Turnover of U.S. teachers is a major reason for our educational problems. In 1987, the modal value of teacher experience has dropped from 15 years to just one year in 2007-2008. Today, 50% of new teachers drop out within five years of entering teaching. Obviously, more and more students are being taught by novice teachers. Once teachers get enough experience to start having positive effects, they quit.

I won't go into the reasons teaching seems unappealing. While salary can be a factor, in my experience with teachers, it is working conditions they find most objectionable.

Numerous other sources have debated such causes as lack of status, misbehaving students, apathetic parents, our general anti-intellectual culture, etc. What I want to highlight here is a new study that explores the relationship between teaching quality and experience - with emphasis on science teaching.

The first good thing this study did was define teaching quality in terms of value added, and they used huge numbers of students and teachers. They monitored over the course of five years the effect of high school teachers on 1.05 million end-of-year exams from over 624,000 individual students and 7,961 teachers. The study covered all the science and math courses and three non-science courses (history, civics, English). Of the issues they examined, two were especially noteworthy: 1) do novice science and math teachers improve with experience? and 2) does the experience effect vary by subject matter? The researchers framed the study this way because numerous prior studies made it clear that teachers and teacher experience are clearly the most important variable affecting student learning.

The next good thing they did was use sophisticated statistics that adjusted for other variables such as prior achievement of students and classroom and school environments. They also disentangled any possible affect due to the possibility that teachers who leave teaching early are less effective, thus giving a misleading impression that the remaining teachers are more effective than they really are.

Results showed that in all subject areas, teacher effectiveness increased in near-linear fashion for the first three years. But then a plateau was reached, and actually showed a decrease by year five for biology teachers. Results also showed science and math teachers who quit by year four were typically less effective than those who stayed on the job.

The implications seem clear. New teachers benefit greatly from early teaching experience, but soon "top out" in effectiveness. Moreover, it is the science and math teachers who have the greatest capacity for improvement, with the steepest growth curves observed for teachers of physics and chemistry. Obviously, students don't learn as well from new science and math teachers as they do from more experienced ones.

In terms of the job market, the high turnover of teachers leads to lower average effectiveness. Another way to think about it, not mentioned in the paper, is the possibility that students don't learn science and math very well because their teachers are not very effective in early years and many of them leave teaching by the time they get good at it. This could account for the poor showing of U.S. math and science students compared with students in other countries.

It's time that educational policy makers addressed three problems revealed by this study: 1) attract better science and math students into teaching careers, 2) provide better initial preparation of science and math teachers, and 3) reduce high teacher attrition.


Henry, G. T., Fortner, C. K., and Bastian, K. C. (2012). The effects of experience and attrition for novice high-school science and math teachers. Science. 335: 1118-1121.

You are reading

Memory Medic

Aging Shrinks the Brain

Can you keep your brain from shriveling?

Teaching Children to Be Honorable

Children are naturally dishonorable.

Do We See the World Like a Movie?

Maybe the answer will help us understand working memory, thinking, and free will