Measuring Teacher Quality: Do Teachers Impact Students’ Adult Outcomes?

Value-added measurements of teachers, which capture the ability of a teacher to raise student test scores, are a controversial tool as more districts use them to evaluate teachers and even tie them to teachers’ salaries. Some critics discount this tool by arguing that the impact of a teacher raising students’ test scores is short-lived. In “Measuring the Impacts of Teachers II: Teacher Value-Added and Student Outcomes in Adulthood,” a September 2013 working paper from the National Bureau of Economic Research, authors Raj Chetty, John N. Friedman, and Jonah E. Rockoff continue their research on value-added measures and find that a high quality teacher in elementary or middle school improves students’ outcomes over time, even into adulthood.

The authors use a large urban school district’s administrative data from 1989-2009 to obtain students’ math and English test scores, teacher assignments, and demographics, and track 87 percent of those students in federal income tax records from 1996-2011. In total, the study follows approximately one million students, 77 percent of whom are eligible for free or reduced lunch, from elementary school to early adulthood.

The authors use two research designs. The first tracks students, controlling for observable characteristics such as previous test scores and demographics, as the students were assigned to teachers with various value-added scores. The authors find that encountering a teacher whose value-added score is one standard deviation above the mean improves the students’ chance of college enrollment at age 20 by .82 percent and improves the chance of enrolling at a higher quality college. These students at age 28 also earned 1.3 percent more income than their peers. Higher teacher quality was also associated with improved non-monetary life outcomes for students, such as a reduced chance of teen pregnancy and increased chance of living in a higher income neighborhood and saving for retirement.

The authors’ second research design uses a quasi-experimental approach that takes advantage of teacher turnover. Researchers compare changes in adult outcomes for consecutive cohorts of students to changes in the mean value-added scores of their teachers. This approach yields results on college enrollment and higher quality colleges similar to those from the first design but cannot replicate the results on non-monetary adult outcomes due to a lack of data on the cohorts into adulthood.

Although the study concludes that higher teacher quality leads to improved outcomes for students over time, female students tend to experience more improvement in life outcomes than male students, as do students with higher family income than those in lower income brackets. Additionally, even though raising students’ English test scores is more difficult than raising their math scores, the authors find that a high quality English teacher corresponds to a greater increase in students’ college quality than a high quality math teacher. Heavy attention is paid to early childhood interventions, but the authors highlight that fourth grade to eighth grade teachers have significant impacts on students’ adult outcomes. Finally, the authors suggest that the cost of improving teacher quality – whether by teacher training, removal of low value-added teachers, or other policies – is worth the expected gains in adult earnings later in life, estimated to be an additional $14,500 per person.

One drawback of the study is that the additive impact of teacher quality over consecutive years could not be researched because students’ previous test scores were one of the control measures. They suggest performing a similar study on students’ adult outcomes with a different evaluation tool like principal or classroom observations to determine how much of students’ outcomes can be measured by value-added tools versus other tools. The authors also caution against the use of value-added tools when tied to teacher evaluations and pay due to the risk of teachers teaching to the test or cheating.

Measuring teacher quality continues to be a controversial topic as Race to the Top funding has caused many states to revamp their teacher evaluation systems. This report by Susanna Loeb and Christopher A. Candelaria describes potential inaccuracies to consider when using value-added tools. Future policy decisions should adopt an approach to measuring teacher quality that employs multiple measurement tools that also respond to the economic inequality that underlies the education system.

Feature Photo: cc/(Merrimack College)