More teacher experience may have a few more benefits than the research has been touting (not that we really think it didn't). A new study by Helen "Sonny" Ladd and Lucy Sorenson finds a positive relationship between a teacher's years of experience in the classroom and students' achievement and non-cognitive behavioral skills, into at least a teacher's 12^th year of teaching, far surpassing the 5^th-year crest that most research is telling us represents the end of the climb.

Using a sample of about 250,000 students in North Carolina (a state known for its robust data system) over five years, the researchers find that years of experience are positively related to student test scores, especially in math. The effect is not insubstantial: the positive impact on student achievement from a teacher with five years of experience is about twice as much as the disadvantage presented by a student's race or family income. Those benefits from experience continue to grow until a teacher has about 12 years of experience. The study also found positive (albeit weaker) correlations with students' non-cognitive and behavioral skills, including student absenteeism, disruptive classroom behaviors, time spent completing homework and time spent reading for pleasure.

Notably, Ladd and Sorenson found that "high ability" teachers leave the profession more often than those of lower ability.

There are a few methodological concerns to keep in mind: non-random assignment of students to teachers (which happens often when principals reward veteran teachers with higher-ability classes or parents request certain teachers) could throw off estimates. Also, because the data set only includes five years of data, even with the use of teacher fixed effects, no teacher is compared to herself across 12 or more years (for example, the data set does not show Ms. Smith's effectiveness as both a first-year and a 15^th-year teacher). To sum up, while this research finds a relationship between experience and efficacy, it cannot conclude that experience causes teaching efficacy.