How much does it cost to educate a child? It often feels like policymakers pick numbers out of a hat. Utah spends less than $7,000 a year on a student from kindergarten through high school. New York spends more than $20,000, federal data show. Within the state of Illinois, a wealthy district typically spends $3,400 more than a poor district, according to a February 2018 study by The Education Trust, a nonprofit group that conducts research and advocates for low-income students. Cost of living differences account for some of these gaps but not all or even most of them, says Ary Amerikaner of Education Trust.
Now a team of five researchers from Rutgers University in collaboration with the Education Law Center, a nonprofit advocacy organization, has created a complicated model that predicts how much money it would cost each school district in America to get its students to reach average test scores in math and reading, as recorded from 2013 to 2015. This is not a particularly ambitious goal; the average test score in the U.S. is well below what is considered “proficient” for each grade level.
Yet most low-income school districts across the country aren’t spending enough to reach even this timid target, never mind closing the achievement gap between rich and poor, according to these researchers’ calculations. Spending shortfalls are particularly large and common across the South and Southwest (see all the red districts in the accompanying map illustration). High-poverty districts in California, Arizona and Mississippi would need to spend an additional $15,000 above what they are already spending just to become average, the authors estimate.
“We decry our PISA results as some kind of national shame,” said Bruce Baker, referring to the Program for International Student Assessment, an oft-cited test where U.S. performance, especially in math, has slid. “We’re being dragged down by states that have thrown their (school) systems under the bus.”
Baker, the lead researcher in creating the cost model published earlier this month, is a professor in Rutgers’s Graduate School of Education. He’s made a career out of analyzing school funding inequities around the nation. His work has been used in court cases in which advocates have pressed for better funding for low-income districts. Now he has gone one step further, not only documenting the gaps between rich and poor but also connecting how much a school spends to performance. “We’ve never been able to do this before,” he said. “This is our first shot at asking the question: how much more do higher poverty districts need than lower poverty districts to achieve a particular goal?”
Many might argue with the premise that you can spend your way to academic success. It’s easy to point to poor districts with seemingly generous budgets, such as the District of Columbia, which spends more than $20,000 per student but still produces below average test scores. Meanwhile, other districts beat the odds, managing to achieve with fewer funds. However, scholars have found that higher school spending tends to be associated with better results. A 2015 paper documented how court orders to increase school funding in low-income communities eventually led to better-educated adults, higher wages and a reduction in poverty. Another 2016 paper found that states that sent additional money to their lowest income districts (usually through court orders) eventually saw higher test scores in those low-income districts compared with states that didn’t reallocate school funds.
In the past, it was hard to compare academic achievement in different parts of the country because every state has its own standards and related tests. Thanks to painstaking calculations at the Stanford Education Data Archive, there is now a dataset that maps every state test onto a single national yardstick. And that allowed Baker to compare school funding for every district in America with academic achievement.
According to Baker and his colleagues, it can cost anywhere between roughly $5,000 and $30,000 a year per student in order to hit average test scores. Two factors mainly determine where a district lies along that range: location and mix of students. Some school districts bear higher costs because they’re located in expensive regions where salaries, including those of teachers, are high. Population density matters too. The costs of educating poor children escalate faster in urban areas, the researchers found.
The more low-income students, the more expensive it is to pay for the resources to make up for the disadvantages of poverty. To achieve the same academic results, a poor district would need to spend three times the amount that a rich one does, the model estimates. A district where more than 40 percent of the students are poor would have to spend between $20,000 and $30,000 per student to achieve average test scores. In a district where fewer than 10 percent of the students are in poverty, the bill would be only $5,000 to $10,000 per pupil.
It would take even more money to get students to perform at grade-level standards or above proficiency. And this paper doesn’t attempt to make these calculations. (To see Baker’s estimated per-student costs for each state in the nation, adjusted by district poverty levels, click here.)
Baker cautioned that this early version of a cost model might be “overstating” how much extra money low-income districts need. He pointed out that similar studies for individual states have found that doubling the costs per pupil was sufficient to equalize outcomes between rich and poor. In a report this month to Kansas lawmakers, a professor at Texas A&M University calculated that Kansas would need to spend an additional $2 billion a year to help under-performing students meet academic targets.
It’s exactly the numerical details, with hosts of assumptions, that make models like this controversial. And the results can be potentially misleading to the general public. “The overall direction is right and intuitive,” said Matt Chingos, a school finance expert at the Urban Institute, a nonprofit research organization in Washington, D.C. “Yeah, it’s going to cost more to make up for things at school that low-income kids aren’t getting at home. But it’s hard to put exact numbers on these things.”
Chingos is skeptical that if individual districts raised their spending, according to Baker’s calculations, their test scores would rise to the level that Baker’s model predicts. Chingos points out that teacher quality is one of the most important factors in influencing student achievement. Because of low teacher salaries in Arizona, the state might not attract the best teachers it could. If the state were to suddenly start spending more, and use that money to give teachers raises, then, initially, the state would just be paying the same set of teachers more. Student achievement probably wouldn’t budge much. It could take 30 years for the current teachers to retire and be replaced with a higher quality teaching force.
In education, nothing moves quickly.