Over at the Quick and The Ed, Kevin Carey writes
I'm not one of those people who believe that "money doesn't matter" in education. That's absurd; money matters a great deal, and there are plenty of schools that don't get their fair share.
While I tend to agree with Kevin Carey on education policy matters (at least the non-financial ones), I tend to disagree with all of his non-education policy posts over at The Quick and the Ed
. And since I try to focus this blog on education policy, let's leave it at that. (I also don't mean to single Kevin out since this is a common meme in the edusphere.)
Nonetheless, the "money doesn't matter" meme is not as absurd as Kevin is suggesting, provided that it's qualified with "at today's funding levels."
At today's funding levels, money doesn't matter. The correlation between school expenditures and student performance is very low to non-existent.
The fact that some schools "don't get their fair share" is irrelevant. The important question is whether schools are getting sufficient funding to educate their students. No one knows the answer to that question. What I do know is that many schools with low funding outperform many schools with much higher funding.
Let me cherry-pick some data points like Kevin did in his School Funding’s Tragic Flaw
From the 2005 Pennsylvania dataset from SchoolDataDirect, I found ten high-performing/low-funded school districts and ten low-performing/high-funded school districts based on 11th grade test results and total expenditures. The districts had similar poverty rates and minority attendance. Here's a graph of those 20 schools.
None of the high performance/low expenditure schools got their fair share. On average the low-performance/high-expenditure schools had $6880 less in funding (44% less) than the high-expenditure schools. That's not fair, now is it? Of course, it didn't stop these schools from performing over a standard deviation better than these high-expenditure schools. Perhaps if the high-expenditure schools received less funding they would have performed better?
Not content to merely cherry pick schools to "prove" my point, I decided to take a closer look at the dataset and focus on the performance of the economically-disadvantaged students.
First I looked at the performance of these economically-disadvantaged students in the best (top decile) schools. The average proficiency was 66.7%, well above the state-wide mean of all students (59.1%). These schools had expenditures averaging $11,640.
Then I looked rate at the performance of economically-disadvantaged students in the worst (bottom decile) schools. The average proficiency in the worst schools was 20.1%, well below the state-wide mean of all students. These schools had expenditures averaging $11,926.
Similar expenditures, vastly different outcomes.
Then I looked at it the other way.
I looked at the performance of economically-disadvantaged students in the best funded school districts (top-decile). The average proficiency in these rich schools was 41.1%, well below the state-wide mean of all students. The average expenditures for these schools was a whopping $16,260.
Then I looked at the performance of the worst-funded schools. The average proficiency in these poor schools was 45.1%, well below the state-wide mean of all students. The average expenditures for these schools was only $8,604.
You read that right. The economically-disadvantaged performed about a quarter of a standard-deviation better in the poorly funded schools. To put that in perspective, that's about the differential you'd expect from reducing class-sizes or hiring more effective teachers. (See here
.) But, in this case, it worked the opposite way.
You would think that these highly funded schools pissed most of that funding away on facilities and other window dressings, but surely some of it went to reducing class-sizes and hiring more experienced teachers -- two of the more favored educational panaceas. But, if it did, it still didn't work as the literature suggests it might. Which shouldn't be surprising if you are familiar with the way monopolies work, especially state-run monopolies.
You'd be hard-pressed to torture the data in a way that permits you to conclude that "money matters" at these funding levels.