How To Build a Better Teacher
Is firing bad instructors the only way to improve schools? New evidence suggests
that poor teachers can learn to be good ones.
By Ray Fisman|
July 25, 2012

Perhaps good teachers can be made after all, and not just born.

The conversation about how to improve American education has taken on an
increasingly confrontational tone. The caricature often presented in the press
depicts hard-driving, data-obsessed reformers—who believe the solution is getting
rid of low-performing teachers—standing off against unions—who don’t trust any
teaching metric and care more about their jobs than the children they’re supposed
to be educating.

But in some ways the focus on jobs misses the point. As New York State Education
Commissioner John King has pointed out, with the exception of urban hubs like
New York and L.A., few school districts have the luxury of firing low-performing
teachers with the knowledge that new recruits will line up to take their places.*

If we take firing off the table, what else can be done to resolve America’s education
crisis? The findings of several recent studies by psychologists, economists, and
educators show that—despite many reformers’ claims to the contrary—it may be
possible to make low-performing teachers better, instead of firing them. If these
studies can be replicated throughout entire school systems and across the
country, we may be at the beginning of a revolution that will build a better
educational system for America.

The view that good teachers are born, not made, is based on the many studies
that have found that various training credentials and certifications have no effect
on a given teacher’s “value-added,” the amount by which he or she increases the
test scores of students above and beyond what you’d expect based on their
performance in earlier grades. A degree in education seems to make no
difference. Nor do higher salaries. (Value-added measures have their own set of
critics, who wonder whether the measures—or even the underlying test scores—
capture anything of use. Yet recent research does suggest that the students of
high value-added teachers go on to earn significantly more later in life.)
But there’s a big difference between saying that we have yet to find an approach
that has been shown to have a measurable impact on a teacher performance and
claiming that none exists. Indeed, the Gates Foundation and others are making big
bets that the secret to teacher improvement can be found, and there’s reason to
hope that they may carry the day. They point to success stories like the Harlem
Children’s Zone and the Knowledge Is Power Program charter schools, which
consistently improve the test scores of students randomized into their classrooms
through lotteries. Pedagogy gurus like Norm Atkins, meanwhile, have developed
teacher-training curricula that they believe can make anyone teach better.
But do the training regimens at these schools actually contribute to their success?
There are many reasons why a particular school or charter system might succeed
in getting its students to excel, and not all of these have to do with improving the
classroom performance of teachers who are already on staff: Superstar charter
networks may be particularly well led (education commissioner John King once ran
the Uncommon Schools, another high-performing charter system), or maintain a
policy of getting rid of bad teachers (or even unmotivated students). It may also
simply be too expensive to scale their approaches across the country. The Harlem
Children’s Zone erased the black-white achievement gap in math, but doing so was
very expensive. As Harvard economist Roland Fryer put it, “The HCZ model
demonstrates that the right cocktail of investments can be successful. The
challenge is to find lower-cost ways to achieve similar results in regular public
Yet there is growing evidence that you may not need to hand out stacks of pink
slips—or have a very tall stack of greenbacks—to improve teacher quality. When I
asked education scholar Doug Staiger where the most promising evidence lay, he
referred me to an assessment of the Teacher Evaluation System that was
implemented in Cincinnati public schools in 2000-01.
Cincinnati’s approach combines evaluation by expert teachers—who observe
classroom performance and also critique lesson plans and other written materials—
with feedback based on those evaluations, to help teachers figure out how to
improve. The study that professor Staiger described, by Eric Taylor of Stanford
and John Tyler of Brown, focused on teachers in grades 4-8 who were already in
the school system in 2000, which allowed the researchers to examine, for a given
teacher, the test scores of their pupils before, during, and after evaluation was
performed and feedback received. And because the TES was phased in gradually,
the researchers could compare the performance of teachers who had already
been evaluated and received feedback to those who were still awaiting their TES
treatment. This ensured that any change in test scores wasn’t just the result of a
general improvement in Cincinnati’s schools concurrent with the implementation of
The results of the study suggest that TES-style feedback and coaching holds
promise—Taylor and Tyler estimate that participating in TES has an effect on
students’ standardized math test scores that is equivalent to taking a teacher that
is worse than three-quarters of his peers and making him about average. The
effects of participation only get stronger with time: If teachers were simply
performing better because they saw their evaluator sitting at the back of the
classroom, you’d expect only a onetime improvement in student outcomes during
the evaluation year. Instead, TES participants’ performance is even greater in
subsequent years. And the expense of creating, if not a great teacher, at least a
decent one, is fairly modest—the cost of TES was about $7,000 per teacher.
(Unfortunately, Cincinnati’s approach to evaluation and feedback has yet to catch
on—a 2009 survey by the New Teacher Project found that school districts rarely
use evaluation for any purpose other than remediation and dismissal.)
But the results are still tentative: The researchers followed only 105 teachers into
the program, and were only able to observe 61 of these following participation in
TES—the others stopped teaching in grade 4-8 classrooms where students took
state math tests, or participated in TES only in the final year of the study. Also, the
authors note that TES had no effect on reading scores (though generally there’s
less evidence of wide variability in the ability of individual teachers to consistently
raise students’ test scores in language arts).
How did the teaching style of TES participants change as a result of their
experience? And which aspects of teaching actually mattered for helping students
perform better? A separate study by Taylor, Tyler, and others tries to deconstruct
Cincinnati’s teaching evaluation into its constituent components to see which ones
really mattered. For example, which is more important: to make sure that students
are well-behaved and focused on their work, or to promote critical thinking among
students? The results are speculative at best—for the most part, teachers who do
well on one aspect of teaching also do well on others, making it difficult to parse
out the effect of any one component.
But if it’s unrealistic to expect that we’ll ever discover a pedagogical silver bullet
that makes a great teacher, it may be possible, guided by insights from social
psychology, to find individual interventions that do have outsize effects on student
learning at relatively modest cost.
Another intriguing example of this comes from my colleague at Columbia
University, psychologist Valerie Purdie-Vaughns. Vaughns is part of a team of
researchers that has run a series of experiments at schools in New York City and
elsewhere in the Northeast to examine whether the way teachers provide feedback
to students can have a material effect on their performance, particularly for
minority students who often feel threatened by stereotypes of low academic
Providing feedback to underachieving students presents a catch-22. Constant
criticism can be demotivating. But Lake Wobegon-style feedback, where even low-
quality work wins lavish praise, quickly loses any meaning, or may even lead
students to believe that they’re incapable of doing well and thus held to a low
Vaughns and her colleagues conjectured that providing feedback that is critical
and holds a student to a high standard, but at that same time makes clear that the
student is capable of excellence, would achieve better results. Through a series of
double-blind experiments, the researchers tested their theories in high schools
and middle schools to see the effect that this type of “wise” feedback had on
student effort and accomplishments.
In one experiment, for example, Vaughns and her colleagues simply appended a
note to teachers’ comments on student essays that said, “I’m giving you these
comments because I have very high expectations and I know that you can reach
them.” 64 percent of black students who received the note were motivated to
revise their papers, as compared to 27 percent of a control set of students who
received a note that simply stated, “I’m giving you these comments so that you’ll
have feedback on your paper.” Students who received “wise” feedback ended up
with higher grades, as well. Based on the results of a related experiment, the
researchers suggest that simply explaining to minority students that critical
feedback from teachers should be seen not as putdowns but as an indication of
high potential may go a long way in reducing the achievement gap between blacks
and whites.
A second study, released this week by the National Bureau of Economic Research,
exploits the tools of social psychology to motivate teachers rather than students.
The researchers (including Roland Fryer of the Harlem Children’s Zone study)
experimented with incentive pay for teachers at a group of K-8 schools. Past
efforts at pay-for-good-teaching haven’t been very effective in improving student
performance. Fryer and his colleagues tweaked the pay-for-performance
approach to give teachers a $4,000 bonus payment upfront, and informing them
that some or all of the money would have to be returned if their students failed to
meet performance targets. This approach takes advantage of the fact that most
people will work harder to avoid a loss rather than earn a bonus—so-called loss
aversion—an idea first developed by famed psychologists Daniel Kahneman and
Amos Tversky. The authors’ calculations suggest the “loss incentive” was
extremely motivating, enough to transform a bad teacher into a mediocre one, or
to provoke excellent results from a teacher that had previously been just average.
Both the wise feedback and pay-for-loss-avoidance studies should be seen as
promising, yet preliminary—there’s a lot that can go wrong in scaling up such small
trials to a nationwide rollout. But they represent exactly the sort of experimentation,
built on well-founded insights, that can help to get better teachers in front of
America’s students—and maybe even help to bring the warring factions in the
education debate a bit closer together.
San Diego Education Report
San Diego
Education Report
Team dysfunction
School Reform
Silence is Golden
Public records

Brown Act Permanent
Media, Secrecy v. Free
Teachers Union CTA
San Diego County Office
of Education
List of School Districts
Evaluating teachers

Gates Foundation
Retaining the best teachers