Home Articles Reader Opinion Editorial Book Reviews Discussion Writers Guide About TCRecord
transparent 13

External Contextual Factors and Teacher Turnover: The Case of Michigan High Schools

by Elizabeth Covay Minor, Guan K. Saw, Kenneth A. Frank, Barbara Schneider & Kaitlin T. Torphy - 2019

Background/Context: All organizations face turnover in their workforce; however, in schools high turnover can interfere with the effectiveness and efficiency of the school. While past research has examined school-related factors linked to teacher turnover, few studies have examined how external contextual factors are related to teacher turnover. This study examines the role of two external contextual factors in teacher turnover: economic downturns and changes in state curricular policy (the Michigan Merit Curriculum [MMC]).

Purpose/Objective/Research Question/Focus of Study: This study asks the extent to which the economic crisis of 2009 and the implementation of the MMC are related to school-level teacher turnover rates and whether those relationships vary by school locale and within the school year.

Population/Participants/Subjects: The data include full-time high school teachers in the state of Michigan aggregated to the school level.

Research Design: Using eight years of statewide longitudinal data from Michigan, the study employs school fixed effects models to account for possible differences in unobservable school characteristics that are constant over time and may be related to teacher turnover. The study examines teacher turnover at both the mid-year and the end of the year as teachers leave schools at various points during the school year. Additionally, this study considers how turnover is experience differentially by urbanicity.

Findings/Results: Between 3.2% and 15.5% of teachers left their school over the eight-year period. The rates of turnover varied by the time of the school year with more teachers leaving at the end of the year than during mid-year. There was a significant increase in teacher turnover rates around the announcement of the MMC as well as the economic downturn. While all locations were impacted by the announcement of the MMC, the largest amount of turnover occurred in urban areas and the lowest for suburban areas. In terms of the economic downturn, towns were impacted the most, followed by rural and suburban schools. Urban areas did not see a significant increase in teacher turnover related to the recession.

Conclusions/Recommendations: The authors conclude that external contextual factors are related to increases in teacher turnover independent of each other. How these factors relate to teacher turnover does depend on school locale. While this study was based in Michigan, all states have their own policy and economic pressures to consider in related to school-level decision making and teacher turnover.

Schools, like many organizations, contend with turnover in their workforce (Guin, 2004; Holme & Rangel, 2012, Ingersoll, 2001; Ronfeldt, Loeb, & Wyckoff, 2013). Though some turnover within any organization is normal and sometimes positive, high turnover rates in schools can decrease school effectiveness and efficiency (Holme & Rangel, 2012; Ingersoll, 2001). Past research shows that high teacher turnover has negative effects on school operations and, even more troublesome, on student achievement (Ingersoll, 2001; Guin, 2004; Ronfeldt et al., 2013). Several school-related factors such as principal leadership, student behavior, and school climate have been shown to be linked to teacher turnover (Ingersoll, 2001; Kukla-Acedvedo, 2009; Stuit & Smith, 2010), yet little is known about the role of external contextual factors. In this study, we seek to examine how economic conditions and statewide curricular changes are related to school-level teacher turnover by analyzing eight years of statewide longitudinal data from Michigan.

Economic downturns and changes in state policy increasing course taking rigor could have implications for school-level decisions—especially for the staffing of instructional personnel, particularly the retention of teachers. Although there are individual factors related to a teacher leaving a school, such as personal preferences or family needs, structural school-level changes may also impact teacher turnover. With external factors placing constraints on school leaders, they must make decisions for the good of their students, which may mean doing more with less and a readjustment of staff. For example, literature on managing school funding with reduced budgets suggests that reducing teaching positions strategically can help schools to function effectively within their means (Odden, 2012). Additionally, statewide curricular changes sometimes lead school leaders to prioritize teachers with specialized subject areas such as mathematics and science whereas those in other subject areas such as the arts are considered less essential for schools resulting in a reduction of the teaching staff (Dynarski, Frank, Jacob, & Schneider, 2012a, 2012b).

In addition to limited research on the role of external contextual factors in school-level teacher turnover, questions about the retention of teachers are often fraught with serious methodological problems. Typically, researchers examining these issues use only end of year estimates to determine the explanations for teacher turnover (e.g., Goldring, Taie, & Riddles, 2014; Rondfeldt et al., 2013;). However, teachers leave schools at various points during the school year, and estimates on retention based on end of the year calculations may miss these mid-term fluctuations (Guin, 2004; Redding & Henry, 2018), which may have important implications for school functioning. Another limitation in much of the teacher turnover research is that researchers view turnover only within a limited time period. Using longitudinal information provides a description of the long-term fluctuations in teacher turnover resulting in a fuller understanding of trends in teacher turnover.

Moreover, we argue that one must attend to how turnover is experienced differentially by urbanicity because the opportunities and constraints in the labor market vary considerably across urban, suburban, town, and rural areas. More specifically for the purposes of our study, when examining statewide contextual factors, it is important to explore the differential effects by urbanicity. By including variation by locale, we are able to improve our understanding of how macro-level (state level) forces interact with meso-level (local level) structures in influencing teacher turnover.

This study extends previous work by improving the methodological approaches and the understanding of external contextual factors related to school-level teacher turnover. Using Michigan state data from 2004 to 2011, we leverage twice yearly longitudinal statewide data to differentiate school-level teacher turnover across two points during the school year placed within the context of variation in geographic locations. More specifically, we consider both semester and year for each school and examine our models, which account for time-varying and unobserved characteristics, for all schools and separately by urbanicity. Our study presents empirical evidence on the variation in external contextual factors, specifically economic recession and state educational reform, on school-level teacher turnover across time and locale.


Teacher turnover is an important indicator of school organizational disruption. In 2011–2012, according to the National Center for Education Statistics (NCES), about 16% of the nation’s teachers either moved schools or left teaching altogether to pursue other work (Goldring et al., 2014). Most schools experience some teacher turnover; however, turnover is not equally distributed. A growing body of evidence shows that novice teachers and teachers in schools with significant proportions of minority, poor, and academically struggling students are more likely to leave their school (Allensworth, Ponisciak, & Mazzeo, 2009; Keigher & Cross, 2010; Goldring et al., 2014).

Past research on teacher turnover focuses on a mixture of working conditions and personal reasons/preferences as explanations for teachers leaving. In terms of working conditions factors, recent research on turnover finds that teachers leave due to lack of administrative support, lack of student motivation, poor school social climate, a disruptive behavioral climate, or unfulfilled salary needs (Feng, Figolio, & Sass, 2010; Ingersoll, 2001; Johnson, Kraft, & Papay, 2012; Kukla-Acedvedo, 2009; Stuit & Smith, 2010). Using data from Massachusetts public schools, Johnson et al. (2012) examine the relationship between working conditions, job satisfaction, and career plans. When teachers perceive their own work environment (e.g., colleagues, community support, facilities, governance, principal, professional expertise, resources, school culture, and time) as positive, they are more likely to be satisfied and less likely to plan to leave the school. Indeed, in a study of public and private schools Ingersoll (2001) finds that 27% of those teachers who leave one school for another and 25% of those who leave the profession do so because of dissatisfaction.

Teachers’ personal preferences have also been examined in relation to turnover. For example, teachers who leave disadvantaged schools may be seeking schools with students of racial and socioeconomic backgrounds similar to their own (Kennedy, 2005) or school conditions where they perceive themselves to be more effective (Frank, Maroulis, Belman, & Kaploicohentz, 2011; Jackson, 2013). By examining teacher turnover as an individual decision, it seems that such choices occur within a school-centric bubble independent of larger contextual factors (e.g., Kukla-Acedvdeo, 2009).

We have taken a different approach focusing on more macro conditions as school-level teacher turnover has implications for school-level functioning. When schools are faced with a reduced staff, the remaining staff members may be expected to fill the gaps in the instructional and organizational needs of the school. In other words, teachers may be asked to teach additional courses and participate in more committees, reducing the time and attention that they usually devote to their instruction. Placing more demands on the teachers may reduce their effectiveness as teachers and in turn reduce student achievement. This may also change the social conditions within the schools, which is likely related to teacher effectiveness and whether a teacher remains at the school (Johnson, 2006; Johnson et al., 2012; Kraft, Marinell, & Yee, 2016).

High teacher turnover at any point during the school year may also restrict the flow of resources, especially local knowledge critical for teaching within schools, creating disruption in the functioning of the organization. In particular, when teacher turnover is high, teachers may not develop an identity with the collective of the school. In turn, they may allocate their help or other resources only to specific colleagues that they have come to trust (Frank, 2009). The result is that resources such as curricular materials or local knowledge become isolated in pockets within the school, restricting the flow from where it could be most valuable. Johnson et al. (2012) also find that the social conditions within a school are important for whether or not a teacher stays at that school. If the flow of resources is restricted, then this may lead to the remaining teachers becoming unsatisfied with the social conditions within their school and considering leaving as well.

In our school-level analysis, we examine how external contextual factors are related to teacher turnover. More specifically, our approach includes statewide economic constraints and education policy mandates and their relationship to school-level teacher turnover across geographical locations.


Economic Pressure

Given the public funding of schools, the teacher labor market is closely related to local and state economic conditions. When the nation experienced the recession in 2009, Michigan was already in an economic downturn (see Figure 1). Michigan experienced a “one-state recession” in the early 2000s.1  Figure 1 shows a sustained decline in per-capita personal income for Michigan residents relative to the nation, beginning in 2003 (Michigan Department of Technology, Management, and Budget [DTMB], n.d.). The jobless rate for Michigan steadily increased, from 2003 to 2007. The number of those employed in non-farm jobs decreased by nearly 150,000, including 21,000 in the auto manufacturing industry alone (Labor Market Information, n.d.)—a key employment sector in Michigan.

In 2009, coinciding with the national recession, Michigan experienced a further dip in per-capita personal income (Michigan DTMB, n.d.) and many more employees lost their jobs (Labor Market Information, n.d.), further reducing state revenues. These losses in jobs and resulting out of state migration had a major impact on state revenues. The general education fund decreased from close to 29 million in 2003 to around 20 million in 2010, with a low of about 6 million in 2007 (State Budget Office, 2013). Michigan schools saw a 3.7% decrease in state per pupil funding from 2008 to 2012 (Leachman & Mai, 2014). In other words, the economic downturn in Michigan greatly impacted the financial conditions of public schools.

Figure 1. One State Recession: Real Per-Capita Income in Michigan and the United States from 2000–2010


Source: “Real Per-Capita Personal Income by State: 2000–2010” data table from the Michigan Department of Technology, Management, & Budget website

With less funding, school leaders had to examine their budgets and make decisions about the areas to fund, reduce or eliminate. Reducing staff is one area that school leaders potentially seek to reduce spending. Nationally, Ingersoll (2001) found that 41% of those teachers who move schools and 12% of those who leave teaching did so because of staff cutbacks. Nationally, from mid-2008 to 2012 school districts cut over 300,000 teacher and school personnel jobs (Leachman & Mai, 2014). Goldhaber, Strunk, Brown, and Knight (2016) find that the Great Recession impacted teacher turnover to levels not seen in the past in Los Angeles and the state of Washington. Even the threat of layoff leads teachers to search for opportunities in other districts creating considerable structural churn within schools and districts.


At the same time the country was experiencing a recession and a decrease in school funding, Michigan instituted a new policy to increase college going that required students to complete more advanced level courses in high schools. In the spring of 2006, the Michigan Department of Education announced that beginning with incoming ninth graders in the 2007–2008 school year, all high school students were required to take a rigorous set of courses, called the Michigan Merit Curriculum (MMC), in order to graduate. The MMC specifies 16 required credits, including math courses such as Algebra II and science courses such as Chemistry or Physics (see Jacob, Dynarski, Frank, & Schneider, 2017, for more information on the MCC). Though some schools had previously been offering these courses, a majority of school leaders had to make significant changes to their curricular offerings and staff composition (Byrd & Langer, 2010; Dynarski et al., 2012a, 2012b; Obenauf, 2014). With MMC requirements, many school leaders had to redirect resources, withdraw certain courses such as art and music, reduce the number of staff in de-emphasized subjects, and augment staff in the other core subjects to meet curricular demands (Obenauf, 2014).

In response to state-mandated reform programs requiring changes to school practices such as the MMC, school leaders are likely to adjust or reorganize their teaching staff to meet the needs of the course requirements, which may result in the hiring, retaining or dismissal of some teachers (Keesler, 2010). For example, in Michigan the new curriculum standard requires all high school students to take Algebra II, which means school leaders need to make sure they have the capacity to instruct all students in Algebra II by changing teaching assignments or hiring new teachers. As Michigan schools were coping with the new curricular agendas, some teachers were forced to either change schools or find other sources of employment.


Though changes in personal circumstance, school context, and economic context may impact teacher turnover, decisions are made within one’s local context. In addition to limited focus on the role of external contextual factors on school-level teacher turnover, geographic location is a factor that should be considered to understand teacher turnover, especially with variation in the opportunity structure based on locale. While recent research has examined teacher turnover on a national level (e.g., Goldring et al., 2014; Keigher & Cross, 2010; Ingersoll & May, 2012; Ingersoll, Merrill, & May, 2016) at the state level (e.g., Hanushek, Kain, & Rivkin, 2004; Ladd, 2011), and within particular districts (e.g., Allensworth et al., 2009; Fulbeck, 2014; Ronfeldt et al., 2013), the Midwest has not been yet been studied. Neglecting variation in teacher turnover by locale masks important differences in the local context. For example, since teachers are hired at the district level, the opportunity structure within large urban districts may enable schools to redistribute their teachers from one high school to the next depending on particular school needs. However, that is not the case for those school districts in small towns or rural areas since there are fewer schools in the districts, which means fewer options for redistribution. Michigan is an interesting case to examine this phenomenon given the mix of urban, suburban, town, and rural areas in the state. Michigan’s variation by urbanicity allows important consideration of how school-level teacher turnover may differ across geography. We argue that one must attend to how turnover is experienced differentially by urbanicity because the opportunities and constraints in the labor market vary considerably across urban, suburban, town, and rural areas.

A concern for measuring teacher turnover is timing. It can be challenging to capture larger contextual factors without multiple years of data to gauge longitudinal changes (Singer & Willett, 2003). Using Michigan’s statewide panel data, we are able to study variation in teacher turnover before and after major contextual shifts such as curricular and economic fluctuations, while also paying attention to within- and between-school variations over time. Though teachers generally leave at the end of the school year, those who leave during the year may be the most disruptive, as they create changes in classroom experience within a given year. This study examines these variations in the timing of teacher turnover as well as the impact of external factors on school-level teacher turnover. We suspect economic and policy changes will be related to increases in school-level teacher turnover and these will vary by location. This study examines three questions: (1) To what extent is the economic crisis of 2009 related to school-level teacher turnover rates? (2) To what extent is the implementation of the Michigan Merit Curriculum, a state mandated policy, related to school-level teacher turnover rates? (3) To what extent is there variation in school-level teacher turnover rates by school locale and within the school year?



We use the Michigan Statewide Longitudinal Data System (MSLDS) for the academic years of 2003–2004 to 2011–2012 for this analysis.2 The MSLDS includes detailed information on students, teachers, administrators, and schools for all public schools in the state, enabling us to estimate the impact of contextual factors on teacher turnover rates while controlling for a set of school characteristics.

Our teacher sample is limited to full-time high school teachers (with a full-time equivalent [FTE]3 of one or above) with at least part of their teaching assignment in a high school.4 One unique feature of the MSLDS data is that each year there are two records of teaching assignments early October and late June. Having two collections of data each year allows for the examination of teacher turnover between the two time points during specific school years. Depending on the specific school year-semester over the course of the eight years, our sample ranges from about 18,000 to 23,500 teachers with about 650 to 700 regular high schools in Michigan each semester. We focus on 607 schools that were open during our study time frame for which we have complete data across eight years with two semesters of teacher turnover within each year.  


To compute the school-level teacher turnover rate we first use individual teacher records, which identify where the teacher was teaching for a given semester, by school. If a teacher is not working at her/his current school the following semester or exits from the MSLDS system, that teacher is considered to have left the school. Since teachers may have teaching assignments in more than one school (only about 1%–3% in our data each semester), we focus on the school in which the teacher has his/her highest FTE. We then compare this school assignment to the school in which that teacher has his/her highest FTE the following semester. If a teacher’s primary FTE the following semester is in a different school, that teacher is considered to have left the school. Finally, we aggregate the total number of teachers who left each semester to the school level and divide it by the total number of FTE teachers in the school to produce a school-level rate of teacher turnover. In doing so, our analytic sample includes 16 semesters of calculated teacher turnover rates, from 2004 to 2011 (not including the year 2011–2012). As noted above, we measure teacher turnover twice during the school year. The first compares teaching assignments from October to June, which we call mid-year (MY) turnover. The other measure, which we called end of year (EOY) turnover, compares teachers’ assignments from June to October.

Our calculation of teacher turnover rates includes teachers who have retired because the external contextual factors may have led to unscheduled retirements, but teachers who left due to illness are excluded from the count (accounting for only less than 1% of total teachers leaving). For the most part, around 1% of the teachers retire at each time with the exception of key years. In Table 2 and Figure 2, we separate out the number of teachers who retired to allow the reader to see how many teachers this is. Due to data limitations, there is not enough information to fully categorize why teachers leave. While knowing more about the reason why individual teachers leave would be informative, that examination is beyond the scope of our school-level analysis.


To measure the contextual factors we use school year dummy variables that correspond to the policy change and the recession. The MMC was officially announced in the spring of 2006 and implemented in the 2007–2008 school year for high schools. If we see a significantly higher teacher turnover rate in 2006 compared to 2005, this would suggest that there are immediate effects following the announcement of the MMC at the EOY 2006 as school leaders start to make changes in anticipation of the implementation. It may be the case that the impact of the MMC will only take place when the MMC is in effect; therefore, we focus on effects of dummy variables for 2006, 2007, and 2008 with years corresponding to the announcement and implementation of the MMC.5

The inclusion of year dummies in the models also allows us to capture the possible effects of the economic recession on teacher turnover rates at the school level. According to the Bureau of Labor Statistics (BLS, 2012), when the most recent national recession began in December 2007, employment reached its bottom during late 2009 and early 2010. Thus, we expect to observe the impact of the recent economic recession on the teaching labor force in Michigan in the 2009–2010 and 2010–2011 school years.

We argue that in addition to the external context factors of the implementation of a state-mandated educational policy and the recession, the local context of the school also plays an important role in teacher turnover. To characterize school locale we use the definitions from the CCD (Common Core of Data) with categories of city, suburban, town, and rural.6


With the rich information offered by the MSLDS, together with the data from the CCD, we include a series of additional external and internal contextual factors, which may be related to the teacher turnover rates at the school level, in our models. The first is the unemployment rate, which is one indicator of a recession (BLS, 2012), for the area in which the school is located and where most of the students reside. This local monthly unemployment rate is obtained from the BLS (2015); we average across the year and weight by student enrollment in a given county.

Taking advantage of the comprehensive personnel data from the MSLDS and given the research on factors related to teacher turnover (e.g., Goldring et al., 2014), we construct two school organizational measures for teacher composition and include them as covariates in our estimation models. The first measure of teacher composition is the percent of full-time teachers within a school. Past research finds that full time teachers are more likely to stay than are part-time teachers (Goldring et al., 2014). We calculate the percent of full-time teachers by dividing the number of teachers with an FTE of at least one in a given high school by the number of full- and part-time teachers within that school. Additionally, since Goldring et al. (2014) find that novice teachers are less likely to stay in the same school, we also create a measure for the percentage of novice teachers within a school where a novice teacher is in her first three years of teaching. Lastly, our analyses control for a set of standard school covariates based on student population, including the percentage of students receiving free or reduced-price lunch, the percentage of minority students, and school size as past research has found that the demographic composition of the school is related to teacher turnover (Allensworth et al., 2009; Goldring et al., 2014; Ingersoll, 2001; Keigher & Cross, 2010). In our study, these school covariates come from the CCD.

Student mobility and leadership instability in schools are two important time-varying school organizational factors that are likely to be related to teacher turnover. Yet, prior studies estimating teacher turnover rates rarely considered the influences of these two aspects. Allensworth et al. (2009) do find that student mobility and principal turnover are related to teacher turnover in Chicago elementary schools but not in Chicago high schools providing some support that student mobility/principal turnover and teacher turnover are related in some contexts. In this study, we are able to calculate the student mobility rates by semester for each school by following the same procedure used for computing teacher turnover rates. A student’s school for a given semester is compared to the student’s school the next semester, and then aggregated to the school level. Similar to the identification of full-time teacher, the indicator of leadership instability is computed based on principals who have an FTE equal to one in a given semester. If the principal in a school leaves by the end of the semester, the school has a value of one for leadership instability. However, there is considerable missing data for the leadership instability measure (roughly 27%), and the inclusion of this variable in our models would reduce our sample by a fourth. To decrease the amount of missing data, we also include assistant principals in the measure as they are part of the leadership/administration in a school. When there are more than one principal or assistant principal within a school, the school is considered to have stability in leadership even if one of the principals/assistant principals has left the school in the following semester. While there would still be some instability when one of the principals/assistant principals leave, our measure captures that there is some consistency in the school leadership. Furthermore, to maintain the sample size we also include a measure of whether the school is missing information about leadership instability as a dummy variable. We did run our models with only principal turnover. Since our results are consistent, we opted to include the models with a broader indicator of leadership instability. Table 1 presents descriptive statistics for key variables in these analyses.    

Table 1. Descriptive Statistics (n = 607)



Standard Deviation

2004–2011a School variables


   Teacher turnover rate



   County unemployment rate



   % of free/reduced lunch students



   % of black students



   School size















   % of full time teachers



   % of novice teachers



   Student mobility rate



   Leadership stability



   Leadership instability



   Leadership stability – missing



Source. Michigan Statewide Longitudinal Data System (MSLDS), Common Core Data (CCD), and Bureau of Labor Statistics (BLS).

Note. n = number of schools. aYear in which spring semester of academic year occurred, for example, “2004” refers to 2003–2004 school year.


Studying changes in teacher turnover rate at the school level is of primary interest in the current study. Taking advantage of the longitudinal feature of the MSLDS, where each school has 16 observations (two time points over eight years), we employ school fixed effects models represented by the following equation:


where Yti is the teacher turnover rate for a given school i at a given time point t, [39_22816.htm_g/00006.jpg] is a set of year dummy variables (2005 is omitted), [39_22816.htm_g/00008.jpg] is a dummy variable indicating spring semester (reference group is fall semester), and [39_22816.htm_g/00010.jpg] is a vector of time-varying school characteristics, including county unemployment rate, student mobility rate, principal instability, percent of free/reduced-lunch students, percent of black students, school size, percent of full time teachers, and percent of novice teachers. The regression models include school fixed effects (denoted [39_22816.htm_g/00012.jpg]) to account for possible differences in unobservable school characteristics that are constant over time and may be related to teacher turnover. This implies a control for pre-tests. [39_22816.htm_g/00014.jpg] is assumed a zero mean normally distributed error term. We also conducted hierarchical liner models with our data. The results are substantively the same (results available upon request). We present the school fixed effects models as our final models as they allow us to take into account unobserved differences among schools.



From the beginning of the 2003–2004 school year to the end of the 2010–2011 school year, the number of FTE high school teachers in the state of Michigan decreased by nearly four thousand teachers7 (see Table 2 and Figure 2). Whereas there were 21,779 high school teachers in the fall of the 2003–2004 school year, there were only 17,931 high school teachers in the spring of 2010–2011. This is an overall decrease of about 18%.8 In general, most teachers were retained at the same school from semester to semester; however, anywhere from 3.2% to 15.5% of teachers left (which includes teachers fired, changing schools, and leaving the teaching profession) a school with 1.0% to 4.0% of teachers retired, depending on the semester and the year9 (see Table 2). The percentages for the spring semester were fairly consistent with national surveys of teacher turnover. The 2011–2012 Schools and Staffing Survey (SASS) and Teacher Follow-up Survey (TFS) showed that most public school teachers stay at the same school from one year to the next, but about 16% of teachers change schools or left the profession (ranging from about 12-17% depending on the year; Goldring et al., 2014).

Table 2. Distribution of Michigan Public High School Teachers who will be Retained or Leave, 2004–2011


Time Point

Number of Teachers


Not Retained

Left school/profession


















End of Year


















End of Year


















End of Year


















End of Year


















End of Year


















End of Year


















End of Year


















End of Year








Notes: n = number of teachers. The total number of regular high schools is 607. In each row, the number of teachers who will leave or be retained do not add up to the total number of teachers because teachers who leave due to illness are not reported here (only about 0.1% each semester).

 aYear in which spring semester of academic year occurred, for example, “2004” refers to 2003–2004 school year.

Figure 2. Teacher Turnover Rates Between and Within School Years


Source: Michigan Statewide Longitudinal Data System (MSLDS)

Conventional wisdom would suggest that we would see most of the teacher turnover at the end of the school year (EOY). In Michigan, while most of the observed teacher turnover was EOY (9.4% to 16.7%), there were also sizeable numbers of teachers leaving for various reasons mid-year (MY), which may create more organizational disruption. From 2003–2004 to 2010–2011, the percentages of teachers not retained MY range from 4.7% to 8.1% (sum of retired teachers, those who left for other career options, or who were not retained at their school). For the most part, the percentage of teachers leaving MY was around 5% except for the 2009–2010 and 2010–2011 school years, corresponding to the economic downturn in Michigan. Not only do we note an increase in the percent of teachers no longer at their school during the 2009–2010 school year but also we note that there is an increase in retirements during that school year as well. For the most part, between 1%–2% of teachers retire in a given semester; however, during 2009–2010, 3%–4% of teachers retired.10  

The percentages of teachers leaving for various reasons EOY increased steadily from the 2003–2004 school year, with the exception of spring 2005–2006 and 2009–2010. For EOY 2003–2004, 9.4% of teachers left their current school and the percentage of teachers leaving increased to 12.8% by EOY 2010–2011 school year. The two notable exceptions to the pattern were two large increases for EOY 2005–2006 with 16.7% of teachers not retained and 2009–2010 with 16.2%. These statistics lend preliminary support that teacher turnover increases correspond to the announcement of the MMC (in 2005–2006) and to the economic downturn (2009–2010). Our data also show that teacher retirement rates for both MY and EOY 2009–2010 were unusually high, 4.0% and 3.1%, respectively (more than doubled from the prior years), which contributed to the large increase in the percent of teachers not retained, following the economic downturn.


Using the full analytic sample, our analyses of school fixed effects models suggest that both school internal and external contextual factors play a significant role in influencing school-level teacher turnover (as shown in Column 1 of Table 3). As expected, student mobility and leadership instability are positively related to teacher turnover. This suggests a vicious cycle of a lack of investment when we look at the effects of administrator turnover and student mobility. Our coefficients are consistently positive across settings, suggesting that as some people leave a school, others are more inclined to leave.

Our data also show that school size, percent of full-time teachers, and percent of novice teachers are negatively associated with teacher turnover. An increase in school enrollment may be an indicator that the school is doing well, resulting in less school-level teacher turnover, whether that is because teachers do not want to leave the school or the school needs the staff to support the number of students in the growing school. A similar argument could be made around the percent of full-time staff and novice teachers. If the school is doing well, there is less need and/or desire for teachers to leave. The percent of low-income students and percent of black students are not significantly related with teacher turnover rates. This is likely because the models include school fixed effects leaving little variation in the composition from one year to the next.  

Finally, as shown in the descriptive statistics, teachers are more likely to leave their school EOY rather than MY. At the EOY about 5% more teachers leave their school as compared to the percent of teachers who leave MY in the multivariate analyses.

Table 3. Fixed Effects Models for Predicting Teacher Turnover Rates 2004–2011 (n = 607)


Full sample
(n = 607)


(n = 88)

Suburb (n = 165)


(n = 69)


(n = 285)

County unemployment rate




-0.6098 *








Student mobility rate

0.2323 ***

0.4362 ***

0.2581 **









Leadership instability

0.0291 ***



0.0529 **

0.0240 *







% of free/reduced lunch students


-0.1268 *










% of black students




-0.4916 **








log(school size)

-0.0396 ***




-0.0532 **







% of full time teachers

-0.0537 **




-0.0498 **







% of novice teachers

-0.0488 **



-0.1566 **

-0.0388 *







End of Year (ref: Mid-Year)

0.0497 ***

0.0336 *

0.0338 ***

0.0534 ***

0.0726 ***







Year dummy – 2005 (ref: 2004)

0.0098 **

0.0319 **










Year dummy – 2006

0.0425 ***

0.0674 ***

0.0231 ***

0.0507 ***

0.0412 ***







Year dummy – 2007

0.0109 **

0.0204 *



0.0148 **







Year dummy – 2008












Year dummy – 2009




0.0386 *








Year dummy – 2010

0.0429 ***


0.0400 *

0.0801 ***

0.0608 ***







Year dummy - 2011

0.0214 **



0.0543 ***

0.0359 **








0.3360 ***




0.4137 ***







Number of observationsa






Note. a Some school observations were dropped from estimations (less than 1.2%) because of missing values for certain variables in a given semester. All models include a dummy variable indicating if the school is missing information about principal instability in order to maintain the sample size. Coefficients in italics correspond to the year of the MMC announcement (2006) and the national recession (2010). Tests of significant differences are based on standard errors clustered by school. Robust standard errors in parentheses. p < 0.1; * p < 0.05; ** p < 0.01; *** p < 0.001.

There is evidence of variation in school-level teacher turnover rates within the school year as well as across school years. These years tend to coincide with our external contextual factors of interest (i.e., state-mandated educational reform and the economic recession). Recall that the MMC was announced in spring 2006. Our school fixed effects results indicate that immediately following the announcement of MMC, statistically significantly more teachers left their school compared to the prior year. In particular, in 2006 there was an increase of about 4.3% in the teacher turnover rate, compared to 2004. In 2010, the teacher turnover rate increased by about 4.3%, compared to 2004.

Overall, we find that in the state of Michigan teacher turnover shows increases in the year of the MMC announcement and the year following as well as during the recession. However, the impact of the MMC and the recession may not be consistent across locale given variation in the opportunity structure. Therefore, we estimate school-level teacher turnover rates using city, suburban, town, and rural sub-samples separately.

As was the case for our full sample analysis, there was a statistically significant increase in teacher turnover in 2006 for all locations, compared to 2004. In terms of the announcement of the MMC, which corresponds to 2006, the largest increase in teacher turnover was in city schools with a 6.7% increase compared to 2004, followed by town schools (5.1%), rural schools (4.1%), and suburban schools (2.3%). We see a somewhat different pattern in relation to the recession. The largest increase in teacher turnover was in town schools, with 2010 seeing a statistically significant increase of 8.0% compared to 2004, followed by rural schools (6.1%) and suburban schools (4.0%). The negative effects of the economic recession on school-level teacher turnover in town and rural areas persist into year 2011 but with smaller magnitudes, 5.4% and 3.6%, respectively. Interestingly, the school-level teacher turnover rate in city schools increased by less than 1% in 2011 and was not statistically significantly different from the baseline.11  Taken together, we conclude that for schools in city areas, the economic crisis had less impact because those schools had already been contending with tight budgets. There may also be other factors such as working conditions and difficult job situations captured in the fixed effects portion of the model that are related to turnover in city schools.


The findings of this study have important implications for theoretical and empirical understanding of teacher turnover issues, yet the degree of confidence for any interpretation will depend on the robustness of our inferences (e.g., with respect to potentially omitted variables). To address this issue, we quantify how much bias there must be in our estimates to invalidate our inferences, following the sensitivity analysis procedures suggested by Frank, Maroulis, Duong, and Kelcey (2013). We focus on key findings—the estimated effects of MMC (year 2006) and economic recession (year 2010) for the full sample. As defined by Frank et al. (2013), the calculation of proportion of bias to make inference invalid is as the following:

      % bias necessary to invalidate an inference = 1- threshold for inference/estimated effect,

where the threshold for inference= s.e. × [39_22816.htm_g/00018.jpg]. Applied to our estimates, to invalidate the inference of the MMC effect (year 2006) on teacher turnover rate, bias must have accounted for (1 - 1.96 x .0038/.0425 = .824) about 82.4% of the estimated effect to invalidate the inference.

For interpretation (Frank et al., 2013), to invalidate the inference one would need to replace about 82.4% of our total school sample (82.4% × 607 = 500 schools) with cases in which there is zero MMC effect on teacher turnover. Note that the zero MMC effect must pertain even after conditioning on the school covariates used in this study. Similar calculations suggest that about 48.8% bias must be present to invalidate the inference of the estimated economic recession effect (year 2010). These levels of robustness are considerably above the median of about 30% for observational studies recently published in Educational Evaluation and Policy Analysis (EEPA) (Frank et al., 2013).



Teacher turnover is a dominant factor behind school staffing problems in the United States (Ingersoll, 2001). Thus, understanding the patterns and causes of teacher turnover is of critical importance for both theory and policy regarding the teacher workforce or labor market. A vast majority of previous studies have attempted to investigate factors affecting teacher turnover; however, they are limited by neglecting the external contextual factors surrounding schools and their personnel. In this study analyzing statewide longitudinal administrative data from Michigan, we find that external contextual factors are, indeed, related to increased rates of teacher turnover. Our results add to a growing body of research that finds that state and local policies are related to changes in teacher turnover (e.g., Feng et al., 2010; Fulbeck, 2014). While the factors that we examined are unique to Michigan, implications of external contextual factors transcend our study and have consequences for all schools across the nation. Other states may not have the same experiences as Michigan but will have their own policies and economic pressures to consider in relation to school-level decisions and teacher turnover.

In the Michigan context, we find that both the introduction of a state-level educational reform, the MMC, and economic pressures from the recession are related to increases in school-level teacher turnover rates, independently of each other. Moreover, to understand and reduce teacher turnover we must fully understand it by looking beyond what is occurring within schools and examine schools situated within their larger context as the external contextual factors of educational reform and the recession are not related to turnover in the same way across school locales. There was a significant increase in school-level teacher turnover in city schools following the announcement of the MMC but not during the recession. However, schools in town, rural, and suburban areas had statistically significant increases in teacher turnover to varying degrees following the announcement of the MMC and the recession for multiple years.  

One possible explanation for the patterns that we see by school locale may be the local opportunity structures. For example, in areas where there are multiple schools within a short distance (such as would be the case in city locations), school leaders are likely to have more options and opportunities to redistribute staff to fit their needs and this may occur on a regular basis. A change in the economic conditions in city locations may not have been a new challenge related to staffing. It is likely these schools have been working with tight budgets for years.  

There may also be differences by locale in the need for such redistribution. We may see less of a need for the redistribution of teachers in certain locales, such as suburban schools, if the schools already had a curriculum similar to the MMC in place. That is, if a school already has a large proportion of their students completing an MMC-like curriculum then the school leaders would not need to make as many adjustments to the teaching staff. If a school does not have a large proportion of their students taking these college preparatory classes, the school leaders will need to supplement their teaching staff from other schools to ensure capacity to comply with the MMC.  

Another explanation for the differential effects of external contextual factors on teacher turnover is that the local economy impacts its local schools in different ways. Just as Michigan’s economy was not the nation’s economy we also find that the suburban economy is not the urban, rural, or town economy within Michigan. It may be that in town and rural areas, where economic downturns have been a long-term reality, the latest economic shock was the breaking point for schools; placing significant constraints on their funding. In order for those school leaders to stay within their budgets and deal with the loss of funding, they may have been forced to reduce their teaching staff. In locales, such as cities, school leaders are likely to be constantly working with tight budgets, and so the impact of the recession may be in areas other than staffing.


Not only does this study contribute to a fuller understanding of teacher turnover by noting the importance of external contextual factors in such research but also it makes important methodological contributions to the literature on teacher turnover. The present study is the first large-scale analysis to investigate teacher turnover across an entire state between and within school years. By exploiting detailed longitudinal administrative data, we construct an analysis file with semesters nested within years for each school, which allows for analyses that provide a more complete and refined picture of teacher turnover in schools. This study demonstrates that the magnitude of teacher turnover for mid-year and end of year can be very different. Our results suggest that efforts to examine the factors related to teacher turnover need to consider the timing of when teachers leave during school year.

Furthermore, this study improves upon the existing literature on the geography of teacher labor markets by examining teacher turnover by school locale. With multiple semesters/years of statewide data, we are able to show the patterns and consequences of teacher turnover for varying urbanicities (i.e., city, suburban, town, and rural) with little compromise to estimate precision. Our study finds important differences across locale that may be overlooked in studies which aggregate nationally or examine only particular local districts.


Our study makes important substantive and methodological contributions, but it is not without limitations. First, with a single state sample, our findings may not generalize to other states because they may not have the same experience as Michigan. Other states, however, will have their own external contextual factors to consider, thus providing opportunities for researchers to test the contextual hypothesis in other settings.

Some of our measures have limitations. We are able to calculate if a teacher is not retained in a school from one semester to the next, but we do not have good information on why teachers leave. While we are not conducting a teacher-level analysis, being able to differentiate between voluntary and involuntary leavers would be informative as these trends are conflated in our results. Additionally, our measure of MY and EOY is not as precise as we would like. The MY measure occurs early in the school year.

We are also limited in our ability to measure and closely examine the internal school mechanisms that may explain the relationship to teacher turnover. We speculate that teacher turnover can lead to vicious cycles including prompting some teachers to restrict their investments in other teachers and their students. This phenomenon could affect even teachers with high intentions and likelihoods of staying in the school. In turn, the restricted investments reduce the quality of the workplace, and thus reduce the probability that any given teacher will remain in the school (see Johnson et al., 2012). In some sense this is the inverse of a positive cycle in which actors increasingly contribute to a well-functioning collective action (Macy, 1991).

Why an individual teacher leaves has not been the focus of this analysis; rather, our main goal is to examine the cyclical nature of school-level teacher turnover rates across different locales while accounting for educational policy reforms and economic conditions. In some instances, the exiting of a teacher from a school is not a personal decision but one spurred by administrative actions beyond their control and those administrative actions are influenced by larger external forces. On the basis of our results when estimating probabilistic models of teacher turnover, we need to factor in these larger conditions and their impacts over time. Statewide longitudinal administrative data is valuable when estimating teacher turnover, especially for determining potential teacher shortages, revealing that there is no consistency in the precise factors at play and not all locales are affected equally. If teacher turnover research does not take into account variation by locale, we may have an inaccurate understanding of the extent of teacher turnover and the relating factors. Additionally, if policymakers do not take into account variation in the impacts of educational reforms by locale, they may have inaccurate expectations for how the policy will play out and the policy may result in unintended consequences for student learning.


1. See “Real Per-Capita Personal Income in Michigan and the U.S.:2000–2010” by the Michigan Department of Technology, Management, & Budget

2. When referring to the school year we use the second year listed to represent the year. For example, we use 2004 when discussing the 2003–2004 school year.

3. FTE is calculated by dividing the amount of time employed by the time normally required for a full-time position. For example, if a teacher is assigned to a chemistry class for three-fourths of his/her schedule and a biology class for one-fourth of his/her schedule, "0.75" FTE for chemistry is reported for the first position, and "0.25" FTE for biology in the second position. With these two assignments combined, the teacher has an FTE of “1.”

4. Often with prior studies of teacher turnover, it is unclear whether they make a distinction between part-time and full-time teachers. Therefore, their calculations of teacher turnover rates may be underestimated (if a teacher goes from full-time to part-time but still teaches in the school) or overestimated (as part-time teachers tend to move more frequently). Our study focuses on full-time teachers since the opportunity structure within the school, district, and/or locale may be different for full- and part-time teachers. An examination of the role of context on teacher turnover for part-time teachers is beyond the scope of this study.

5. We recognize that we are interpreting the year dummy variables aggressively. The changes that we see could be a function of demographic changes in the high school aged population; however, our examination of the census data did not suggest that there were large declines in the size of the high school student population.

6. Our locale classifications come from the Common Core of Data. The urban and suburban classifications are determined based on their proximity to an urbanized area, which is a population center with greater than 50,000 people. Towns are classified based on their relation to urban clusters, which have less than 50,000 people but greater than 25,000. Finally, rural locations are labeled based on their proximity to both urbanized areas and urban clusters (National Center for Education Statistics, n.d.).

7. Our original sample, which includes closed and new opened schools over the years, shows similar results— the number of FTE high school teachers in Michigan dropped from 22,641 in fall 2003–2004 to 18,655 in spring 2010–2011. We did run the analyses with 2004-2005 as our base-year in order to calculate change. The results are substantively the same.

8. Nationally there has been a slight increase in the number of full-time teachers (1.6%) from 2009–2013; however, this is not separated by high school, middle school, and elementary (author calculations from Keigher & Cross, 2010; Goldring et al., 2014).

9. Supplemental analyses (available upon request) show that compared with school districts with only one high school (n = 502) school districts with two or more high schools (n = 36) tend to have slightly higher teacher turnover rates each semester (from 3.2 to 16.8%) (note both semesters in year 2011 are an exception).  

10. We note that the retirement rates MY and EOY for the most part are similar. This is somewhat surprising given that we would expect most teachers to retire at the end of the year. However, this could be a function of teachers using accrued personal and sick days to retire earlier in the year.

11. In additional analyses using multilevel models, we find that indeed the magnitudes are larger in city schools.


This research result used data structured and maintained by the Michigan Consortium for Educational Research (MCER). MCER data is modified for analysis purposes using rules governed by MCER and are not identical to those data collected and maintained by the Michigan Department of Education (MDE) and/or Michigan’s Center for Educational Performance and Information (CEPI). Results, information and opinions solely represent the analysis, information and opinions of the author(s) and are not endorsed by, or reflect the views or positions of, grantors, MDE and CEPI or any employee thereof.


Allensworth, E., Poniscak, S., & Mazzeo, C. (2009). The schools teachers leave: Teacher mobility in Chicago Public Schools. Chicago: Consortium on Chicago School Research at the University of Chicago Urban Education Institute.

Bureau of Labor Statistics. (2012). The Recession of 2007–2009. Retrieved from http://www.bls.gov/spotlight/2012/recession/

Bureau of Labor Statistics. (2015). Frequently Asked Questions. Retrieved from http://www.bls.gov/lau/laufaq.htm#Q01.

Byrd, S., & Langer, K. (2010). Mandating change: Assessing the implementation of the Michigan merit curriculum. Retrieved from http://closup.umich.edu/files/closup-wp-29-michigan-merit-curriculum.pdf

Dynarski, S. M., Frank, K., Jacob, B. A., & Schneider, B. (2012a). Impacts of the Michigan Merit Curriculum on student outcomes: Preliminary findings from the first cohort. Retrieved from http://michiganconsortium.org/research/

Dynarski, S. M., Frank, K., Jacob, B. A., & Schneider, B. (2012b). The Michigan Merit Curriculum and teacher compositional change. Retrieved from http://michiganconsortium.org/research/

Feng, L., Figlio, D. N., & Sass, T. (2010). School accountability and teacher mobility. NBER Working Paper 16070.

Frank, K. A. (2009). Quasi-ties: Directing resources to members of a collective.
American Behavioral Scientist, 52(12), 1613–1645.

Frank, K. A., Maroulis, S., Belman, D., & Kaplowicohentz, M. D. (2011). The social embeddedness of natural resource extraction and use in small fishing communities. In W. W. Taylor, A. J. Lynch, & M. G. Schechter (Eds.), Sustainable fisheries: Multi-level approaches to a global problem (pp. 309–332). Bethesda, Maryland: American Fisheries Society. Retrieved from https://msu.edu/~kenfrank/social%20embeddedness%20of%20natural%20resource%20use%20in%20small%20fishing%20communities.pdf

Frank, K. A., Maroulis, S., Duong, M., & Kelcey, B. (2013). What would it take to change an inference? Using Rubin’s Causal Model to interpret the robustness of causal inferences. Education Evaluation and Policy Analysis, 35(4), 437–460

Fulbeck, E. S. (2014). Teacher mobility and financial incentives: A descriptive analysis of Denver’s ProComp. Educational Evaluation and Policy Analysis, 36(1), 67–82.

Goldhaber, D., Strunk, K. O., Brown, N., & Knight, D. S. (2016). Lessons learned from the Great Recession: Layoffs and RIF-induced teacher shuffle. Educational Evaluation and Policy Analysis, 38(3), 517–548.

Goldring, R., Taie, S., & Riddles, M. (2014). Teacher attrition and mobility: Results from the 2012–13 Teacher Follow-up Survey (NCES 2014-077). U.S. Department of Education. Washington, DC: National Center for Education Statistics.

Guin, K. (2004). Chronic teacher turnover in urban elementary schools. Education Policy Analysis Archives, 12(42).

Hanushek, E. A., Kain, J. F., & Rivkin, S. G. (2004). Why public schools lose teachers. Journal of Human Resources, 39(2), 326–354.

Holme, J. J. & Rangel, V. S. (2012). Putting school reform in its place: Social geography, organizational social capital, and school performance. American Educational Research Journal, 49(2), 257–283.

Ingersoll, R. (2001). Teacher turnover and teacher shortages: An organizational analysis. American Educational Research Journal, 38(3), 499–534.

Ingersoll, R., & May, H. (2012). The magnitude, destinations, and determinants of mathematics and science teacher turnover. Educational Evaluation and Policy Analysis, 34(4), 435–464.

Ingersoll, R., Merrill, L., & May, H. (2016). Do accountability policies push teachers out? Educational Leadership, 73(8), 44–49.

Jackson, C. K. (2013). Match quality, worker productivity, and worker mobility: Direct evidence from teachers. Review of Economics and Statistics, 95(4), 1096–1116.

Jacob, B., Dynarski, S. Frank, K., & Schneider, B. (2017). Are expectations alone enough? Estimating the effect of a mandatory college-prep curriculum in Michigan. Educational Evaluation and Policy Analysis, 39(2), 333–360.

Johnson, S. M. (2006). The workplace matters: Teacher quality, retention, and effectiveness. Washington, DC: National Education Association Research Department.

Johnson, S. M., Kraft, M. A., & Papay, J. P. (2012). How context matters in high-need schools: The effects of teachers’ working conditions on their professional satisfaction and their students achievement. Teachers College Record, 114(10), 1–39.

Keigher, A., & Cross, F. (2010). Teacher attrition and mobility: Results from the 2008-2009 Teacher Follow-Up Survey. Washington, DC: National Center for Education Statistics (NCES 2010-353).

Keesler, V. (2010). Teacher turnover and undersupply: A school organizational analysis.  Michigan State University. Unpublished dissertation manuscript.

Kennedy, M. M. (2005). Inside teaching: How classroom life undermines reform. Cambridge, MA: Harvard University Press.

Kraft, M. A., Marinell, W.H., & Yee, D. (2016).  Schools as organizations: Examining school climate, teacher turnover, and student achievement in NYC. New York: The Research Alliance for New York City Schools.

Kukla-Acedvedo, S. (2009). Leavers, movers, and stayers: The role of workplace conditions in teacher mobility decisions. The Journal of Educational Research, 102(6), 443–452.

Labor Market Information. (n.d.) Industry employment. Retrieved from http://milmi.org/cgi/dataanalysis/cesReport.asp?menuchoice=ces.

Ladd, H. F. (2011). Teachers’ perceptions of their working conditions: How predictive of planned and actual teacher movement? Educational Evaluation and Policy Analysis, 33(2), 235–261.

Leachman, M., & Mai, C. (2014). Most states still funding schools less than before the recession. Washington, DC: Center on Budget and Policy Priorities.

National Center for Education Statistics. (n.d.). Identification of rural locales. Retrieved from https://nces.ed.gov/ccd/rural_locales.asp

Macy, M. W. (1991). Chains of cooperation: Threshold effects in collective action. American Sociological Review, 56(6), 730-747.

Michigan Department of Technology, Management, and Budget. (n.d.). Per-Capita Personal Income by State: 2000-2010. Retrieved from http://www.michigan.gov/cgi/0,4548,7-158-54534-253826--,00.html

Obenauf, K.T. (2014). Curricular reform and economic context: The case of the Michigan Merit Curriculum. Michigan State University, East Lansing, MI.

Odden, A.R. (2012). Improving student learning when budgets are tight. Thousand Oaks, CA: Corwin.

Redding, C. & Henry, G.T. (2018). New evidence on the frequency of teacher turnover: Accounting for within-year turnover. Educational Researcher, 47(9), 577–593.

Ronfeldt, M., Loeb, S., & Wyckoff, J. (2013). How teacher turnover harms student achievement. American Educational Research Journal, 50(1), 4–36.

Singer, J. D., & Willett, J. B. (2003). Applied longitudinal data analysis. New York: Oxford University Press, Inc.

State Budget Office. (2013). Executive budget: Fiscal years 2012 and 2013. Lansing, MI: Department of Technology, Management, and Budget. Retrieved from http://www.michigan.gov/budget/0,4538,7-157-11460_18526---,00.html

Stuit, D., & Smith, T. M. (2010). Teacher turnover in charter schools. Nashville, Tennessee: National Center on School Choice.

Cite This Article as: Teachers College Record Volume 121 Number 11, 2019, p. 1-30
https://www.tcrecord.org ID Number: 22816, Date Accessed: 1/22/2022 9:49:39 PM

Purchase Reprint Rights for this article or review
Article Tools
Related Articles

Related Discussion
Post a Comment | Read All

About the Author
  • Elizabeth Minor
    National Louis University
    E-mail Author
    ELIZABETH COVAY MINOR is an associate professor at National Louis University in the Educational Leadership Program. Her research interests focus on educational leadership as well as inequality in student opportunities to learn and how social context is related to differences in access to, returns to, and experiences within student opportunities to learn. She recently published “Developing Effective Leaders Requires Valid, High Quality Psychometrically Sound, and Reliable Tools: A Test-Retest Analysis of The Vanderbilt Assessment for Leadership in Education” in Educational Assessment, Evaluation, and Accountability with co-authors Andrew C. Porter, Joseph Murphy, Ellen Goldring, and Stephen N. Elliott.
  • Guan Saw
    University of Texas at San Antonio
    E-mail Author
    GUAN K. SAW is an assistant professor at the Department of Educational Psychology at the University of Texas at San Antonio. His research focuses on educational inequality, STEM education, and college access and success. Recent publications include: Saw, G. K., Chang, C.-N., & Chan, H.-Y. (2018). Cross-sectional and longitudinal disparities in STEM career aspirations at the intersection of gender, race/ethnicity, and socioeconomic status. Educational Researcher; and Saw, G. K. (2018). Remedial enrollment during first year of college, institutional transfer, and degree attainment. Journal of Higher Education.
  • Kenneth Frank
    Michigan State University
    KENNETH FRANK received his Ph.D. in measurement, evaluation and statistical analysis from the School of Education at the University of Chicago in 1993. He is MSU Foundation professor of Sociometrics, professor in Counseling, Educational Psychology and Special Education; and adjunct (by courtesy) in Sociology at Michigan State University. His substantive interests include the study of schools as organizations, social structures of students and teachers and school decision-making, and social capital. His substantive areas are linked to several methodological interests: social network analysis, causal inference and multi-level models. His publications include quantitative methods for representing relations among actors in a social network, robustness indices for sensitivity analysis for causal inferences, and the effects of social capital in schools and other social contexts. Dr. Frank’s current projects include how beginning teachers’ networks affect their response to the Common Core, how schools respond to increases in core curricular requirements, cognitive linkages among aspects of knowledge, the diffusion of knowledge about climate change, and how the decisions about natural resource use in small communities are embedded in social contexts.
  • Barbara Schneider
    Michigan State University
    E-mail Author
    BARBARA SCHNEIDER uses a sociological lens to examine interactions and social structures and how they impact educational inequality. She is the John A. Hannah University Distinguished Professor in the College of Education and Department of Sociology at Michigan State University—Her most recent publications include Broda, M., Yun, J., Schneider, B., Yeager, D.S., Walton, G.M., & Diemer, M. (2018). Reducing Inequality in Academic Success for Incoming College Students: A Randomized Trial of Growth Mindset and Belonging Interventions. Journal of Research on Educational Effectiveness; and Schneider, B., Krajcik, J., Lavonen, J., Salmela-Aro, K. (Expect publication 2019) Learning Science: Crafting Engaging Science Environments. New Haven: Yale University Press.
  • Kaitlin Torphy
    Michigan State University
    E-mail Author
    KAITLIN T. TORPHY, Ph.D., is a Research Scientist and the developer and Director of the Teachers in Social Media Project. This project considers the intersection of cloud to class, nature of resources within virtual resource pools, and implications for equity as educational spaces grow increasingly connected. Kaitlin conceptualizes the emergence of a teacherpreneurial guild in which teachers turn to one another for instructional content and resources. She has expertise in teachers’ engagement across virtual platforms, teachers’ physical and virtual social networks, and education policy reform. Kaitlin has published work on charter school impacts, curricular reform, teachers’ social networks, and presented work regarding teachers’ engagement within social media at the national and international level. Kaitlin’s other work examines diffusion of sustainable practices across social networks within The Nature Conservancy. Kaitlin holds a Ph.D. in education policy, a specialization in the economics of education, and is a Teach for America alumni and former Chicago Public Schools teacher.
Member Center
In Print
This Month's Issue