| January 13, 2014
Does Per-Pupil Spending Matter?
Advocates of increased funding for public schools love to trot out the Holy Grail of their argument, the claim that per-pupil spending in Oklahoma lags behind levels in other states, and that if we’d just boost how much we spend per student we would cross an invisible line into some sort of educational Utopia. Add $500, or $1,000, or whatever the figure of the moment may be, and Johnny would become a scholar of renown.
That was the logic behind the 2010 push for State Question 744, which would have mandated specific levels of education spending keyed to the regional per-pupil spending average. Thankfully, voters saw through that charade (or at least they were frightened off by the massive tax increases SQ744 would have required) and rejected it overwhelmingly.
But the per-pupil spending mythmakers are ever busy, dusting off their argument for each legislative session to call for x new school dollars because, well, because if we can just hit that magic number things will be hunky-dory in the classroom. A low per-pupil spending figure means less learning, they imply; a higher one will lead to more.
Unfortunately, data collected here in Oklahoma and nationwide fail to support that claim. In fact, there is ample evidence that the more you spend on schools, the dumber some kids become.
* * *
Utah is supposed to be ashamed that it has the lowest per-pupil spending average in the nation, just $6,612, according to the National Center for Education Statistics (NCES).
The highest per-pupil spending can be found in the District of Columbia public schools, which is not surprising, since those schools are effectively federalized. NCES says per-pupil spending for D.C. schools was more than triple Utah’s, at $19,698.
So given the thesis that more spending equals more learning, D.C. kids ought to be miniature geniuses, while those poor neglected kids in Utah should be mired in ignorance.
Let’s look at the ACT college entrance exam, a good end-of-instruction benchmark since it is usually taken early in a student’s senior year of high school. Just 32 percent of D.C. seniors took the ACT in 2012, which means that those who did were the cream of the crop, the most likely to be college bound.
Fifty-one percent of the D.C. students who took the test were rated as college-ready in English. Forty-two percent were rated as college-ready in reading. Just 37 percent met college-entrance levels in math, and 26 percent were seen as college-ready in science.
Remember, these scores came from the top third of the D.C. class. And those kids had allegedly benefited from 12 years of the highest per-pupil spending in the land.
Out in supposedly underfunded Utah, an astonishing 97 percent of high school seniors took the 2012 ACT test. Statisticians will tell you that when almost everyone in school takes a test, that will inevitably water down the aggregate results, since a lot of kids who have no intention of going to college are included in the final scores.
Yet 64 percent of Utah seniors were rated college-ready in English, 54 percent in reading, 40 percent in math, and 29 percent in science. The Utah test-taking sample was triple that of the District of Columbia, Utah spends comparative peanuts to educate them, and those kids out west still out-performed their cousins in D.C.
The National Assessment of Educational Progress (NAEP) annually tests millions of kids at several grade levels in core subjects and reports its findings by state. In 2011, fourth graders in Utah and D.C. were tested in reading. NAEP ranks students in four groups—those who are advanced at the tested skill, those who are proficient, those who perform at basic level, and those deemed below basic (essentially, those who fail).
Six percent of Utah’s 2011 fourth graders were advanced. Twenty-seven percent were proficient. Thirty-four percent ranked at basic level, and another 34 percent were below basic. Overall the Utah fourth graders achieved a numerical score of 220, which happened to be precisely the national NAEP average for that year.
In D.C., the same six percent ranked as advanced readers, while 13 percent were proficient. Another 25 percent rated scores at basic level, but a whopping 56 percent—a majority of DC fourth graders—read at below basic level. The aggregate score was just 201.
So Utah is spending $6,612 per student—$132,240 for a class of 20—and outperforming the District of Columbia, where the taxpayers are forking over $19,698 per kid, or $393,960 per classroom. Put simply, D.C. spends more than a quarter-million dollars more per classroom yet gets worse results.
These inconvenient truths are not confined to the extremes. A 2012 paper by Drs. William E. Bibb and Larry McNeal assessed the impact of per-pupil spending on schools in their native Tennessee. “This research,” they concluded, “revealed that per pupil expenditure did not have a significant relationship to ACT scores or to the TCAP (statewide) Writing Assessment scores. An implication is that giving schools more money does not necessarily raise student achievement, but rather how the money is spent can raise student achievement.”
In short, mere dollars are not the answer. How and where those dollars are spent matter more.
Next came State Budget Solutions, which issued a study based on the 2009 and 2011 school years that concluded “states that spend the most on education as a portion of their total budget didn’t graduate students at a higher rate, nor did their students score better on the ACT than their peers.”
* * *
Even a cursory glance at the Oklahoma Educational Indicators Program reports will show that there is no detectable correlation between per-pupil spending and what students are learning, as measured by a wide range of tests, in different Oklahoma schools.
To accurately assess data from Oklahoma schools, it is important to recognize some variables.
First, per-pupil spending can vary widely between schools, since local property tax bases and dollars funneled to them in state aid may be quite different. Small rural schools, schools with high levels of minority or even special-education enrollment, and those with other demographic factors may receive considerably more in state aid per student.
So it is not uncommon to find two schools in the same county with per-pupil spending levels that can differ by hundreds, or even thousands, of dollars.
It is also important to keep in mind the large number of small rural school districts in Oklahoma, the product of historic resistance to any form of consolidation. We are propping up hundreds of small schools—in some cases a dozen or more in a single county—at excessive additional cost. That results in a less than efficient expenditure of school dollars, if only to pay all those small school superintendents.
So how are some of our schools doing on the per-pupil spending front as it (allegedly) translates into more learning?
Bethany Schools is an average school district in many ways. It has about 1,500 students from a middle-income suburban population. Per-pupil spending there is well below the state average, at $7,017 for 2010. But Bethany kids are doing well.
They score at or above the state average in all 26 state-mandated assessments of student learning, from tests given in grades 3, 4, 5, 6, 7, and 8, and at the end of instruction in seven high school subjects. In addition, Bethany’s class of 2010 averaged 21.6 on the ACT, nearly a full point above the state average of 20.8.
Bethany’s four-year dropout rate was just three percent, compared to a state average of 11.1 percent. Almost 73 percent of Bethany grads went on to college (the state average was about 57 percent) and five percent fewer Bethany graduates had to take remedial courses as college freshmen than the state average.
All in all, Bethany is doing well—at a per-pupil spending level of $7,017, some $800 below the state average, which the advocates of more school spending would be quick to tell us means that education in Bethany is “underfunded.”
Now drive east to Heavener Schools in LeFlore County, which has a similar enrollment (1,027 students), more Native Americans, but otherwise is not that different from Bethany. Heavener spends $9,383 per student, more than $2,300 more than Bethany, and a hefty $1,500 more than the state average.
Heavener students score below the state average on 14 of the 26 state mandated assessments, which is not terrible, but less than you’d expect with that increased financial investment if those dollars actually led to more learning.
The ACT average for Heavener was 19.0, well below the state average.
The four-year dropout rate was 4.7 percent, not bad again, but still worse than Bethany’s.
About 44 percent of Heavener grads went on to college, but 54.6 percent of them—more than half—had to take at least one remedial freshman class.
The bottom line? Kids in Bethany are coming out of that school district significantly better educated than their peers in Heavener, at less cost. If the per-pupil spending gurus are correct, you’d expect results that are exactly the reverse.
You can do these comparisons all day long. Some are startling. Out in the Panhandle at tiny Plainview School (nine total elementary students) we are spending an astounding $48,647 per student, but the kids there are still scoring below state average on every grade tested.
Jenks Schools have one of the lowest per-pupil spending levels in the state ($7,352) and one of the highest academic records, with an ACT average of 23.8. Caney Schools in Atoka County spends a hefty $10,956 per kid and recorded an ACT average of just 18.7 and a college attendance rate of barely one in three.
Down in Norman they are virtually poverty-stricken, at a mere $7,093 per student—almost Utah territory. Somehow they are acing all 26 state tests, scoring an astronomical 23.3 on the ACT, and sending three-fourths of their graduates to college.
So clearly, how much you spend per student has little, if any, impact on educational outcomes. What does?
Back in 1966 sociologist James Coleman issued a landmark report on a massive study of public education. He isolated two primary factors that have more to do with student success than any other—demographics and family background.
Not surprisingly, kids from stable two-parent homes where there are books on the shelves, limits on television time, and parental educational levels that foster an expectation of academic success do better than those from poor single-parent households where drugs, violence, sloth, and other factors send a signal that it doesn’t really matter how you do in school, or whether you go at all.
And there is nothing school dollars can do to alter those conditions. We could triple Oklahoma’s per-pupil spending average to surpass D.C.’s and those numbers in Norman and Bethany and Heavener and Jenks would barely budge.
Of course demographics is not destiny. We all know people from stable backgrounds and supportive communities who dropped out of school, and we all know children of poverty and chaos who overcame those obstacles to do well and succeed. But on average, the family backgrounds and socioeconomic environments most kids come from are much better predictors of school success or failure than how many dollars, out to the last decimal point, we are spending on their education.
It’s time to bury the per-pupil spending myth that has driven too many of our educational and appropriations policies for decades. How much we spend on schools, above a sensible basic amount, has almost nothing to do with the results those schools produce. How we spend those dollars—creating a more efficient network of schools where more courses are offered, for example—can have a minimal positive impact, but dollars in are never going to magically turn into geniuses out.
Mike Brake is a journalist and writer who has recently authored a centennial history of Putnam City Schools. He served as chief writer for Gov. Frank Keating and for then-Lt. Gov. and Congresswoman Mary Fallin, and has also served as an adjunct instructor at OSU-OKC.