In 2009, a national study by Stanford’s Center for Research on Educational Outcomes, or CREDO, found that charter schools do not produce better outcomes than traditional public schools. If you have ever come across a column denouncing charter schools, you have probably seen a reference, either direct or indirect, to this finding. “Evidence is mounting that the reforms are not working. Stanford University’s 2009 study of charter schools — the most comprehensive ever done — concluded that 83 percent of them perform either worse or no better than traditional public schools,” wrote Joanna Barkan in Dissent. “Known as the CREDO study, it evaluated student progress on math tests in half the nation’s 5,000 charter schools and concluded that 17 percent were superior to a matched traditional public school; 37 percent were worse than the public school; and the remaining 46 percent had academic gains no different from that of a similar public school,” proclaimed Diane Ravitch. “Social science research,” writes Matthew Yglesias, “shows very clearly that on average, charter schools freed from the shackles of collective bargaining agreements are no better than traditional public schools.”
When moderates cite the study, it is in a spirit of resignation: “We tried a new experiment, and it failed.” When leftists cite it, they generally leap directly to accusation: “We know charters do not help children, so only a nefarious motive can explain their existence.” The ubiquitous reliance on this research led many people to permanently dismiss any chance charter schools could produce significant improvements.
But this finding is badly out of date. A few years ago, Margaret Raymond, CREDO’s director, told me that its studies of states and districts were noting a distinct upward trend line in charter schools, especially those located in cities. At the time, CREDO had not conducted a national study since 2013. But now the center has released its latest report, and it confirms the trend Raymond had already seen locally. The new study finds, unambiguously, that students on average gain more learning in charter schools than in traditional public schools.
What happened since 2009? Simply, charter schools got better. This chart shows the overall trend in charter schools from CREDO’s first study, in 2009, to its 2013 follow-up, to its latest:
The entire purpose of the charter-school model was to encourage experimentation with different models so that the best models could be tested and spread. The first charter school opened barely more than three decades ago, so it is hardly a surprise that it has taken time for educators to determine best practices.
I’ve found even highly educated people who are not specialists in this field don’t understand how studies of charter schools work. When they hear charter schools produce learning gains, they assume those studies simply make unweighted comparisons of the test results of charter-school students against students in traditional public schools. That would be a terribly inaccurate way to measure the effect of a charter school, since it would leave open the possibility that charter schools simply enroll smarter or more motivated students.
Instead, researchers employ a couple methods to separate the effect of the school from any differences in students. One method is lottery studies. Charter schools, unlike private schools, can’t select their students, and if they have more applicants than spaces, they must dole out spaces by random selection. Scholars have used these lotteries to measure the differences in outcomes between winners and losers.
Lottery studies are a highly valuable method, though they suffer from the drawback that they only examine students whose families were motivated enough to enter a charter lottery to begin with. (Students from under-privileged families may be more likely to simply default to the local school without exploring alternatives.) CREDO’s method is different. It uses sophisticated statistical controls of matched students with similar demographic and economic characteristics, then compares the outcomes for matched pairs in traditional public schools and charters. This method also allows CREDO to study outcomes on a larger scale than a lottery, which can only be conducted within a city or school district.
CREDO finds that a given student is now likely to learn more in a charter school than a demographically identical student in a traditional public school. Beneath that top-line finding, there is tremendous variation. The details matter tremendously. But the specifics do not undermine their conclusion that charter schools offer a promising model for closing the achievement gap. On the contrary, the details make the case far stronger than the top-line figures would indicate.
Some forms of charter schools are predictably unimpressive, or even worse. Charters in rural or suburban areas have never outperformed traditional schools. (This is probably both because they lack the geographical concentration to allow a real school choice and because charters have been most effective at providing alternatives for low-performing schools, which are concentrated most heavily in cities.) Online “virtual” charter schools are a catastrophic failure. Their students do so badly that even a tiny number of them drags down the overall average of charter students. As CREDO notes, “despite having only 6 percent of charter school students enrolled, their impact on student progress of 58 fewer days of learning in reading and 124 fewer days in math has damaging consequences for students and exerts an outsized drag on overall national results.”
Another factor detected by CREDO is the success of charter management organizations, or CMOs, which is a network of charter schools, as opposed to a single, individually operated charter school. CMOs produced greater learning gains than stand-alone charters.
What this variation means is that there are networks of charter schools serving low-income urban students that produce extremely large learning gains. It has been the case for many years that charter schools produce large learning gains for urban minorities, but that important finding was obscured by the endlessly cited, obsolete finding that charter schools on the whole perform no better than traditional schools.
The nationwide average was never a good reason to oppose expanding charters in the places where they perform very well, but now the 2009 nationwide result is not merely irrelevant but false. It is perhaps an opportunity to focus on the transformative potential of urban charter schools, which the latest report details.
Many of these charter networks can consistently allow low-income urban students to wipe out the achievement gap. This passage from CREDO’s study shows how powerful and widespread these effects can be:
More than 1,000 schools have eliminated learning disparities for their students and moved their achievement ahead of their respective state’s average performance. We refer to these schools as “gap-busting” charter schools. They provide strong empirical proof that high-quality, high-equality education is possible anywhere. More critically, we found that dozens of CMOs have created these results across their portfolios, demonstrating the ability to scale equitable education that can change lives.
This is not just a handful of exceptional schools here and there. A large number of charter schools have developed scalable models that can allow Black and Latino students in cities with awful neighborhood schools to get the same education as white kids in suburbs enjoy.
Skeptics point out, accurately, that charter schools do not solve every problem in public education. What they do solve is one particularly important problem: the fact that generation after generation of Black and brown kids come through the school system years behind their affluent suburban peers. This achievement gap is a social crisis of the highest order. Indeed, many of the most bitter fights in American society concern its downstream effects. Affirmative action is a worthwhile but far less effective way to compensate for the deeper problem that American schools produce scandalously few high-achieving Black and Latino graduates.
It is difficult for me to understand why, in the face of evidence that charter schools can make such a large difference for this neglected cohort, they are so often met with fatalism or antipathy. Opposition by teachers unions, which is rooted in charter schools’ need to circumvent union contracts that make it difficult to fire ineffective educators, has entrenched itself as the official progressive line on the subject; just as most progressive organs will trust pro-choice organizations on abortion and climate activists on the environment, so too will they generally defer to teachers unions on education reform.
I suspect conservatives consider the achievement gap an unalterable genetic or social divide and have difficulty believing Black and Latino children genuinely have the potential to achieve at the same level as white kids. (I suspect quite a few liberals have trouble believing it, too.) Self-interest surely plays a role here: It is probably not a coincidence that Black and Latino Democrats have a much more favorable view of charter schools than do white Democrats.
Even those liberals who implicitly grant that charter schools can make a positive difference for minority students tend to linger on the limits of what a better education system can accomplish. There is almost a mania for refuting the proposition that education reform will eradicate poverty and inequality. Here are just a few headlines turned up by a simple Google search: “Education Alone Won’t Solve Poverty, Social Isolation, and Resentment”; “Education Alone Won’t End Income Inequality”; “Why Education Won’t Solve America’s Inequality Crisis”; “Education Can’t Solve Poverty — So Why Do We Keep Insisting That It Can”; “No, Education Still Won’t Solve Poverty”; “Better Schools Won’t Fix America.”
I can’t think of any social intervention that is so frequently measured against the baseline, “Does this single-handedly eliminate poverty/inequality?”
So, fine, I agree. Education alone won’t end inequality or poverty or save America. Neither, for that matter, will universal health care or full unemployment or ending racism. So what?
The standard we usually apply to a social intervention is not whether it solves all our problems but whether it makes the world a fairer or better place. In this case, we have overwhelming evidence of something that works. Rather than throwing up our hands at how politically difficult it is, or despairing over the other social problems that would remain in the world, why don’t we just try to make it better?