public schools added nearly 200,000 teachers and 210,000 instructional aides from 2002 to 2020. But even when including spending on those new hires, increases in education revenues often don’t make their way into paychecks.

Using data on average teacher salaries from the National Center for Education Statistics, which excludes classroom aides and accounts for new teacher hires, the picture becomes bleaker. Despite education revenues increasing by 25 percent nationally from 2002 to 2020, inflation-adjusted average teacher salaries fell by 0.6 percent, from $64,522 to $64,133.

Figure 2 shows how national trends in total K–12 revenues and average teacher salaries play out at the state level. Despite every state except North Carolina increasing inflation-adjusted education revenues per student between 2002 and 2020, 26 states saw declines in their average teacher salary. These declines probably aren’t attributable to declines in average teacher experience, since NCES data indicates salaries nationally have also been flat for teachers with different levels of experience.

Generally, states with the largest increases in education funding have seen the largest increases in average teacher salary. However, states like Pennsylvania, Illinois, and Delaware saw notable declines in average teacher salary despite increasing education funding by more than $5,000 per student from 2002 to 2020. Conversely Washington, California, and Massachusetts increased teacher pay substantially while making comparable additional K–12 investments per student over the same period.


Just how bad are states at driving money toward instructional salaries?

But even in states like New York where large increases in education spending have corresponded with substantial teacher raises, it’s worth examining how instructional salary growth measures up to the size of additional investments those states have made. In other words, how much in additional instructional salary spending does an additional education dollar per student buy?

Returning to the Census data, Figure 3 illustrates how inefficiently each state is in converting a new education dollar per student into spending per student on instructional salaries (higher instructional salaries and new instructional hires, including instructional aides). Nationally, only seven cents of each new dollar devoted to education between 2002 and 2020 went to instructional salaries. Massachusetts led the nation in effectively translating new education spending into teacher take-home pay—but even in that case it was only 34 cents of each new education dollar per student.

This lens allows for better apples-to-apples comparisons between states that make varying amounts of new investments in public education. Nebraska, for example, ranks 33rd in the country in revenue growth between 2002 and 2020 and 29th in average teacher salary. But as Figure 3 demonstrates, Nebraska ranks second in the nation during this timeframe by translating 32 cents of each new dollar into higher instructional spending. By contrast Pennsylvania ranks 6th in per-student revenue growth and 10th in average teacher salary. But from 2002 to 2020, the state performed below the national average by only directing five cents of each new dollar to instructional salary spending. Pennsylvania has invested more money into public education than Nebraska over the last two decades and has higher overall teacher pay, but Nebraska has done a much better job at funneling new dollars into the classroom.

High spending growth states generally fare better at converting new funds into instructional salaries. That said, some states with modest spending growth (Ohio, Mississippi) have been more efficient than states with higher spending growth (California, New Jersey) at devoting new funds to instructional pay. And low spending growth states such as Georgia and Missouri shouldn’t be let off the hook, since their new dollars have resulted in less funding going to instructional salaries per student. No state does a very good job at getting new dollars to instructional salaries, suggesting that the problem is systemic.


Why giving teachers raises is easier said than done

Why do states do such a poor job of prioritizing teacher salaries? First, salaries are not the only component of total compensation. The compensation gains that teachers made from 2002 to 2020 were largely eaten up by higher costs in benefits. Figure 1 shows that instructional benefit costs per student, which primarily comprise retirement and healthcare expenses, have grown by nearly $1,000, or 79 percent. For every new dollar going to instructional salaries, four went to benefits. Importantly, benefit cost growth is due largely to rising unfunded liabilities in teacher pension systems, which helps explain why states like Pennsylvania and New Jersey with ailing pension systems do a poor job at increasing teacher salaries despite ample K–12 investments.

Second, Public Education at A Crossroads details how states have prioritized hiring new staff, especially non-teachers like school-level support personnel, over salary bumps for existing teachers. Nationally, total public school staff grew by 13.2 percent from 2002 to 2020 while enrollment grew by 6.6 percent. Non-teaching staff alone grew by 20 percent. The issue of benefit costs crops up here as well because new personnel are also affected by the rising costs of non-salary compensation.

But most fundamentally, compensation and staffing decisions are made primarily at the school district level, and district leaders have different incentives than governors or state legislators. While state officials might focus on how their teacher salaries compare with those of other states, district leaders are more concerned with day-to-day school operations and competing with neighboring districts for staff. Additionally, district budget officers are risk-averse and thus inclined to deploy new dollars toward marginal support personnel additions rather than increases to teacher salary schedules that lock them into longer-term commitments. Consequently, district leaders rarely plan to take advantage of staff attrition and re-purpose funds for raises.

To be fair, district leaders may have limited latitude to make these kinds of shrewd budget decisions, especially in localities with strong teacher unions. A chief budget officer in Los Angeles or Chicago would likely roll their eyes at any suggestion that they should make long-run budget tradeoffs; even if financial course corrections are desperately needed, their collective bargaining agreements would never allow it. And since nearly one-quarter of the country’s public school students live in one of the largest 120 school districts, a handful of contracts in union-friendly states can exert outsized influence on overall staffing and salary trends.

Notably, Figure 3 shows that union stronghold states generally fare better at driving new funds toward teacher salaries. This trend is corroborated by research showing that stronger collective bargaining agreements are associated with higher teacher salaries. But even if a strong union presence leads district leaders to do a comparatively better job at channeling new funds into teacher paychecks, that doesn’t mean they therefore do a good job at balancing raises with other budget priorities. Teacher unions also have competing incentives like adding more members and securing disproportionate benefits for senior members, both of which cut against broad-based salary increases. Union-friendly states also tend to have more staffing bloat and higher benefit costs.


No easy solutions

If their goal is to ensure new funds result in higher teacher salaries, state legislators may need to force the issue on local school districts. States would first need to shore up their pension systems and free up dollars for salaries. But beyond that, the tools states already use—minimum state salary schedules, maximum class sizes, staffing prescriptions—would have to be put on steroids.

Of course, that approach would curtail school district autonomy and scramble local salary schedules. It’s unclear how tightly funding restrictions could be enforced on wealthier school districts funded largely with local property tax revenue. Similarly, less wealthy districts with fewer dollars to go around would have their hands tied most tightly. For a glimpse of how difficult this option is, look at how Arkansas districts are handling the recent $14,000 increase in starting teacher salaries on the statewide salary schedule. Seventy-one of the state’s districts have resorted to compressing salary schedules so that teachers of all experience levels temporarily get paid the same until local leaders can fully determine how the new law will affect their budgets.

The only other available option within the current public education system is equally unattractive, even though it’s what states are currently doing: pump large shares of new dollars into the education system and hope some of it trickles into teacher paychecks.

Christian Barnard is assistant director of education reform at Reason Foundation.

Last Updated


Notify Me When Education Next

Posts a Big Story

Business + Editorial Office

Program on Education Policy and Governance
Harvard Kennedy School
79 JFK Street, Cambridge, MA 02138
Phone (617) 496-5488
Fax (617) 496-4428

For subscription service to the printed journal
Phone (617) 496-5488

Copyright © 2024 President & Fellows of Harvard College