The IMPACT policy was compared to other studies of district-specific policies or ran-
dom trials related to performance incentives in a single district. One conclusion the research-
ers posit is that “teacher incentives are more effective when they are viewed as enduring rath-
er than a temporary pilot.” The authors also suggest that the descriptive evidence presented
show there are “strong incentive contrasts” for teachers at both the highest and lowest levels
of performance, but especially for teachers at the lowest levels of performance (2015, p. 284).
Dee and Wyckoff abridge these findings in the following manner:
One way to frame the magnitude of these effects is
to note that just above the ME [Marginally Effective]
threshold, roughly 20 percent of teachers did not
return to DCPS [Washington, D.C. Public Schools] in
the subsequent year. An ME rating that increases this
attrition by 11 percentage points implies an increase
in teacher attrition of more than 50 percent (2015, p.
288).
Further Data on Teacher Movement
National Data Sources
The National Center for Education Statistics is a project of the United States Department of
Education. Over the past twenty years, the NCES has compiled data on a range of topics related to
education. Some of these statistics center specifically on teacher preparation and movement. Of the surveys
completed by NCES, the most relevant source of data will be the two linked surveys known as the Schools
and Staffing Survey (SASS), and the Teacher Follow-Up Survey (TFS). The SASS and TFS are sent to
district personnel, school personnel and school media centers. These specific surveys attempt to distinguish
between three types of teacher movement. Teachers are classified according to the following methodology:
“teachers who were still teaching in the same school were classified as ‘stayers,’ teachers who were teaching
in a different school were classified as ‘movers,’ and teachers who left the profession since the SASS
interview in the previous year were classified as ‘leavers’” (Goldring, Taie, and Riddles, p. 1). Further analysis
of teacher movement within DCPS or across the state of Florida could be completed using similar
methodology to the surveys described previously. Since Florida has several large school districts, a statewide
program could be modeled after this example. This would provide a rich source of data for analyzing the
movement of teachers across the state.
Duval County Data
Jacksonville Public Education Fund research found that “Duval County is losing on average about
one of every two new teachers hired within the first five years of their employment” (Patching the Pipeline:
Addressing Teacher Satisfaction and Retention, 2013). This longitudinal study examined personnel data
from the 10 years prior to the publication of the study. This report also includes quantitative analysis of
survey results related to the future plans of current Duval County Public School teachers. This research
could be expanded by building on the research of Boyd, et. al. (2010) and Ronfeldt (2012) with the goal of
creating policies related to teacher retention.
Risk Modeling
An increasing number of for-profit and not-for-profit firms have implemented organizational
psychology research to develop predictive risk modeling for employee turnover. Allen, Weeks, and Moffitt
(2005) suggest that there are “…specific dispositional traits that influence the intentions–behavior
relationship” between intending to quit a job and actually quitting. Further research into employee risk
modeling at the P-12 level could yield potential policy solutions that proactively address teacher turnover.
10