Get The Weekly Bancroft & Annual Newsletter delivered straight to your inbox!

Good Teachers Equal Lasting Gain

Good teachers make a big difference in their students’ lives.  Most of us remember a teacher or two who was particularly special.  For me it was Myrtle S. Sherman, who taught French at Cape Elizabeth High School in the 1950’s and 60’s.  One did not come to Madame Sherman’s class unprepared.  She did not tolerate slacking.  I was someone who glided along without doing much work – until I encountered Madame.  Two years of French with her left me with a new respect for hard work, and a glimpse of worlds beyond Cape Elizabeth that lifted my aspirations significantly.


Unfortunately for education in this country we do not know how many Madame Shermans we have, nor do we have any way of telling just how much better they may be than their contemporaries. Recently, however, we have had a breakthrough in research. A study: Long Term Impacts of Teachers by economists Raj Chetty and John N. Friedman of Harvard and Jonah E. Rockoff of Columbia suggests that student test scores are actually a better indicator of teacher quality than is commonly accepted.


These economists tracked a large sample of students, 2.5 million, over twenty years in a study much more comprehensive and longer-term than any that has hitherto been published. The study is dense, full of economist’s formulas, but the conclusions are striking:  replacing a poor teacher with an average teacher would raise the lifetime earnings of a single classroom by $266,000; replacing a poor teacher with an outstanding teacher and lifetime earnings increases to $532,000.  Think about that in the context of the number of classrooms in the United States.


Central to the study’s analysis is that student test scores are, in fact, a good proxy for teacher quality.  Two of the study’s authors were highly skeptical of the validity of test scores as an indicator of teacher quality when they initiated the work.  However, the authors found that, after controlling for factors such as socioeconomic background, and looking at test scores over a period of years, the scores consistently identified good, average, and poor teachers. The differences are substantial. For example, poor teachers’ students had results that were two standard deviations lower than an average teachers’.


After categorizing teachers in this way, the researchers looked at the histories of their students over the long-term, assessing such factors as earnings, college matriculation rates, and where they ended up living.  Students with good teachers were less likely to become pregnant as teenagers, more likely to enroll in college, and, as noted earlier, likely to earn more money as adults.


The study’s authors argue that school districts should use test scores and other so-called value-added measures and remove teachers who are the lowest performers, regardless of the disruption such action might cause.  As Professor Friedman puts it:  “The message is to fire people sooner rather than later”.


As you might imagine, this is not a message that America’s teachers unions are happy to hear.  The teachers unions, particularly the largest and most powerful, the National Education Association, have been fundamentally opposed to using value-add measures to evaluate teacher performance.


In Maine this drama played out at the end of the last legislative session of the Brennan Administration when a special Task Force with representatives from the teachers union (Maine Education Association), and the associations representing superintendents,  principals, and School Board members tried to hammer out an approach to teacher evaluation that included value-added measures.  The MEA resisted all attempts to add such factors.


The teachers unions have their heads in the sand.  This would not be so bad if our children’s future were not at stake.  Bravo to Professors Chetty, Friedman, and Rockoff for doing the kind of systematic analysis that makes clear what the stakes are in identifying poor performing teachers and indicates a way to determine just who those teachers are.


The solution to teacher evaluation is clear – value-add measures need to be an essential part of teacher evaluation, and superintendents and principals need to have the authority to remove poor teachers.


There are lots of details still to be worked out.  Student test scores need to be supplemented with other assessments to make value-add measures even more accurate .  Value-added measures themselves are necessary but not sufficient in an appropriate teacher evaluation approach.


Nonetheless, we are close to a solution that can have substantial impact on improving education in this country.  The teachers unions need to wake up and be a partner in this process.