During a , Rep. Kevin Kiley, R-Calif., used a graph on an easel behind his chair to make a dramatic point about the last decade of American education.
鈥淲e鈥檝e just received the latest test scores for students across this country, and they are absolutely alarming,鈥� said Kiley, the newly minted House subcommittee chair for K-12 and early-childhood education. He pointed to a chart with upward-facing curves above the x-axis showing increased education funding, and downward-facing curves showing declining test scores.
鈥淭his is a steady growth in spending in real dollars which is proceeding in tandem with a steady decline in student achievement,鈥� Kiley said, just before previewing broad plans to 鈥渞eimagine federal funding鈥� for schools and expand public subsidies for private education nationwide.
With his comments, Kiley joined a chorus of politicians, pundits, and researchers who have questioned why U.S. students haven鈥檛 gained more academic ground after Congress doled out close to $200 billion in pandemic relief funds for K-12 schools in 2020 and 2021.
But the story Kiley told about the , published Jan. 29 by the Georgetown University-based research group the Edunomics Lab, doesn鈥檛 precisely line up with what the graph actually shows鈥攐r what its authors say they want to get across.
In the two weeks following the release of the latest NAEP scores, the Edunomics Lab charts have helped spark the latest round in a perennial debate over how much money really matters in education. The discussion is nothing new for the field, but it raises a deeper philosophical question: Does plotting spending figures against test scores wind up increasing confusion rather than clarity?
Funding matters, but what about how it鈥檚 used?
Researchers reached consensus roughly a decade ago that funding increases make a tangible difference in student outcomes. But debate has continued since then over which students benefits most from increases, and what else states, schools, and educators need in order to use money efficiently.
These discussions are especially consequential now, as many states gear up to revise decades-old school funding formulas, and the Trump administration plows forward with an unprecedented and sweeping crackdown on federal spending that has already hit education research hard and tested centuries-old constitutional principles.
Graphics like the Edunomics analysis come at an inopportune time, said Morgan Polikoff, a professor of education at the University of Southern California Rossier School of Education.
鈥淭he message it perpetuates is, school spending is unrelated to student achievement. And there is very ample research that shows that to be false,鈥� Polikoff said.
This message is spreading as the Trump administration and Republican majorities in both houses of Congress have threatened to slash federal education funding and eliminate the U.S. Department of Education altogether.
鈥淭hey could point to this figure and say, well, school spending doesn鈥檛 matter,鈥� said Polikoff. 鈥淏ut that isn鈥檛 true.鈥�
Marguerite Roza, director of the Edunomics Lab and a research professor at Georgetown University, said she didn鈥檛 intend to suggest that money doesn鈥檛 matter for education. Instead, she hoped her group鈥檚 analysis would start a conversation about why funding increases have contributed to academic improvements in some states more than in others.
鈥淥f course money matters,鈥� Roza told Education Week. 鈥淲e鈥檙e looking at where it matters more, when it matters more, what role do people play in making sure it matters more. It鈥檚 a how much, not a yes or no.鈥�
Despite the confusion, Roza sees the vigorous dialogue about the 鈥淩OI鈥� analysis as a sign that her analysis is prompting essential, if uncomfortable, debate and introspection. Lawmakers, state education departments, school district leaders, and labor unions that represent educators all have a role to play in setting America鈥檚 students on a more positive academic trajectory, she said.
鈥淭houghtful, smart people can disagree on stuff and still have a conversation where we all end up smarter, regardless of whether we changed our view,鈥� she said.
Comparison of scores and funding demands context, experts say
The Edunomics chart and its were published Jan. 29. Earlier that day the federal government released the latest round of scores for the NAEP exam, administered to a sample of the nation鈥檚 4th and 8th graders in reading and math every few years. The results generally showed declining or stagnant performance and widening gaps between struggling students and their higher-performing peers.
The Edunomics team mapped NAEP scores from the last decade onto a chart for each state that also includes federal data on K-12 education spending, and a curve that illustrates the inflation rate as published by the federal Bureau of Labor Statistics. The charts have since prompted a wide range of reactions.
Some conservative pundits and right-leaning policy groups have included the charts in posts arguing for decreasing or rethinking school funding.
鈥淭eachers鈥� unions and their apologists continue to say the problem with public schools is funding 鈥� More money correlates with more problems rather than solving them,鈥� Andrew Clark, president of the yes. every kid. foundation, an arm of the Koch Network which advocates for school privatization.
Some education researchers, in interviews and , said the charts don鈥檛 fully contextualize how school spending and academic outcomes are linked.
Of course money matters.
For instance: It鈥檚 impossible to know whether NAEP scores might have been even lower without spending increases over the last decade. State-level figures paper over the substantially different spending strategies school districts took with the resources at their disposal. There鈥檚 no evidence of a directly proportional connection between increased spending and higher test scores because so many other factors are at play.
鈥淵ou can鈥檛 compare [test scores] that go from 0 to 500 on a plane with spending that goes from $0 to $30,000 or $0 to $100,000,鈥� said Jess Gartner, the founder of Allovue, a tech company now owned by PowerSchool that helps school districts manage spending. 鈥淚f you want to share that information side by side, it needs to come with a lot of context.鈥� (Gartner serves on the board of trustees for Editorial Projects in Education, the nonprofit publisher of Education Week.)
Other research suggests that COVID relief funding has made a difference. Analyses published last fall and earlier this month by professors at Harvard and Stanford universities show that, in low-income districts, the additional federal dollars had about as much influence on student achievement as a general revenue increase鈥攁nd did, in fact, 鈥減revent larger losses,鈥� the researchers wrote. Districts that used more of their money to fund academic recovery saw bigger gains in student test scores.
But that nuance is lost in the Edunomics charts, some say.
鈥淚f you went to any state legislature in the country and presented this, the idea that they would get is not, we should figure out how to make dollars more effective,鈥� said Josh McGee, associate professor of education policy at the University of Arkansas. 鈥淭he idea they would get is, we don鈥檛 have to invest anymore because the dollars right now are being wasted.鈥�
The pandemic brought unpredictable expenses
That exact scenario has already begun to play out on Capitol Hill. And state legislatures are taking notice.
Late January in the Oregon House of Representatives, lawmakers as they tried to grapple with their state鈥檚 academic performance. Some lawmakers were skeptical that test scores alone fully illustrate the state of Oregon鈥檚 schools, .
The Oregon Parent Teacher Association advanced the debate a few days later, when the group sent a to state house lawmakers detailing numerous complaints about the Edunomics analysis and warnings about how to interpret it.
If you went to any state legislature in the country and presented this, the idea that they would get is ... we don鈥檛 have to invest anymore because the dollars right now are being wasted.
Among several concerns, the group echoed experts who say the graphs highlight overly simplistic spending numbers.
On the , for instance, a curve shows that school spending increased by 71 percent between 2013 and 2024. Another curve below it shows the inflation rate increase of 35 percent over the same period.
The accompanying text spotlights only the percentage increase in unadjusted dollars, leaving out the lower percentage increase when accounting for inflation.
News reports in , and have cited their states鈥� respective spending increase figures as characterized by Edunomics, without mentioning that the numbers haven鈥檛 been adjusted for inflation. The Oregon parents group told lawmakers the unadjusted numbers leave out that 2013 was a low point for post-recession education funding in the state.
Roza said her group鈥檚 choices were deliberate, noting that in focus groups that 鈥減eople don鈥檛 understand what adjusted dollars are. Laying it out separately like that allows people to see what鈥檚 going on.鈥�
As for Oregon, she said, 鈥渋t is hard to ignore that other states have been more successful in leveraging incremental dollars to drive progress for students.鈥�
Numbers don鈥檛 tell the full story, researchers say. Some districts had far less funding than others even before unevenly distributed pandemic aid; the price of goods and the makeup of the labor pool vary widely from place to place; and growing fixed costs for staff salaries and benefits, utility bills, and pension debts can鈥檛 be avoided.
鈥淲hat we really care about is the extent to which additional dollars allow us to purchase more educational goods and services.鈥� McGee said. 鈥淭his overstates how much more we could purchase.鈥�
All of this context complicates arguments about how school funding is or isn鈥檛 working. But the nuance is essential said Katie Roy, partner and general counsel at Education Resource Strategies, a firm that helps large school districts with budget planning.
Roy helped found the nonprofit , and used to spend much of her time explaining the particulars of school finance to state senators and representatives who didn鈥檛 know the first thing about funding formulas or district expenses. Her team used a 100-page PowerPoint presentation and applied one main rule to avoid confusion: Each chart should only communicate one piece of data.
鈥淧eople who do not have a lot of experience looking at data, charts, and graphs can very easily, just by mistake, misunderstand,鈥� Roy said.
Student achievement may be slow to rebound, experts say
Roza and many of her critics share the view that reversing downward trends in student achievement will require a wide range of thoughtful new approaches.
The effects of the pandemic, a once-in-a-lifetime disruption to students鈥� lives, weren鈥檛 likely to be fully offset on NAEP in a few years, said Martin West, the vice chair of the National Assessment Governing Board, which sets policy for NAEP. Lost learning tends to persist.
鈥淭hat is, there鈥檚 nothing that would make us expect that students would become faster learners after a disruption. In fact, if anything, you might think the opposite: They would have lost some knowledge or skills or habits of mind that would slow their progress going forward,鈥� he said.
The tutoring and summer school programs that many schools stood up to make up for lost time varied widely in scale and intensity. The call to 鈥渁ccelerate鈥� students鈥� learning by simultaneously teaching new content and refreshing previous years鈥� skills felt practically impossible for many educators.
鈥淚f we could improve schools hugely, rapidly, we probably would have done it beforehand,鈥� Susanna Loeb, the executive director of the National Student Support Accelerator, told Education Week in January.
Despite these steep challenges, research from the Education Recovery Scorecard, a project from researchers at Harvard and Stanford that tracks pandemic-related learning loss, found that some districts have bucked the trend of general decline and raised student scores above pre-pandemic levels. But the data can鈥檛 pinpoint what combination of policy decisions and spending choices caused these results in every case, the researchers said.
NAEP is notoriously hard to use to draw cause-and-effect conclusions, and generally speaking, improvements on it are incremental.
鈥淣AEP trends don鈥檛 move very much in terms of points and percentages. They move slow, and even in good times they move up by a small sliver of a standard deviation, whereas staffing and spending go up in percentage terms quite a bit more,鈥� said Chad Aldeman, a policy analyst and columnist for The 74 who has published reports on school funding and educator pensions. 鈥淵ou can鈥檛 just assume that because one trend goes up at one rate, the other trend goes up at the same rate.鈥�
And, at least in reading, the declines go much further back than the pandemic. This slide comes after an upward trend in NAEP scores during the 1990s and 2000s.
This fuller timeline of scores on the nation鈥檚 report card complicates the connection between dollars spent and student outcomes, said Polikoff. School spending has risen for much of the past three decades, though it dipped in the years following the Great Recession. Achievement has gone up and down over the same period.
Roza said she hoped the Edunomics analysis would spur policymakers and experts across the ideological spectrum to embrace the iterative work of improving students鈥� test scores, rather than ignoring the nationwide trend and the states defying it.
鈥淚 actually think the work of getting more value for the dollars is kind of a slog,鈥� Roza said. 鈥淚 think it requires prioritizing what we鈥檙e trying to do in schools, and I think it means using data constantly to see what鈥檚 working and what鈥檚 not.鈥�
Some districts cited labor shortages and hiring challenges to explain why their ambitious plans to scale up tutoring and summer school for the last few years fell short of expectations. Roza isn鈥檛 convinced those hurdles were insurmountable.
鈥淲e have to be willing to be eyes wide open and pivot and check if the investments are delivering the value we hoped,鈥� Roza said. 鈥淚f not, try something else, or try to fix the thing that鈥檚 not working.鈥�
