Charter scores are lower

Charter schools lag traditional public schools in test scores, according to a New York Times story.

The data shows fourth graders attending charter schools performing about half a year behind students in other public schools in both reading and math. Put another way, only 25 percent of the fourth graders attending charters were proficient in reading and math, against 30 percent who were proficient in reading, and 32 percent in math, at traditional public schools.

Because charter schools are concentrated in cities, often in poor neighborhoods, the researchers also compared urban charters to traditional schools in cities. They looked at low-income children in both settings, and broke down the results by race and ethnicity as well. In virtually all instances, the charter students did worse than their counterparts in regular public schools.

The story, fed by a tip from the American Federation of Teachers, distorts the data, writes Eduwonk.

Most importantly, though, when one controls the data for race it turns out there is no statistically significant difference between charter schools and other public schools. You’ll search in vain in the Times story for that information though. In fact, to the contrary, a chart accompanying the story fails to offer readers any significance tests for the numbers they’re looking at, inaccurately indicating that there are significant differences by race.

Is this important? Yes, since charters in this sample disproportionately serve minority students by an almost 2-1 margin compared to traditional public schools.

Eduwonk also doubts that the AFT is acting “more in sorrow than anger” when it promotes anti-charter news.

They just don’t like charter schools, they’re not reluctantly concluding that they don’t work, they’re fervently hoping and working to ensure that’s the case.

The Times story did raise the critical question: Do charters “cream” the children of motivated parents or attract students who were having trouble in traditional schools? There’s some evidence that charter students start out behind comparable students but make faster progress.

One previous study, however, suggests that tracking students over time might present findings more favorable to the charter movement. Tom Loveless, director of the Brown Center on Education Policy at the Brookings Institution, who conducted a two-year study of 569 charter schools in 10 states found that while charter school students typically score lower on state tests, over time they progress at faster rates than students in traditional public schools.

Unfortunately, the federal data won’t track charter students over time to see how they improve.

Most charters are new schools still getting up to speed. Charter teachers often are young and inexperienced, though idealistic. I’ve seen a lot more troubled than cream-of-the-crop students in charters. But the bottom line is that the charter concept doesn’t guarantee that every new school will work; it promises that ineffective schools will improve quickly or shut down.

Update: Education Secretary Rod Paige complains that the Times story ignored the differences between charter and non-charter students.

Center for Education Reform cites data showing strong achievement by charter students.

About Joanne


  1. Jim Martin says:

    Read the data closely, public schools creamed the charter schools. I await the spin that is already beginning, the demonization of public schools has been going on too long to allow a few facts to get in the way.
    You recognize the truth when you see how hard the government tried to hide these results.

  2. Annaeus Seneca says:

    Strange, it appears to me that Ms Jacobs has read the data pretty closely. If controlling for race evens the test scores of public and charter schools, what closer reading changes that? If that is a fact that is getting in the way, whose preconceived notions is it in the way of?

    Since she is the one who actually gave specific facts, it seems to me that you should consider whether you are the one who isn’t letting facts get in the way of his preconceived notions.

  3. As EduWonk points out, when controlling for race the differences are not statistically significant. From the AFT report linked:

    “Compared to their peers in regular public schools, black and Hispanic charter school students scored lower both in math and reading in grade 4, but the differences were not statistically significant. The achievement gaps between white and black students and between white and Hispanic students were about the same in charter schools as in regular public school”

    I would argue that including income, as well as other controls, would show that other differences are statistically insignificant as well.

  4. Jim Martin says:

    This is just spin, spin, spin. Read the report for yourself. Ms. Jacobs is vehemently opposed to public schools and takes part in the right-wing demonization and and unjust criticism.
    This report is a hit against Mr. Bush and this group will work hard to spin this to the right. Read the report and don’t let them think for you. Again, the public school scores totally creamed the highly touted charter schools.
    This bunch is not out to educate but to indoctrinate.

  5. Richard Aubrey says:

    Right, Martin. Now, about the facts….

  6. Michelle Dulak says:


    Given that the only reason I have access to the report at all (the vaunted NYT site doesn’t provide a link) is that Joanne linked to Eduwonk, who linked to it, I can’t see how she’s involved in suppressing it.

    Charter schools don’t take off except where the existing public schools are lousy and parents want an alternative. It’s safe to assume, I think, that the children leaving standard public schools for charter schools have had a sub-average public-school education up to that point. If they were to hit educational parity a year later, it would be almost literally miraculous. Consider me unstaggered by the revelation that they didn’t.

  7. “Where Is Your God Now? eh?”
    From Edward G. Robinson’s performance
    in “The Ten Commandments”

  8. mike from oregon says:

    Jim –

    Did you go to Eduwonk, or read the pdf file version of the report? Here is a quote from the end of Eduwonk article “…notes that the NAEP tested less than one percent of charter school students” Give me a break, there is no way that the report is anywhere near indicative of ANYTHING. Geez, give me a break, test a minimum 10% or how about going for the whole enchalada and test 100%. Now your looking at some data with teeth, but not even 1%. I could do that, let me find a kid who is doing really horrible in whatever school (public, charter, private) allow me to test him and then let’s proclaim that it shows how that school setting isn’t working. Sheesh.

  9. Good for Michelle D and Mike from O. I can only resonate with them and respond to both camps on this issue “So what?” . (More on my blog if anyone’s interested.)

  10. This report doesn’t tell me much. Most states set very specific conditions for charter schools. I have seen some pretty strange thematic charters – often for kids that public schools don’t want anyways. It seems that in many states you have to be quite different to get a charter. (… and approved by the state’s public education establishment) There are also bitter battles over charter schools whose themes are higher expectations and more rigor. The idea is that if a town’s schools are meeting or exceeding the state’s (low) expectations, then why should the town have to pay to send a student to a charter school. (Especially one that takes away the better students.) This has led to moratoriums on charter schools in many states.

    I like the idea of charter schools and choice, but, unless the laws governing the creation of charter schools are altered, I don’t think you can compare charter schools with public schools. I have yet to see a charter school (in our area) to which I would send my son.

    Of course, I see this argument as all about control versus individual choice. The rich get to choose, the poor do not. Charter schools are the best attempt (outside of full vouchers) to break the monopoly of public schools and to provide choice. It is just not much of a choice yet. Some would be more than happy to use this report to see even this limited choice go away.

  11. Richard Brandshaft says:

    1) Few of us have the time and skills to trace a study back through the raw data and how it was gathered. Doing it right would mean weeks to years of tedious work. Fortunately, there is a short-cut: If the study results agree with your politics, the study proves your point. If it contradicts your preconceptions, that means someone was fiddling with the numbers. It only remains to construct a plausible argument about how the numbers were fiddled with. (If all else fails, an accusation of outright lying will do.) In this case, when it looked like charters were doing better, it was because they had children of the most motivated parents. When it seems they are doing worse, it?s because they have the problem students. A life?s lesson from computer programming: things are a lot quicker and easier if you don?t insist on correct answers.

    2) I didn?t read the study (see (1)), just the New York Times report. The article talks about what are effectively averages. That?s a mistake that crops up in many forms. In this case, its a shade less explicit than in most. For the purpose of this argument, I will take the results at face value.

    Charter schools are experiments. (Numerous people have pointed out, correctly, that experimenting on children is justified because the normal system is doing so badly.) Big surprise: more experiments fail than succeed. Average over them all, and it?s not surprising that charters do worse. The proper way to approach the problem is to see which (if any) charter schools are doing better than public schools, and try to figure out why.

  12. Shannon Rundquist says:

    Ah, but Richard – if some were doing better than public schools, that would mean the public school was doing worse and that would be a slap in the face to them.

  13. Title of a book seen displayed on the library counter a few eons ago when I was in High School:
    “How to Make Statistics Lie”.
    First thought in my mind when I heard about this result from the AFT on the news – are they REALLY comparing apples to apples? Somehow Joanne’s input confirms my suspicions. And a 1% sample!!! Give me a break!! Let’s at LEAST start with a statistically significant sample. 1% MY FOOT!

  14. DragonFly says:
  15. Some posters have criticized the study because the NEAP tests less than 1% of charter school students. But the important number is the number of students tested, not the percentage of the total.

    Most national surveys by major polling groups are conducted using samples less than a few thousand people, a fraction of a percent of the US population, but these surveys are still considered to be the gold standard of polling. The fact is that if the subjects are randomly selected, using more than that number does not substantially increase the reliability of the results.

    There may be problems with the study, but as long as enough individuals are tested, sample size as a percentage of the whole isn’t one of them.

  16. James Eliason says:

    As an educator of over 30 years, working in such diverse populations as Micronesia, Papua New Guinea, and Houston, Texas, I believe the system is broke and needs to be fixed. In reading of this study, I too was suspicious, and immediately went to the detailed results of Delaware’s DSTP testing of all the schools in the state (found here: ). Unfortunately, there appears to be some collaborative evidence of charter schools failing to achieve a difference (particularly noting that five of the worst seven schools in 3rd grade math are charter schools).

    However, not all charter schools are equal, in my experience. They are not really immune from the same public school tendancy to be “cash cows” for the local community. When a charter school is seen as a way to put money back into the community through hiring practices, one cannot expect to see significant educational advances.

  17. I can’t speak for all charter schools, but I can say that at our charter school, we take in a lot of students who are getting behind in reading and math. Some of them are behind because they are bored with public school’s one-size-fits-all programs, and some because they really need the extra attention that the charter school can give (in our local school district, giving extra attention is discouraged by policies). There are also some students (like my daughter) who were never allowed to get behind, and are now way ahead of where they might have been if they were enrolled in the public school from the start.

    However, by the end of the year they are all charter school students, and our school kicks the collective tails of all the local schools on standardized tests. A few years ago, before the formation of the charter school, one of our school districts’ officials was on the radio bragging that half of his students were reading at or above grade level. The charter school is making us all realize how dreadful that statistic is– and how preventable.

  18. Mike in Texas says:

    Quote from the article:

    “Most importantly, though, when one controls the data for race it turns out there is no statistically significant difference between charter schools and other public schools. You’ll search in vain in the Times story for that information though.”

    In other words, if you manipulate the statistics, as this article is accusing the AFT of doing, you can get the desired results. Of course, the next questions is how do public schools do if you factor out race?

    Race and ethnicity are actually the REAL factors by which test schores are measured here in Texas (except for Asians). You can have 100% of your Caucasian and Asian kids pass the test but if an ethnic group scores below a certain score your overall rating is lowered.


  1. Charter Schools vs. Public Education

    On the whole, the article leaves you with the feeling that the whole charter school movement is a bust, a failure, and a drain on the funds necessary to improve public education, not to mention an indication of the failures of the Bush administration’…