Search EdWorkingPapers

Search EdWorkingPapers by author, title, or keywords.

Educator labor markets

Luis A. Rodriguez, Julie Pham, Briana K. Gonçalves.

Disparate turnover among teachers of color remains a persistent educational challenge, yet little research explores the link between school leadership, peer teaching staff, and turnover disparities. This study explores whether principal and peer teacher demographics predict teacher turnover in New York City, and whether they do so differently for teachers of color. We find teachers are less likely to exit when their principal and a higher share of peer teachers are of the same race/ethnicity, with Black teachers having especially lower transfer rates with a higher share of Black peer teachers. However, results suggest school leadership style and positive teacher relationships are not differentially associated with turnover for teachers of color. We conclude with a discussion of implications.

More →


Tuan D. Nguyen, Elizabeth Bettini, Christopher Redding, Allison F. Gilmour.

Many studies rely on public sector employees’ reported career intentions instead of measuring actual turnover, but research does not clearly document how these variables relate to one another. We develop and test three ways in which measures of employee intentions and turnover might relate to one another: (a) intention may measure the same underlying construct as turnover; (b) intention may be distinct from but strongly related to turnover; or (c) intentions may be distinct from turnover. Using nationally representative data on 102,970 public school teachers, we conduct a descriptive and regression analysis to probe how teachers’ turnover intentions are and are not associated with attrition. While there is some variation across measures of intent, we find evidence most consistent with the second scenario; intention is distinct from, but strongly related to, turnover. We offer recommendations for how researchers should use public sector employee intentions in research.

More →


Brendan Bartanen, Andrew Kwok.

Strengthening teacher supply is a key policy objective for K12 public education, but understanding of the early teacher pipeline remains limited. We leverage the universe of applications to a large public university in Texas from 20092020 to examine the pipeline into teacher education and employment as a K12 public school teacher. A unique feature of Texas's centralized higher education application is it solicits potential interest in teacher certification. We document sharply declining interest in teaching over the period. Further, we show that nonwhite, male, and high-achieving students are substantially underrepresented in teacher education. Particularly for race/ethnicity, these disparities are only partially explained by differences in interest at application.

More →


Virginia S. Lovison, Cecilia H. Mo.

While investing in the teacher workforce is central to improving schools, school resources are notoriously limited, forcing school leaders to make difficult decisions on how to prioritize funds. This paper examines a critical input to resource allocation decisions: teacher preferences. Using an original, online discrete choice survey experiment with a national sample of 1,030 U.S. teachers, we estimate how much teachers value different features of a hypothetical teaching job. The findings show that (a) teachers value access to special education specialists, counselors, and nurses more than a 10% salary increase or 3-student reduction in class size, (b) investments in school counselors and nurses are strikingly cost-effective, as the value teachers alone place on each of these support roles far exceeds the per teacher cost of funding these positions, and (c) teachers who are also parents treat a 10% salary increase and a child care subsidy of similar value as near perfect substitutes. These novel estimates of teachers’ willingness to pay for student-based support professionals challenge the idea that inadequate compensation lies at the root of teacher workforce challenges and illustrate that reforms that exclusively focus on salary as a lever for influencing teacher mobility (e.g. transfer incentives) may be poorly aligned to teachers’ preferences.

More →


Michael Bates, Michael Dinerstein, Andrew C. Johnston, Isaac Sorkin.

We study whether reallocating existing teachers across schools within a district can increase student achievement, and what policies would help achieve these gains.  Using a model of multi-dimensional value-added, we find meaningful achievement gains from reallocating teachers within a district. Using an estimated equilibrium model of the teacher labor market, we find that achieving most of these gains requires directly affecting teachers' preferences over schools. In contrast, directly affecting principals' selection of teachers can lower student achievement. Our analysis highlights the importance of equilibrium and second-best reasoning in analyzing teacher labor market policies.

More →


Jessalynn James, Matthew A. Kraft, John Papay.

We explore the dynamics of competitive search in the K-12 public education sector. Using data from Boston Public Schools, we document how teacher labor supply varies substantially by position types, schools, and the timing of job postings. We find that early-posted positions are more likely to be filled and end up securing new hires that are better-qualified, more-effective, and more likely to remain at a school. In contrast, the number of applicants to a position is largely unassociated with hire quality, suggesting that schools may struggle to identify and select the best candidates even when there is a large pool of qualified applicants. Our findings point to substantial unrealized potential for improving teacher hiring.

More →


Todd R. Jones, Arielle Sloan.

We use roster data of 96 top U.S. economics departments to document the academic origins of their tenure-track faculty. Academic origins may have implications for how undergraduate (B.A.) and doctoral (Ph.D.) students are trained and placed, as well as the type of research produced. We find that faculty educated at top-ranked Ph.D. universities are overrepresented; e.g., over half of our sample attended a top 15 university, and over a third attended a top six university. We find similar, but less pronounced, patterns for B.A. origins; e.g., over a third of those with a U.S. B.A. attended a top 15 university.

More →


Katharine Meyer, Kelli A. Bird, Benjamin L. Castleman.
With rapid technological transformations to the labor market, many working adults return to college after graduation to obtain additional training or credentials. Using a comparative individual fixed effects strategy and an administrative panel dataset of enrollment and employment in Virginia, we provide the first causal estimates of credential “stacking” – earning two or more community college certificates or degrees – among working adults. We find stacking increases employment by four percentage points and quarterly wages by $375 (four percent). Returns are larger for individuals studying in Health and who return to college after first completing a short-term certificate.

More →


Andrew C. Johnston, Jonah Rockoff.

As unfunded pension liabilities grow, governments experiment with ways to curb costs. We examine the effect of a representative cost-cutting reform on the retention and productivity of workers. The reform reduced pension annuities and increased penalties for early retirement, projected to save 8 percent of revenues. We leverage administrative records and a discontinuity in the reform to estimate its effect on labor supply. The reform slightly increased worker retention, and we can rule out small attrition effects. The reform had no effect on worker output. The extensive and intensive margins of labor supply appear to be maintained under the reform.

More →


Aaron Phipps.

Using administrative data from D.C. Public Schools, I use exogenous variation in the presence and intensity of teacher monitoring to show it significantly improves student test scores and reduces suspensions. Uniquely, my setting allows me to separately identify the effect of pre-evaluation monitoring from post-evaluation feedback. Monitoring's effect is strongest among teachers with a large incentive to increase student test scores. As tests approach, unmonitored teachers sacrifice higher-level learning, classroom management, and student engagement, even though these pedagogical tasks are among the most effective. One possible explanation is teachers ``teach to the test'' as a risk mitigation strategy, even if it is less effective on average. This is supported by showing teaching to the test has a smaller effect on student test score variance than other teaching approaches. These results illustrate the importance of monitoring in contexts where teachers have the strongest incentive to deviate from pedagogically sound practices.

More →