Charter schools that post unusually high academic gains are often accused of having unfair advantages over traditional public schools, including more advantaged students and more private money at their disposal. A new and highly contentious study released today attempts to prove that the Knowledge is Power Program (KIPP), the largest charter-school network in the country, is inundated with both in comparison to its regular public-school counterparts and other charter schools.
The study is likely to give ammunition to charter-school critics as evidence that KIPP’s high test scores can be attributed to extra cash and a population of students that’s easier to educate. But the study’s findings are far from conclusive: The data used in the financial analysis are limited and, according to KIPP, often inaccurate, and the methodology used to examine KIPP students is problematic.
In the national battles over whether to increase the number of charter schools, research has been a weapon wielded aggressively by both sides. (Teachers’ unions and their supporters are typically on the anti-charter side, and ed-reformer-types like Michelle Rhee, former chancellor of the D.C. schools, and Joel Klein, former chancellor of the New York City schools, are on the other.)
But this study is different than many others because it accepts the fact that KIPP’s academic outcomes are indisputably extraordinary, and seeks instead to dig more deeply into “the reasons for its success.”
Most notably, the study, by Western Michigan University researchers at the Study Group on Educational Management Organizations, addresses the question of whether KIPP receives more money per student from government and private sources than other schools. Critics have wondered whether the chain’s reliance on philanthropic dollars, which have helped fund its rapid expansion, can be maintained as the network continues to grow.
The possibility that KIPP is getting more money per student than its traditional-school counterparts also raises the question of whether it’s reasonable to expect regular public schools to match KIPP’s achievements, and whether increasing the number of charter schools is an efficient use of money – an important question in tough economic times.
Here is what the study found:
In the 2007 school year, 12 KIPP school districts encompassing 25 schools received $12,731 per pupil from local, state and federal governments. Public-school districts where the KIPP schools were located received $11,960 (a few dollars more than the national public school average). Charter schools in general received much less on average: $9,579. Compared to regular public schools and other charters, KIPP received much more federal money, as well as more than double what other charters received in local funding.
Besides the extra government money that KIPP receives, the study found that the 12 KIPP school districts reported $37 million to the IRS in private donations in 2008, about $5,760 per pupil on top of the nearly $13,000 per pupil they received from the government.
“We were surprised they were getting so much,” said Gary Miron, a researcher at Western Michigan University and lead author of the study.
But KIPP vigorously rejected the study’s data after reviewing it yesterday. “This report has multiple factual misrepresentations,” Mancini said.
Mancini noted that the study focused on only 25 KIPP schools out of 58 open at the time when researchers calculated the financial data — missing schools in California, for example, which allocates much less money to charter schools than other states. According to KIPP’s own estimates, its schools receive about $9,000 to $10,000 per pupil, on average, from government sources, a figure that is closer to what other charters receive.
As for the private money, Mancini said the study does not take into account the fact that a significant part of the donations goes toward paying for buildings, often a large cost for charter schools in districts that don’t give them facilities. Miron, the study’s author, said that school districts must also pay for buildings, but Mancini countered that these costs are generally not included in per-pupil calculations.
KIPP estimates that it receives only about $2,500 per student from private sources, putting the total (including government money) at around $11,500 or $12,500 per pupil, right around what regular public schools receive. The study does not include data on the amount of private money other charter schools receive, but, keeping in mind that KIPP is the largest and best-known charter network in the country, it doesn’t seem unreasonable to assume KIPP does better at fundraising and that other charters receive less.
The takeaway is that KIPP’s model is not especially cheap, although KIPP does offer extras that traditional public schools don’t — like Saturday school and longer school days — for a similar amount of money.
“I think what this study does is at least give us pause about inferring that the KIPP model is a low-cost model,” said Jeffrey Henig, a political scientist at Teachers College who briefly reviewed the study before it was published, and who is affiliated with the National Center for the Study of Privatization in Education, housed at Teachers College. (The Hechinger Report is also located at Teachers College.)
The New York Times and Washington Post coverage of the study focused on the money issues, but articles in Education Week and Bloomberg News focused on the study’s examination of KIPP students.
KIPP uses a “no-excuses” model in which students and parents are required to sign performance contracts. Most of the students it educates are low-income. In fact, the WMU study found that KIPP enrolls higher percentages of low-income students than the public-school districts in which its schools are located.
But the idea that charter schools “cream” the best students from surrounding neighborhood schools and push out students who don’t perform well academically is a persistent critique of the schools, and the study claims to have found that the hardest-to-educate KIPP students tend to leave the schools at high rates.
In particular, the researchers argue that 40 percent of African-American male students, a group that generally posts lower test scores, “drop out” of KIPP schools between sixth and eighth grade. (Most KIPP schools are middle schools.)
“KIPP schools are cycling out those low-performing students, but they’re not replacing them,” said Miron. This is thought to be advantageous to KIPP for two reasons: first, the schools get to keep the funding tied to the student for that academic year even after he or she leaves the school; and, second, a school’s test score average goes up when low-performing students quit.
KIPP aggressively contests this finding, however. Mancini pointed to a study KIPP commissioned from the nonpartisan research group, Mathematica, which followed individual students over time. The WMU study used aggregated data taken as a snapshot and compared KIPP attrition rates to the rate of students who moved out of the school districts in which KIPP schools were located. Mathematica researchers said that a student leaving an individual school is not the same phenomenon as a student leaving a district.
“You have to do a school-by-school comparison,” said Brian Gill, one of the co-authors of the Mathematica report, which found that, on average, attrition at KIPP schools is about on par with schools in surrounding neighborhoods. “There’s a real danger from people drawing inferences from this that aren’t supported.”
That some KIPP schools don’t replace students if they leave is true, however, and both Mancini and the Mathematica research team said they have been looking into this phenomenon.
Next week, Mathematica will release a new study on the matter, but as with most charter school studies, it’s unlikely to be the last word.
First, the criticism of KIPP was that it didn’t produce better results. After it became clear that, overall, KIPP schools usually do a better job, the criticism moved on to the “mix” of students. Since charter schools cannot discriminate, the presumption was that students were somehow “tossed” out. Now we see that many students leave because the work is too demanding. (Unfortunately, for most of us, learning can involve hard work.) Finally, after we’ve seen that KIPP schools often have a higher percentage of students in poverty than the surrounding public schools, the new criticism is that KIPP spends too much money. The amounts don’t seem all that much higher, if this new data is even true, but the KIPP schools use that money to pay for their own facilities, longer school days and longer school years — all things that are not included in traditional public school per-pupil expenditures.
One could be forgiven for thinking that all of this cricism of an instructional model that is simply a public school option — no one is required to choose KIPP — is really opposition to school choice, in any form. If choice is the real enemy for these KIPP critics, then they are unlikely to convince most of the public of their position.
If the 8th grade is significantly smaller than the same cohort was in the 5th grade, and this happens year after year, it doesn’t matter in the big picture if some students were held back rather than being counseled out. Overall, the attrition is still attrition and it’s still a red flag.
Mathematica is being dishonest (presumably under duress, dancing with who brung it — KIPP paid for the study) in not discussing the fact that the numbers show that KIPP doesn’t replace an enormous number of the students who leave and public schools do. Whatever the official policy is, that’s what the number show, resoundingly and irrefutably.