Improving Teachers, Improving Communities
A teacher aids her students in completing schoolwork. A child’s place of birth should not determine his or her success. With income inequality growing in the United States, policymakers have begun looking at innovative methods to improve areas with persistent poverty. HUD’s Promise Zones initiative approaches neighborhood improvement as a multifaceted issue. In conjunction with Choice Neighborhood grants, the Promise Zones program looks at education, housing, justice, agriculture, and economic development as a means of improving communities.
In the previous issue of The Edge, we presented research by Raj Chetty and colleagues on economic mobility in the United States. One of the study’s primary findings was that teacher quality has a significant impact on students’ outcomes in adulthood. To demonstrate this correlation, the researchers used a value-added metric to show how improvements in teacher quality can lead to better economic and social outcomes for children.
The value-added model is a means of measuring teacher contribution to student learning. The study, conducted by economists from Harvard and Columbia, examines teacher quality by using a “big data” approach. The researchers analyze large quantities of data over extended periods to identify trends and correlations.
The breakthrough in their research on teacher effectiveness occurred when they obtained data from New York City school districts dating back to 1988. This data included the teacher assignments and test scores of more than 2.5 million students who collectively took 18 million tests. The research team then linked the data with tax records, identifying household income, college attendance, and mother’s age at birth for the same students.
The value-added measurement compares student test scores from one year to another to demonstrate a teacher’s contribution. The metric uses state tests administered at the end of third grade, for example, to predict the score students are likely to obtain at the end of fourth grade. The researchers then looked at the average test score gains (or losses) for all of the students that a particular teacher taught.
In “Measuring the Impacts of Teachers I: Evaluating Bias in Teacher Value-Added Estimates,” Chetty and colleagues ask, “[A]re the [value-added] measures we construct ‘unbiased’ predictors of teacher quality?” Critics of these metrics have raised concerns about “student sorting,” or the differences in the types of students assigned to different classes. Is it fair, for example, to compare two teachers’ value-added measurements when one works with students in special education and one does not?
To address these differences, Chetty and colleagues adjusted students’ test scores based on student characteristics such as race, gender, parental income, eligibility for free or reduced-price lunches, enrollment in English as second language classes, and others. Even with these adjustments, the value-added score was still a clear correlate of student success. The researchers concluded that value-added metrics can realistically depict teacher quality.
Controversy and Criticism
School districts nationwide have been debating the value-added measurement of teacher quality. Measurements of teacher effectiveness are often controversial. In 2010, Michelle Rhee, then the chancellor of the District of Columbia public school system, used value-added measurement to reward high-performing teachers and fire low-value-added employees. The teachers rated “poor performers” as a result of Rhee’s policies criticized the value-added metric as being unfair to educators.
Chetty recently was called upon to testify as an expert witness in a case involving teacher quality and its effects on student success. In Vergara v. California, nine public school students filed suit against the state of California for tenure policies that they claimed “kept poor teachers in place.” In June 2014, Judge Rolf Treu ruled that California’s public school tenure policies violated student rights to a quality education, disproportionately affecting minority students. The decision garnered considerable media attention and protests from teachers unions.
The primary objection to the value-added metric is that it relies too heavily on standardized test scores. The Economic Policy Institute discusses the possible factors that researchers and policymakers do not take into account when looking at student test scores, such as school conditions, parent involvement, family resources, and community participation. Others protest that value-added measures do not include metrics that allow creativity or unique approaches for engaging students. Other teachers object that they experience a ceiling effect when students already achieving high scores do not improve significantly from year to year.
A 2014 study in the American Educational Research Journal compares value-added measures with observation-based teacher measures from school principals. The authors find that these two measures are weakly correlated. The study findings show that evaluations from supervisors incentivize different behaviors than value-added measurements: “How we evaluate teachers will likely affect the character of the learning environment and the teachers and teaching that students experience.”
Indeed, the value-added metric is not the only way to evaluate teacher quality. When asked if the research should be taken at face-value, Chetty responded, “I think the main message of our study is that standardized-test score impacts can be a useful input into evaluating teachers, but by no means are we saying that test scores are the end-all and be-all of how teachers should be evaluated.”
In “The Long-Term Impacts of Teachers: Teacher Value-Added and Student Outcomes in Adulthood,” Chetty and his colleagues explore the impact of teachers on students as adults. His research asks: does the teacher you have in 3rd or 4th grade impact how much you are earning at age 30? After linking the students in the dataset with income tax filings, Chetty and colleagues found that students with high-value-added teachers are more likely to attend college and less likely to have children as teenagers than those with low-value-added teachers. Furthermore, young people with high-value-added teachers earn higher salaries as adults than those with low-value-added teachers.
To demonstrate the impact of a low-quality teacher in monetary terms, Chetty and colleagues estimate the costs of retaining poor-quality or low-value-added teachers. If the poorest performing teachers — those with value-added scores in the bottom 5 percent — were replaced with average quality teachers, the lifetime income of students would increase by approximately $250,000 per classroom. This figure demonstrates that monetary impact of investing in high-quality teachers.
The research of Chetty and colleagues shows a powerful correlation between education and students’ long-term outcomes. Although this research does not cover other aspects of good teaching, such as creativity and student engagement, it does show that investments in education can have promising returns in community development. As the Promise Zones initiative begins to examine innovative approaches to cyclical poverty, education will be a central part of the discussion about how to build stronger communities and improve student outcomes in the long run.