Understanding Teacher Quality Gaps Between Advantaged and Disadvantaged Schools: How Did They Form, and How Can We Close Them?

Research Question #1: How have teacher quality gaps in public schools changed over time, and to what extent are these gaps due to differences across districts, across schools within a district, and across classrooms within a school?

In two states (NC and WA), for each year of observable data and for every measure of teacher quality (experience, licensure and value-added estimates (VAM)), we find that disadvantaged students (measured by race and ethnicity– URM and eligibility for free-reduced priced lunch -- EDS) are more likely to have lower-quality teachers (Goldhaber et. al, 2018b).

• North Carolina: The sources of inequity are largely due to within-district sorting.

1. Teacher Experience: Across the years of available data, disadvantaged students are between 2 – 4 percentage points more likely to be exposed to a novice (5 or fewer years) teacher.
2. Licensure Tests: Disadvantaged students are 5-8 percentage points more likely to be exposed to a lower quartile licensure teacher.
3. Value-Added Estimates: Disadvantaged students are 3-6 percentage points more likely to be exposed to a lower quartile teacher (in a separate paper, Goldhaber et al. (2016), we show that the magnitudes of estimated VAM gaps vary by model specification).

• Washington: The sources of inequity are largely due to between-district sorting; this is similar to evidence from Massachusetts (Cowan et al., 2017)

1. Teacher Experience: Across the years of available data, disadvantaged students are 1-5 percentage points more likely to be exposed to a novice teacher.
2. Licensure Tests: Disadvantaged students are 5-7 percentage points more likely to be exposed to a lower quartile licensure teacher.
3. Value-Added Estimates: Disadvantaged students are 3-8 percentage points more likely to be exposed to a lower quartile teacher.
   1. These differences in average exposure to effective teachers has implications for student achievement gaps: if advantaged and disadvantaged students were assigned to teachers of the same value added between grades 4-8, the achievement gap would be 10% smaller (Goldhaber et al., 2018c).

Research Question #2: To what extent do patterns in teacher hiring, within-district transfers, cross-district transfers, and attrition influence these gaps?

We simulate TQGs to assess the extent to which four processes —teacher attrition from each state workforce, teacher mobility within districts, teacher mobility across districts, and teacher hiring— contribute to TQGs between advantaged and disadvantaged students. We find that for each combination of student disadvantage and teacher quality, all four processes contribute to the inequity between advantaged and disadvantaged students (Goldhaber et. al, 2018a).

• North Carolina: The hiring of teachers into new positions contributes to the most of the TQGs. Within-district mobility also contributes to a substantial portion of the gaps across the different measures of teacher quality.

1. Teacher Experience: Teacher attrition contributes to about one-third of the simulated gaps we see between advantaged and disadvantaged students. Within-district mobility contributes between 10% – 20% of the simulated gaps.
2. Licensure Tests: Teacher hiring explains over half of the simulated gaps between advantaged and disadvantaged students. Both within-district and cross-district gaps contribute to approximately 14% of the simulated gaps.
3. Value-Added Estimates: Consistent with licensure tests, the hiring of new teachers contributes to approximately one-third of the simulated gaps.

• Washington: For each measure of student disadvantage and teacher quality, the hiring of teachers into new positions contributes to the majority of TQGs.

1. Teacher Experience: For URM students, the hiring of new teachers into new positions and teacher-attrition each contribute to about one-third of the simulated gaps. For ED students, each of these processes contributes to approximately 20% of the simulated gaps.
2. Licensure Tests: The hiring of new teachers contributes to over two-thirds of the simulated gaps.
3. Value-Added Estimates: The hiring of new teachers contributes between 80% and 90% of the gaps we simulate.

• In our simple simulation, we are unable to incorporate a truly dynamic model which includes continuous measures and peer effects. Future iterations will allow us to simulate the effects of different policy interventions.

References

Cowan, James, Dan Goldhaber, and Roddy Theobald. "Teacher Equity Gaps in Massachusetts." ESE Policy Brief, Massachusetts Department of Elementary and Secondary Education (2017).

Goldhaber, D., Quince, V., and Theobald, R. (2016). Reconciling different estimates of teacher quality gaps based on value-added. Washington, DC: National Center for the Analysis of Longitudinal Data in Education Research (CALDER), American Institutes for Research.

Goldhaber, D., Quince, V., & Theobald, R. (2018a). How Did It Get This Way: Disentangling the Sources of Teacher Quality Gaps Across Two States. CALDER Working Paper 209-1118-1.

Goldhaber, D., Quince, V., & Theobald, R. (2018b). Has it always been this way? Tracing the evolution of teacher quality gaps in U.S. public schools. American Educational Research Journal, doi 10.3102/0002831217733445.

Goldhaber, D., Theobald, R., & Fumia, D. (2018c). Teacher quality gaps and student outcomes: Assessing the association between teacher assignments and student math test scores and high school course taking. CALDER Working Paper 185.