Roland Fryer Identifies Five Habits of Successful Charter Schools

Harvard economist (and Freakonomics friend) Roland Fryer has a new paper out (full version here) that takes a look at the specific successful habits of charter schools. Along with co-author Will Dobbie, Fryer collected “unparalleled data” on 35 elementary and middle charter schools in New York City by conducting extensive interviews and videotaping classrooms.

Their results are fairly counter-intuitive. They showed that traditional solutions like class size, per-pupil expenditure, and the number of teachers with advanced degrees are not correlated with effectiveness, and in fact, “resource-based solutions” actually lowered school effectiveness.

Instead, they found five qualities that made up about 50 percent of a charter school’s effectiveness.  These are:

1. Frequent teacher feedback
2. Data driven instruction
3. High-dosage tutoring
4. Increased instructional time
5. Relentless focus on academic achievement. 

For example, a high-achieving charter middle-school teacher gets feedback 13.42 times per semester, versus 6.35 times at other charter schools. Similarly, high achieving middle-schools test their students 4 times per semester, compared to 2.4 at other schools.

Perhaps even more interesting, these qualities remained paramount to a successful charter school despite different styles of environment, such as “Whole Child,” “No Excuses” (like KIPP), and teacher-retention.


I am an assistant principal at a public school and have limited knowledge on the operations of a charter school. The only thing I know is that every year I receive one or two students who were dismissed from a charter school, because they could not cut the mustard. Yes, they must take the kids from a lottery but, they don't say anything about getting rid of that kid once they find out the student won't assimilate to their strict rules and procedures.

There are vast differences between public and charter schools and far too many reasons why the public schools are the way they are.

A serious analysis needs to be done to accurately examine both options.


all of you are blessed because u know the difference between charter school, public school, private schools and i dont.can somebody explain to me the difference?me,the school i went to in in small african village is the school of life.we learn differently because our realities are totally different from yours.our way is much long as we can self rely,that is took me too long to type this email because my child is teaching me english and some it skills.he is 10 and i will graduate from his school soon but he too will graduate from mine because motherhood is a free school.he is teaching me free because he says that is his way to pay back the lessons i taught day, i am sure i will join one of all the schools you mentioned in your debate.good luck to you all.

Billy D

So, let's restate those 5 qualities more cynically. The 5 qualities that most influence a charter school's test score effectiveness are:

1. Frequently reminding teachers that tests are important.
2. Instructing to the test.
3. High-dosage tutoring to the test.
4. Extra time spent instructing to the test.
5. Constantly remind everyone about the tests.

So we're back to the old argument: Do high test scores = better education?

Enter your name...

It depends: Does your test happen to measure a good education?

For example, frequent testing of basic arithmetic facts leads to more students being able to quickly and accurately give the right answer to questions like "What is 6 times 7?". If you believe that mastering those facts is part of a "good education", then you ought to be giving those tests, and the scores on those tests will accurately tell you whether the students are getting (that part of) a good education.

This works for both simple issues (for example, how well can you type) and for complex ones (how well can you sort out a complex mechanical problem or how well can you analyze a work of literature). The only real difference is that tests for simple subjects are easier to design and much cheaper to grade.

On the other hand, if you think that "good education" means acquiring certain attitudes (for example, about people, education, government, or society) or character qualities (for example, perseverance, empathy, or love of liberty), then the answer is no: a structured test is not likely to tell you whether you're getting a "good education".



I am a middle school teacher. I can tell you ONE SIMPLE CHANGE that I believe needs to be made to "fix" education.

I can do a good job teaching a class of even 100 WELL-BEHAVED students...or I can do a mediocre job teaching a class of 25 students if a number of them are poorly behaved. Yet school district after school district mixes in "rough" kids with well-behaved ones.

Now, I'm not talking about ESE students who may have learning difficulties. I have some of those in my class and they are some of the quietest, hardest workers! They understand that they will have to exert more effort to get the same result as some of the other students.

No, I'm talking about students who are BEHAVIOR PROBLEMS. The ones who don't care, who may be defiant, constantly in trouble, come late, enjoy skipping, etc. These students should not be mixed in the with general population! It makes teaching and learning difficult when every few seconds you have to play "whack-a-mole" with disruptive students.

SOLUTION: Put a single great teachers in with LARGER (but well-behaved) classes...but put multiple teachers in with difficult students. Two strong teachers in a class of 7-15 difficult students can make a difference for THOSE students! For they are forced to pay better attention, are (hopefully) a bit more intimidated by having so many adult eyes on them at all times, and will, I trust, put forth a much better effort on their papers. I would almost wager my annual salary that this will work wonders--for well-behaved and not-so-well-behaved students!

How do we find out who is who? The teachers will tell you. Names that keep coming up over and over...move them into these special, smaller, better supervised classes. Sometimes, the difference between and great class and a terrible one or two students. It's just the lifeboat game all over--sometimes someone has to go in order that the others can survive. Fortunately, we'er not throwing these kids overboard, but simply loading them onto another lifeboat--one that can do a better job bringing them to shore.


Jack Skellington, Esq

It's interesting the amount of antagonism drawn by teachers unions on this thread. It's almost as if it were an existing talking point being echoed by the downstream masses with zero attempt at intellectual rigor. Surely it can't be that. No, it's far more plausible that organizations who represent people who have chosen education as a profession are ruthless extortionists who care nothing for children. Makes sense.

Unions aside (the effects of which this research doesn't touch upon *at all*), I took the extraordinary, nay, borderline insane, step of actually reading the paper.

Do differentiate from the group of comments who clearly haven't, I'll offer a few points I found interesting.

Let's begin with

"Market-based reforms such as school choice
or school vouchers have, at best, a modest impact on student achievement...(various cites)...This suggests that competition alone is unlikely to significantly increase the efficiency
of the public school system."

So the working basis of the paper basically assumes the fact that the problem absolutely isn't that public school teachers are complacent or that parents (through a market based solution) are very good at choosing the best schools for their children.

There's a fair amount of the paper dedicated to discussing this at length. There's also this interesting correlation, oddly absent from the post summary here, but one might think, at least superficially much more profound.

"Teachers at high achieving schools also work longer hours than teachers at other charter schools; an additional 7.75 hours per week at the elementary level and 10.29 hours per week at the middle school level. Despite this higher workload, the maximum salary of teachers at high achieving schools is the same or somewhat lower than other charter schools."

Interesting that "Instructional staff works 25% longer for no increase in pay" is transformed to "Increased instructional time". I imagine most organizations would be more effective if employees worked 25% more hours for free. One suspects the correlation here is probably stronger than the correlation with "Relentless focus on academic achievement" If only there were a school where they relented, but the staff still worked overtime in their noble pursuit of mediocrity.

Left out, also, from the paper (correctly) and (mostly)the comments here, is the inherent selection bias of a lottery system. Children who have parents motivated enough to opt into a selection process for they perceive is a "better" school are difficult to control for when compared to the populations of schools where all children are not only excepted, but frequently forced by law to attend, regardless of motivation.

While there's a good attempt to control for other variables, the crux of the concept really begs the question on the point of motivated instructors. Teachers who are willing to work additional hours and receive more feedback are more successful educators seems to be the overarching point. The problem, of course, is the "chicken or egg" nature of how you attract these people to a school.

In all cases, the study revolves around teachers who are generally an exception, working outside of the "normal" public education system. Are those without advanced degrees more motivated because they lack opportunity in the traditional public setting? Would they continue to be as motivated if these "philosophies" were applied to the larger set of public schools? What are the teachers ages, and other demographic data. Not really discussed, but would be interesting. Do these teachers not have advanced degrees by choice, or are they 25? There are a lot of open questions. The data is interesting, although the conclusion not particularly compelling.

Challenging children to do well with teachers willing to accept feedback and work long hours for free correlates with achievement. Great. There are numerous reasons this can work in the institutions that are the "exception" and that it can't be applied to the entire set of public schools. No, there probably aren't "hero" teachers, and that myth is dangerous, but it's likely that there isn't a "hero" philosophy, either. Education is a stunningly complex interaction among numerous systems and actors. There aren't "5" anything that easily explain achievement gaps.

"It's complex" rarely sells mangy page views, or gets one tenure, though, I suppose.



Basically, the study shows/concludes that if you obsess over test scores (points 1, 2, 3, and 5), then you will get higher test scores. I don't think that's very revolutionary; in fact, it's called cheating in some places, or gaming the system in other places.

Maybe more qualified teachers get lower test scores because they understand child and cognitive development and refuse to teach to the test.


As a teacher in a high-achieving urban charter school which is beating the odds, I can tell you that what is widely maligned as "teaching to the test" is actually, quite simply, teaching the standards! Do those who object to "teaching to the test" really want to get rid of the curricular standards set for each grade level? I bet that they, as parents, were very on top of their own children being "kindergarten ready." Not all parents have the educational tools to do this. Hence, schools like ours which incorporate all five of the qualities Fryer's research found to be essential.

Steve Chambers

It always comes down to the Unions.

Unions drive mediocrity.

Kill the Union and performance will rise.