Better Schools, or Better Teachers?

How much does school choice matter? Probably less than you think, as Levitt has previously argued. Now, in an analysis of seven years of test-score data from 6,000 Los Angeles teachers, the L.A. Times, assisted by a senior Rand Corp. economist, has found teacher effectiveness to be three times more influential than school attendance on student performance. What’s more, the Times reports, “the best teachers were not concentrated in schools in the most affluent neighborhoods, nor were the weakest instructors bunched in poor areas. … The quality of instruction typically varied far more within a school than between schools.” More data from the study is here. (HT: Amanda Ripley) [%comments]

Tom Kelly

School choice has never been tried on a meaningful scale. Until we have at least a major city worth of private choice schools in operation for about 10 years, it is foolish to judge the idea one way or another.

Markets take time to adjust to consumer preferences. Choice schools need that time.

Tom Kelly

To clarify my first post, we should not be looking at the results of teachers in choice schools or at the results of choice schools. We need to look at the over time results of a choice system compared to our existing system.

The big improvements that we could have using an all private choice system are indeed systemic. There will still be lousy teachers and even some lousy schools but the incentives will be aligned so the problems get better rather than worse.

The bonus of a private choice system would be to slow down urban sprawl. As young families would no longer have to move to exurbia in search of affordable housing in good school districts.


Yet the best teacher's are paid the same as the worst teachers.


I think this analysis ignores the theory that administrations will respond to the ability of students to move by aggressively recruiting the best teachers, removing bad ones.

Jeremy Miles

David - not in Los Angeles. Teachers are never removed in Los Angeles public schools.


More telling than the numbers, were the observations by reporters. Amazingly, the teachers who kept their students' attention and challenged them to think critically were the teachers with the highest gains. The sweet and loved teachers at my school with the ability to do neither of the above are kept around because they are sweet and loved, not because they are good at what they do. That should change.


As long as there are teacher unions we will have the issues we currently have.

A teacher union is actually closer to a professional sports league union than a trade union. In a trade union like carpenters after a worker spends 3 to 4 years learning the trade they become a full-fledge carpenter. Here is the kicker that worker with 4 or 5 years experience makes the same as the worker with 25 years experience. This is not the case for teachers. A teacher with 25 years experience is going to make a ton more than a teacher with say 5 years experience. Who is the better teacher? We don't know. Yet that is the most important factor to our child suceeding.

I would not mind the teacher unions if all full-fledged teachers made the same amount. But having teachers that make over 100K and get a pension is crazy when there are workers in the workforce willing to do the same job for less.

Justin James

At the same time, better schools attract better teachers. Here in SC, schools are very hit or miss, with many more misses than hits. I've talked to a number of teachers, and believe me, the competition between them to teach at the better schools is FIERCE. This is a state with a 65% high school graduation rate. How many teachers who really care about success want to work in an environment where 1/3 of the students would rather sit at home without a diploma than pay attention? Sure, there are a few "superhero" teachers who look forwards to that challenge. but not many. Or to put it this way, Would the Freakonomics folks be willing to leave Chicago U. and head out to... I don't know... some dumpy school with crumbling facilities, toilets that don't work, students that will throw stuff at them and sleep in class, be high or drunk through class, perhaps have a weapon with them, and insult them all day long? Because that's what a "bad school" looks like. If the answer is "yes" then maybe I will say that the school doesn't matter. But the answer isn't "yes" then I've made my point. And do you REALLY think those kinds of conditions do not affect the quality of education much?

Yes, the teacher matters more than the school. But it is so hard to get a good education, and so few good teachers want to teach in those conditions, that a better school environment is a baseline requirement for a good education.



Scott W

Anyone looking at the effect of good teachers on student attendance?

Eric M. Jones

And you have to wonder if Harvard's ranking (for example) is because of their "reputation"....which is due to their reputation....

Lots of things are like this. Good students go to the schools where good students go. It's the brand.

Frankly, I still vote for bulldozing the whole thing and starting again. The USA ranks 22nd in Science, 27th in Math, and 33rd in Reading.


What I cannot understand for the life of me is why parents accept that, in a particular grade or subject, their child will receive an education that is not as good as that of the children in the next classroom.

Why can't we identify the best teachers in the world for every grade/subject, place them on well-produced DVDs, add images, bullet points, music, etc., so that EVERY student gets the advantage of the best teachers?

The other teachers could serve to supplement the DVDs, administer tests, etc.

A child absent from school could access the course on-line (or at a later time in the library, perhaps) and stay up to date.

A brilliant child could perhaps complete the course in a shorter period of time, going on to future lessons.

The slower child could listen to the lesson again, until he/she understood it.

Lessons could be interactive.

As time goes on, it would be relatively easy to edit the DVDs so that updated material could be inserted, or problem areas could be addressed more vigorously.

Yes, I know that will replace many teachers in the classroom setting, but if we then used those teachers for security, discipline, etc., maybe our schools would do even better.

If nothing else, we could do this for just a core group of courses that are essential to further learning, etc.

ALSO, at the college level, why not create a SPECIFIC set of CORE courses--the required courses for every freshman/sophomore--and have them taught via DVD/online?

This would reduce the number of professors needed, since the students would have to study during off-hours, would ensure a more better and more standard result (since everyone is getting the same high-caliber teaching), and it could even reduce students expenses due to charging less for these courses.

Why should a freshman at one college get only a "decent" history teacher, while a freshman at another college gets a brilliant one? License the lectures of the best teachers and allow all of our students to obtain superior educations--at least in these core courses.



Just about any school has at least one incredible teacher, and at least one really terrible one--a great variance at that one school. Of course that difference is going to be a bigger difference than between the average teacher at school A and the average teacher at school B.

It's like saying the difference in height between the tallest woman and the shortest woman is greater than the difference between the average man and the average woman, and concluding therefore that talking about height differences between men and women is meaningless.


My experience with going to "good" schools my entire life is that it is more about the schools, and specifically the intellect of the students that they attract, than about the teachers. The combination of the increase in competition due to smarter peers, the exposure to a higher level of intellect throughout every conversation throughout your day, and the cooperative nature of study sessions combine in some way that I qualitatively felt made much more of an impact on me than my teachers, who were for the most part mediocre at best.


Everyone likes the idea of paying teachers more if they have more experience or more education or a credential -- but those factors are all statistically insignificant when it comes to teacher effectiveness.

No one, I presume, would like the idea of giving teachers added pay for being female or Asian -- but the data show that these two attributes are associated with greater effectiveness.

Of course the best course of action is to find some way of determining who the best teachers are and to given them a pay differential -- no matter what their education, gender, experience, race, etc.

Drill-Baby-Drill Drill Team

Education has many players, and all have to be proactive if the system is to excel. Teachers, students, parents and ancillary staff all have a role to play. BUT teachers tend to get all the blame...Much of the time the children were not well prepared at home and in prior education years and began lagging years ago.

Kids are lazy and don't want to do homework, with their sports interest, skate boarding, video gaming and air guitar band. We have done a poor job at parenting; too permissive and failing to make a child accountable for performance. Our education system needs to be stricter and more demanding of students. The Laid-Back California Approach has failed.

Old ineffective teachers need to be purged(FIRED). Lazy students need to be failed. Everyone has to be made responsible and accountable, right down to spelling homework assignments. Military discipline has a role as required.



School choice is merely a drop in the bucket when discussing student performance. This particular droplet gains greater emphasis due to the nature of the educational system. A system that narrows the selection of schools to choose from also limits the selection of teachers. The real question is whether it is wise to restrict students to the schools within their residential boundary or are we better off allowing the invisible hand balance the scale?


I don't think standardized test scores are an effective way to measure learning. We live in a school district that does not even offer recess. Many of the children attend privates schools. It is difficult to imagine as a layperson that the children attending our public schools are in fact receiving an equivalent education. It is possible that a school that teaches to the test may have better exam scores than the Montessori or progressive schools, but I do not believe the children are better prepared for life.

Teachers are important. Curriculum is also important.


@Jeremy Miles

Absolutely right. Unionized teachers are almost never fired.
True school choice would let students take their funds anywhere (oublic/private/religious) A natural outgrowth of this would the unions would relent or public schools would wither on the vine.
The variations in schools currently is likely the effect of teachers wanting to work at 'better' schools and the small efforts of administrators to be gatekeeper. Imagine if administrators had more incentives and a free hand.


It's interesting to me that we do not accept the variation in teacher ability as a mere byproduct of humanity. It is no different with parents...and they are much more indicative of an individual child's future success.

Bottom line: Some teachers are better than others. But the same is true for all professions. In other words, some doctors are much more effective than others. Airline pilots. Accountants. Or engineers.

While I applaud our continual efforts to measure and understand the differences in ability or effectiveness, we will never eradicate it. On the other hand, we might better price these skills...but in doing so, we will also limit their (fortuitous) reach to those who cannot afford them.


- About the lack of good school choice research --- There HAVE been large-scale implementations of school choice initiatives with positive results. I attended a session at the AERA educational conference where the studies were presented. One is in Washington DC and the other was, I think, in Texas, and positive educational and local economic gains were found.

The Journal of School Choice probably has this data, and more.

- Of course there are differences within schools. Every parent knows that, which is why there are nearly fights to get students with certain teachers. But also, of course there are differences between schools. I lived in an area where the local school district averaged 50% graduation and the district 15 miles to the north averaged 90% graduation. Yes, there were probably socioeconomic factors, but that wasn't the whole story. I did research in the district to the north, and it was simply better run, better managed, had a better vision, and attracted better teachers because they did education well.

- @AaronS - my expertise is in educational technology, and what you propose would never work. Computers and videos cannot replace teachers, because it has been tried multiple times since Skinner's famous teaching machines of the 1950s. Check out Walter Lewin's physics lectures on iTunes because he is about the best professor you'll ever find, especially on video. The videos are definitely great resources, but after viewing several of them, tell me if it's the same as having a real instructor in class.