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

Examining English Language Arts Teachers: Evidence from National Data


by Tuan D. Nguyen & Laura Northrop - 2021

Background: Much of the previous research in teacher attrition and retention focuses on teachers in general, without regard to specific types of teachers. We focus on English language arts (ELA) and English as a second language (ESL) teachers because the United States has stubbornly low achievement in reading, and reading is critical to the success of learning other subjects, and because these are the two groups of teachers most responsible for teaching both native and non-native English speakers to read.

Purpose: Our goal is to understand how the demographics and qualifications of ELA and ESL teachers have changed over time, changes in the student characteristics and school conditions in which they teach, their attrition rate, and the factors that are associated with their attrition behaviors. We also pay close attention to teachers in economically disadvantaged schools.

Research Design: We use nationally representative data from seven waves of the Schools and Staffing Survey from 19871988 to 20112012, as well as the 20152016 National Teacher and Principal Survey to examine ELA and ESL teachers and their turnover behaviors. We employ sampling weights to make the results nationally representative. We use both descriptive and regression analyses to examine these teachers.

Data Analysis: We first describe how characteristics of ELA and ESL teachers and the schools in which they teach have changed from 1988 to 2016. We then examine how these characteristics vary systematically across high- and low-poverty schools and compare the attrition rate for ELA and ESL teachers relative to other teachers. We also examine the factors that are associated with various forms of turnover and how organizational supports may be leveraged to increase retention.

Findings: We find the composition of ELA and ESL teachers has changed substantially over time, with more teachers attending selective schools, being certified to teach, and more likely to teach in high-poverty schools. Relatedly, teacher characteristics in high- and low-poverty schools are consistently different across time, and school working conditions play an enhanced role in high-poverty schools rather than in more affluent schools.

Conclusions: Although we find ELA teachers turn over at similar rates compared to non-ELA teachers, we find ESL teachers are more likely to leave both their current school and the profession. This is particularly concerning given that the ESL population is increasing while at the same time there is a shortage of trained-ESL teachers.

Teacher turnover and attrition are persistent concerns of policy makers, district leaders, and school principals. Although some teacher turnover can be positive, such as when less effective teachers choose to leave the profession (Boyd, Grossman, et al., 2008; Goldhaber et al., 2011; Nguyen et al., 2020), most teacher turnover is disruptive and costly for schools, in terms of both monetary costs and hidden costs, such as quality of staff and lower student achievement (Sorensen & Ladd, 2020).


Much of the previous research in teacher attrition and retention focuses on teachers in general, without regard to specific types of teachers. However, it is important to examine differences in attrition by subject level, because different supply and demand forces can affect the teacher labor market by subgroups. On the supply side, teacher shortages in particular subject areas, such as math, science, special education, and English language, are more acute than in other subjects (Sutcher et al., 2016). On the demand side, teachers with specific skills and subject-area knowledge may have more nonteaching job opportunities in the labor market.


In this paper, we focus on English language arts (ELA) and English as a second language (ESL) teachers, a subgroup of ELA teachers who work with non-native English students. We focus on these two groups of teachers because the United States has, for decades, had stubbornly low achievement in reading, and reading is critical to the success of learning other subjects, and because these are the two groups of teachers most responsible for teaching both native and non-native English speakers to read (Allington & McGill-Franzen, 2018; Carver, 2000). Snapshot data from the 2019 National Assessment of Educational Progress shows low reading achievement for all students, as well as that English learners (ELs) lag their non-EL peers in both reading and math (National Center for Education Statistics, 2019). Because research supports the idea that teachers influence academic achievement for students through their training and abilities, such as college ratings, test scores, and math coursework (Boyd, Lankford, et al., 2008; Wayne & Youngs, 2003); instructional practices (Grossman et al., 2013); and experience (Boyd, Lankford, et al., 2008; Henry et al., 2012), as well as race-matching (Redding, 2019), it is important to examine the characteristics and turnover for the group of teachers primarily responsible for reading instruction in the United States.


We include both ELA and ESL teachers as our focus to fully capture teacher turnover in students’ English learning experiences, whether they are a native or non-native speaker. EL students are a particularly important subgroup of students, because from 2000 to 2016, the percentage of public-school students who identified as EL increased from 8.1% to 9.6%, an increase of approximately 1 million students, with EL students consisting of more than 10% of public school students in more than nine states (McFarland et al., 2019). Moreover, because school districts in the United States have flexibility in how to serve EL students and can choose from a variety of models such as bilingual, immersion, pull-out classes, or inclusion classes,1 EL students are being served by a variety of different English teachers. Lastly, school districts are having difficulty finding enough qualified ESL teachers, with more than 30 states identifying a shortage of teachers in this area (Sutcher et al., 2016), making it even more important to examine both ELA and ESL teachers to fully capture student experiences in school.


In this paper, we use nationally representative data from the Schools and Staffing Survey (SASS) and its new iteration, the National Teacher and Principal Survey (NTPS), to investigate ELA and ESL teacher characteristics and attrition between 1987 and 2016 and specifically ask:


1.

What are the changes in ELA and ESL teacher demographics and qualifications during this time period?


2.

What are the changes in student characteristics and the school conditions in which ELA and ESL teachers work during this time period?


3.

What is the attrition rate for ELA and ESL teachers, and how does it compare to their counterparts’ attrition rates?


4.

What teacher characteristics and school conditions are predictive of ELA and ESL teacher attrition?


BACKGROUND


Rates and Impact of Attrition


Estimates of teacher attrition vary, depending on whether they measure attrition from a school or attrition from the profession, but range from approximately 13% to 18% for one year (Papay et al., 2017; Redding & Henry, 2018), with 3- and 5-year attrition rates ranging from 14% to as high as 45% (Gray & Taie, 2015; Ingersoll, 2001; Kelly & Northrop, 2015). Additionally, approximately 4–6% of teachers turn over during the school year (Redding & Henry, 2018), which can be even more disruptive than turnover that occurs during the summer months. In urban districts, rates of attrition can be even higher, although there is large variation in teacher attrition rates depending on the district (Papay et al., 2017).


Persistent and high rates of teacher turnover can adversely impact students, because teacher turnover is disruptive to schools and students, and has negative effects on learning, particularly for the most vulnerable students. Studying fourth- and fifth-grade students in New York City, Ronfeldt, Loeb, and Wyckoff (2013) found that teacher turnover adversely impacted both reading and math scores, and was more harmful to minority and low-income students. Likewise, a more recent study using data from an urban district in Texas also found that teacher turnover negatively affects quality instruction more in lower-achieving schools than in high-achieving schools, primarily due to the loss of general and grade-specific experience (Hanushek et al., 2016).


One positive effect of teacher turnover is that more effective teachers, as measured by value-added models that capture student test score gains, are less likely to leave teaching, both in the middle of the year and at the end of the year (Boyd, Lankford, et al., 2008; Goldhaber et al., 2011; Nguyen et al., 2020). More effective teachers are also less likely to leave urban districts (Papay et al., 2017), and schools with a high percentage of minority students and students in poverty are more likely to face teacher shortages (Suchter et al., 2106). In addition to district attrition, more effective teachers are also less likely to request transfers to new schools within the same urban district (Boyd et al., 2010).


Reasons for Attrition


There is a robust line of research documenting factors associated with teacher turnover, including teacher characteristics and school characteristics (Nguyen et al., 2020). Teacher characteristics include factors such as preparation program, type of licensure, and subject taught. School characteristics include the type of school, the demographic characteristics of the students who attend the school, and the working conditions at the school.


One teacher characteristic that influences attrition is the type of preparation program teachers completed prior to workforce entry. Teachers who are prepared in traditional programs in the same state in which they teach are least likely to leave (Bastian & Henry, 2015; Redding & Henry, 2018). Teachers prepared via alternative pathways, including Teach for America, are more likely to leave both within the school year and at the end of the school year, although Teach for America teachers are most likely to leave at the end of their two-year commitment (Redding & Henry, 2018; Redding & Smith, 2016). Teachers from highly selective universities have an 85% greater likelihood of leaving the profession, due to lower career satisfaction (Kelly & Northrop, 2015).


Another teacher characteristic that influences attrition is the subject teachers specialize in. Although Ingersoll (2001) found that math and science teachers were not more likely to turn over, other studies have found that teachers who hold math or science undergraduate degrees are more likely to leave (Borman & Dowling, 2008; Nguyen et al., 2020; Nguyen & Redding, 2018). In addition, special education teachers are more likely to turn over (Ingersoll, 2001). To our knowledge, there is very limited research examining the attrition rates of ELA teachers. Using the public version of the 1999–2000 SASS, Hahs-Vaughn and Scherff (2008) found some evidence that salary is related to attrition for beginning traditional English teachers. However, because they used only the publicly available version and for only one wave, they were not able to leverage the rich teacher and school characteristics that we are able to with the restricted version, and across a large timespan. Moreover, they focused on only beginning traditional English teachers, not all ELA teachers. As a result, we know very little about which teacher characteristics may be related to ELA teachers’ attrition.


School characteristics also influence teacher turnover. Teachers are more likely to leave from middle schools and high schools (Nguyen et al., 2020). In one study that examined data from New York City, middle schools received both the most requests from teachers to transfer out and the least amount of applications to transfer in (Boyd et al., 2011). Teacher attrition rates are also higher in schools with more low-income, low-achieving, and minority students (Borman & Dowling, 2008; Guarino et al., 2006; Hanushek et al., 2004), including both middle of year and end of year departures (Redding & Henry, 2018). Likewise, teachers request more transfers out of schools that have higher populations of Hispanic, Black, low-income, and low-achieving students (Boyd et al., 2010). In addition to more teacher churn, these students also typically have less-skilled teachers, including higher percentages of teachers with no prior teaching experience, with no certification, who have failed a certification exam, or who have a college degree from a less prestigious college (Lankford et al., 2002).


School characteristics also include the working conditions teachers face within each school, with teachers more likely to leave schools with poor behavioral climate (Guarino et al., 2006; Ingersoll, 2001; Kelly, 2004). In particular, Ingersoll (2001) found that inadequate administrative support and student discipline problems were main reasons cited by teachers who left their school or the profession. Likewise, using data from New York City, Boyd et al. (2011) found that administrative support, which included supportive and encouraging behavior towards teachers, a well-planned and enforced discipline policy, and teacher evaluation, significantly predicted teacher retention. More recently, research from New York City middle schools also found that school leadership and school safety were associated with reductions in teacher turnover (Kraft et al., 2016).


Current Study


Previous studies that have included attrition rates by subject taught have focused primarily on science and math teachers (e.g., Kelly & Northrop, 2015; Nguyen & Redding, 2018). To the best of our knowledge, there have been very few studies that examine ELA teachers at the national level (e.g., Hahs-Vaughn & Scherff, 2010; Hancock & Scherff, 2010), and what has been examined largely relies on short-term administrative data or data from a single district or state (Borman & Dowling, 2008). For instance, using a single wave of 2003–2004 SASS, Hancock and Scherff (2010) found that being a minority teacher and years of teaching experience are associated with teacher turnover, but it is unknown whether their findings are specific to that year and whether their findings are robust to different model specifications or a richer set of covariates. As a result, we know little about the factors associated with teacher turnover for ELA teachers (Nguyen et al., 2020).


In thinking about attrition rates for ELA teachers, one hypothesis is that they will have similar attrition rates to other types of teachers. ELA, in contrast to the special education, math, and science fields, does not have a shortage of teachers (Sutcher et al., 2016). Likewise, non-STEM teachers are hired into K–12 jobs at a lower rate than STEM teachers (Goldhaber et al., 2021). Lastly, teacher candidates employed as K–12 teachers earn more than those employed outside of education, with the exception of STEM teacher candidates (Goldhaber et al., 2021). Taken together, these points suggest that ELA teachers may be more reluctant to leave a teaching position once obtained, as well as face less lucrative options outside of the teaching labor market.  


However, when looking only at the subset of ESL teachers, a different hypothesis emerges. In contrast to ELA teachers, there is a current shortage of ESL teachers (Sutcher et al., 2016). In addition, ESL teachers can face an increased emotional labor (Hochschild, 2018) due to student concerns unique to the language-minority student population. First, language-minority students who are refugees or immigrants may face discrimination and hate speech (Duran, 2019) or trauma (Schmidt, 2018), with ESL teachers being called to work toward dismantling this xenophobia and support students learning through trauma-informed teaching. Similarly, recent changes in regulations about immigration and refugee status in 2017, which limit immigration and refugee resettlement, create fear for non-native students and a need for teachers to work with students to address those fears (Petrie & Darrag, 2019). Most recently, fears stemming from the COVID-19 pandemic have led to an increase in racism and hatred directed toward Chinese and other Asian students (Liz, 2020; Roberto et al., 2020).


In addressing this gap, our study makes several contributions. First, we examine the changes in ELA teacher characteristics and qualifications over a 28-year period, which corresponds to an increase in EL students in U.S. schools. Second, we examine the changes in student characteristics and the school conditions in which ELA teachers work during this time period. Third, we examine the attrition rates for ELA teachers and in comparison to non-ELA teachers over a substantial period of time. Fourth, we examine the associations of teacher characteristics and school conditions with various forms of teacher turnover. Fifth, we consider how these relationships vary for low-income and more affluent schools. Lastly, we examine ESL teachers as a specific subset of ELA teachers. In short, we make substantial contributions to the sparse literature on ELA and ESL teachers and their mobility patterns.


DATA AND METHOD


We use data from the Schools and Staffing Survey (SASS) and its supplement, the Teacher Follow-Up Survey (TFS), as well as the National Teacher and Principal Survey (NTPS), the new iteration of SASS. SASS and NTPS, administered by the National Center for Educational Statistics (NCES), include more than 30,000 public school teachers every wave and consist of nationally representative samples of schools, principals, and teachers for public schools. These surveys include comprehensive data on teacher characteristics, organizational supports, and school characteristics, making them a rich data source for describing differences among ELA and ESL teachers and the factors associated with turnover. The average teacher response rate is about 85% with a low missing data rate (less than 3%). For this study, we use all seven iterations of SASS and the last four iterations of TFS where teacher turnover data and key covariates for all teachers in the SASS base years can be generated along with the 2015–2016 NTPS. More specifically, we use the 1987–1988, 1991–1992, 1993–1994, 1999–2000, 2003–2004, 2007–2008, and 2011–2012 SASS waves and the 2015–2016 NTPS for the descriptive analyses, and the 2000–2001, 2004–2005, 2008–2009, and 2012–2013 TFS waves for the turnover analyses. We note the NTPS 2015–2016 survey wave does not include turnover data. We employ sampling weights to make the results nationally representative. The overall sample size is more than 100,000 observations, and the turnover sample size is 17,550 ELA teacher observations with 1,330 ESL teachers.


MEASURES OF ATTRITION


A complete description of the variables used in this study is provided in Appendix Table 1. In the next two sections, we provide an overview of these variables. The main dependent variable for this study comes from the principal report of teachers’ employment status in the follow-up year following the baseline survey year. We categorize teacher status into one of three categories: stayers, switchers, or leavers. Stayers are teachers who remained in the same school in the baseline year; switchers are teachers who switched to a new school but remained in the teaching workforce; and leavers are teachers who left the teaching profession.


DEFINING ELA AND ESL TEACHERS


We define ELA teachers as teachers whose main teaching assignments are English, communications, composition, journalism, language arts, reading, and speech. We include English as a second language (ESL) teachers if their main teaching assignment is included above. In secondary analysis, we replicate our main analyses specifically for ESL teachers. (Results can be found in the Appendix Tables 3–7.)


MEASURES OF TEACHER CHARACTERISTICS AND SCHOOL CHARACTERISTICS


We include a number of teacher characteristics such as gender, race/ethnicity, graduate degree(s), certification status, undergraduate college selectivity using Barron’s Admissions Competitiveness Index, annual salary, and union membership. In particular, we include teacher demographics because prior works have shown that teacher–student race-matching can positively influence students’ outcomes, particularly those traditionally underserved (Redding, 2019). College selectivity has been used in previous studies as a proxy for academic ability or to indicate high quality teaching candidates (Cohen-Vogel & Smith, 2007; Nguyen, 2021; Nguyen & Redding, 2018). However, the Barron index is a rough measure of college quality because many colleges have been reclassified as top-tiered or increased their rankings without evidence of changes in instructional quality (Hess & Hochleitner, 2012). In terms of school characteristics, we consider several important characteristics: the school’s urbanicity, enrollment size, secondary or elementary level, percentage of students with free-and-reduced-price lunch (FRPL) eligibility, percentage of minority students, and percentage of limited English proficiency (LEP) students. In terms of school working conditions, we include levels of student disciplinary problems, teacher-reported support from the administrators, teacher cooperation, and measures of professional development provided to the teachers.  


We separate our results between teachers working in high- and low-FRPL schools, because educational resources provided to students in low-income schools are usually lower than in more affluent schools (Goldhaber et al., 2015). We operationalize this school-level measure by considering a school with 50% or more of students eligible for FRPL a low-income school (Ingersoll, 2001; Nguyen & Redding, 2018). To check the robustness of our results based on how we operationalized low-income schools, we also compare the bottom and top quartile for percent FRPL (Appendix Table 2). The results using the bottom and top quartile are substantively similar to the main estimates. Our preferred strategy is to use the majority indicator approach because it provides an absolute level of comparison and it is easier to interpret across waves compared to the lowest/highest comparison that changes from wave to wave. Moreover, the majority indicator allows us to use the full sample of ELA teachers relative to the smaller and restricted sample size when using only the first and fourth quartile of percent FRPL.


ANALYSIS


Our analysis consists of two main parts: descriptive and regression analyses. In the descriptive analysis, we report on changes in ELA teacher characteristics and the conditions in which they teach. With regression analysis, we examine the extent to which ELA teachers are predicted to turn over relative to non-ELA teachers and how this varies for high- and low-income schools. We first estimate a logistic regression model comparing stayers versus movers (switchers and leavers). To examine the different forms of turnover, we employ multinomial logistic regression models comparing stayers versus switchers and stayers versus leavers, respectively, accounting for teacher and school characteristics. This multinomial logistic regression model can be expressed as:

[39_23856.htm_g/00002.jpg]



[39_23856.htm_g/00004.jpg]



where the probability of switching schools (k = 1) or leaving teaching (k = 2) for teacher i from school j in year t is a function of being an ELA teacher (ELATeacheri.), a vector of teacher background characteristics (Ti), a vector of school characteristics (Sj), a state fixed effect (ys), and a year fixed effect (lt). We examine the extent to which ELA teachers are more likely to turn over from low-income schools by interacting the ELA teacher variable with the low-income school indicator. Models are estimated using robust standard errors clustered at the state level while employing appropriate weights to make our results nationally representative. To examine the extent to which qualifications or school characteristics predict lower turnover rates for ELA teachers, we limit the sample to only ELA teachers and we also separate the models based on low-income school status. We note that, when we replicate this analysis for ESL teachers, we do not include state fixed effects due to reduced sample size.


STUDY LIMITATIONS


Although we are able to address many empirical gaps that exist for ELA and ESL teachers, we recognize that there are limitations to our study. First, although attrition is linked to student performance, prior works have shown that standardized assessment may lack validity for EL students (Kieffer & Thompson, 2018). We argue, however, that the existing evidence suggests turnover may have negative effects on student learning, not just on standardized assessment, so ELA and ESL teacher turnover would have deleterious effects on EL learning. Second, due to data limitation, although we do have more current data on ELA and ESL teacher characteristics, we do not have measures of turnover past the 2011–2012 school year. Currently it is unknown whether the NTPS will have turnover measures for future waves due to poor response on the recent Follow-Up Surveys; if it does not, future studies must use state-specific data to examine how turnover may have changed in recent times, particularly in part due to the COVID-19 pandemic.


RESULTS


ELA TEACHER CHARACTERISTICS AND SCHOOL CHARACTERISTICS OVER TIME


We begin by providing background information on the teacher characteristics and school conditions in which ELA teachers work. The results of our descriptive analysis examining ELA teacher and school characteristics are in Table 1. There were some substantial changes in the demographic characteristics of ELA teachers between 1988 and 2016. In particular, 73% of ELA teachers were female in 1988, but by 2016, 84% were female, a 11 percentage point increase from 1988. During this time period, the percentage of White teachers steadily decreased from 91% to 84%, whereas the percentage of Hispanic teachers steadily increased from 2% to 7%. Moreover, the percentage of novice teachers, those who had less than three years of teaching experience, increased from 6% in 1988 to 12% in 2000, but decreased back to 9% by 2016.


Table 1. Descriptive Statistics


 

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

 

Wave: 1988

Wave: 1991

Wave: 1994

Wave: 2000

Wave: 2004

Wave: 2008

Wave: 2012

Wave: 2016

Teacher Characteristics

 

Female

0.73

0.74

0.75

0.80

0.83

0.83

0.85

0.84

Black

0.06

0.06

0.05

0.08

0.08

0.09

0.07

0.08

Asian

0.01

0.01

0.01

0.01

0.01

0.01

0.01

0.02

American Indian

0.01

0.01

0.00

0.01

0.01

0.01

0.01

0.01

Hispanic

0.02

0.03

0.04

0.05

0.03

0.06

0.07

0.07

White

0.91

0.90

0.90

0.85

0.88

0.84

0.85

0.84

Teacher age

42.07

43.77

43.91

42.55

42.53

43.07

43.13

42.82

Novice teacher

0.06

0.07

0.08

0.12

0.11

0.09

0.07

0.09

Graduate degree

0.52

0.51

0.50

0.49

0.50

0.54

0.60

0.57

No certification

0.03

0.02

0.04

0.06

0.01

0.00

0.00

0.01

Most selective college

0.04

0.00

0.06

0.07

0.11

0.11

0.12

0.13

Very selective college

0.10

0.00

0.17

0.16

0.19

0.19

0.22

0.22

Salary per $1,000

54.13

55.74

55.86

54.58

54.76

54.03

55.54

53.41

Union member

.

.

0.78

0.80

0.77

0.78

0.76

0.71

School Characteristics

 

Urban school

0.26

0.25

0.25

0.26

0.27

0.25

0.24

0.26

Percent FRPL

0.21

0.23

0.25

0.30

0.40

0.42

0.46

0.51

Majority FRPL school

0.10

0.11

0.14

0.23

0.33

0.36

0.45

0.49

Percent minority students

0.25

0.27

0.28

0.32

0.36

0.42

0.41

0.47

Majority minority school

0.19

0.21

0.23

0.26

0.31

0.36

0.37

0.42

Percent LEP

.

.

.

0.04

0.05

0.07

0.06

0.05

Student discipline (std)

0.34

0.24

0.16

0.01

0.02

0.01

0.04

0.07

Admin. support (std)

.

.

–0.09

–0.10

–0.08

–0.04

–0.08

–0.10

Teacher cooperation (std)

.

.

–0.09

–0.14

–0.05

0.00

–0.03

–0.10

Observations

2,590

3,770

4,120

4,070

4,660

4,210

4,620

2,970



With respect to teacher education and qualifications, in 1988, 52% of ELA teachers had graduate degrees compared to 57% in 2016. Moreover, the percentage of ELA teachers without certification increased from 3% in 1988 to 6% in 2000, but this dropped to near 0% from 2004 to 2016. As dramatically as this decrease in the lack of certification is the change in undergraduate college selectivity. Although only 14% of ELA teachers had attended the most selective or very selective colleges in 1988, 35% of ELA teachers had attended such a college in 2016. In terms of salary, the average ELA teacher salary, in constant 2016 dollars, had fluctuated from $54,130 in 1988 to $55,540 in 2012, dropping to $53,410 in 2016. The substantial drop in average salary in 2008 was likely related to the Great Recession of 2007–2008. Lastly, from 1988–2012, about 77% of ELA teachers belonged to a teacher union, but this dropped to 71% in 2016.


In terms of school characteristics, about a quarter of the schools were located in urban areas. We observed substantial shifts in the student body characteristics. In 1988, about 10% of the schools were majority low-income with about 21% of students eligible for FRPL lunch. By 2016, half of schools were classified as majority FRPL with 51% of students FRPL-eligible. In other words, ELA teachers are increasingly likely to teach in schools where half of the students are low-income. With respect to LEP students, the percentage of students with LEP is about 5%.


Next we consider the levels of student disciplinary problems, administrative support, and teacher cooperation reported by ELA teachers. These measures are standardized by wave for all teachers, not just ELA teachers, so the overall mean is zero with a standard deviation of one. For ELA teachers, in comparison to all teachers, there were more reports of student disciplinary problems, particularly from 1988 to 1994, although this dropped substantially by the year 2000 to be barely above the mean of zero. In terms of administrative support and teacher cooperation, ELA teachers consistently reported receiving less administrative support and experiencing lower levels of teacher cooperation relative to other teachers nationally. With evidence that there have been substantial changes in the teacher characteristics and school conditions for ELA teachers, next we examine whether these characteristics are systematically different for low- and high-FRPL schools and whether they have changed over time in Table 2.


Table 2. Descriptive Statistics of ELA Teachers by High- and Low-FRPL School Status


 

Wave: 2000

  

Wave: 2012

Wave: 2016

 

Low-FRPL

High-FRPL

Diff

  

Low-FRPL

High-FRPL

Diff

Low-FRPL

High-FRPL

Diff

Teacher Characteristics

Female

0.79

0.82

0.03

  

0.84

0.85

0.02

0.83

0.86

0.03+

Black

0.05

0.18

0.13**

  

0.03

0.11

0.08**

0.05

0.11

0.07**

Asian

0.01

0.01

0.00

  

0.01

0.01

0.00

0.02

0.02

0.00

American Indian

0.01

0.01

0.01

  

0.01

0.02

0.01*

0.01

0.01

0.00

Hispanic

0.03

0.11

0.08**

  

0.03

0.12

0.10*

0.04

0.09

0.05**

White

0.90

0.69

–0.22**

  

0.93

0.75

–0.18**

0.90

0.77

–0.12**

Teacher age

42.38

43.13

0.75

  

42.90

43.41

0.51

42.71

42.93

0.21

Novice teacher

0.12

0.11

–0.01

  

0.06

0.08

0.02+

0.08

0.10

0.02*

Graduate degree

0.49

0.47

–0.02

  

0.64

0.54

–0.10**

0.61

0.53

–0.08**

No certification

0.06

0.09

0.03*

  

0.01

0.00

0.00

0.01

0.01

0.00

Most sel. college

0.08

0.02

–0.06**

  

0.13

0.10

–0.03*

0.16

0.10

–0.06**

Very sel. college

0.16

0.14

-0.03

  

0.25

0.18

–0.07**

0.24

0.20

–0.05*

Salary per $1,000

55.07

52.88

–2.19*

  

57.09

53.63

–3.46**

55.65

51.05

–4.6**

Union member

0.81

0.78

–0.03

  

0.79

0.72

–0.07**

0.73

0.68

–0.05*

School Characteristics

Urban school

0.20

0.46

0.26**

  

0.17

0.32

0.16**

0.20

0.32

0.12**

Percent FRPL

0.19

0.70

0.51**

  

0.24

0.72

0.48**

0.25

0.79

0.54**

Percent minority

0.24

0.62

0.38**

  

0.28

0.57

0.29**

0.34

0.59

0.25**

Major. minority

0.15

0.66

0.51**

  

0.18

0.60

0.42**

0.24

0.61

0.37**

Percent LEP

0.02

0.09

0.07**

  

0.03

0.10

0.07**

0.03

0.07

0.04**

Student discip.

0.01

0.02

0.01

  

–0.05

0.15

0.21**

–0.02

0.16

0.18**

Admin. support

–0.11

–0.08

0.02

  

–0.06

–0.10

–0.04

–0.10

–0.09

0.02

Teacher coop.

–0.13

–0.15

–0.02

  

0.02

–0.08

–0.10**

–0.05

–0.14

–0.09*

Observations

4,070

  

4,620

 

2,970

 


Note. Nationally representative weights are employed. Sample sizes are weighted to the nearest 10 in accordance with NCES nondisclosure rule. Salary has been adjusted to constant 2012 dollar.

+ p < 0.10

* p < 0.05

** p < 0.01

Source: U.S. Department of Education, National Center for Education Statistics, Schools and Staffing Survey (SASS) and National Teacher and Principal Survey (NTPS)


ELA TEACHERS AND HIGH-POVERTY SCHOOLS


Table 2 shows several notable patterns of differences between ELA teachers working in low- and high-FRPL schools, and these patterns have persisted through time. As such, we only show the first and last wave of data with turnover data as well as the recent NTPS data. First, we observe there tend to be more Black and Hispanic ELA teachers working in high-FRPL schools than low-FRPL schools and on the flip side, there are fewer White ELA teachers teaching in high-FRPL schools. For instance, in 2012, 3% of ELA teachers in low-FRPL schools were Black relative to 11% at high-FRPL schools, a significant eight percentage point difference. Similarly, ELA teachers in low-FRPL schools are more likely to come from the most selective or very selective colleges relative to their counterparts in high-FRPL schools. Along this line, ELA teachers in low-FRPL schools also received higher salary than those in high-FRPL schools, an income increase of around $2,100 in 2000, around $3,300 in 2012, and around $4,600 in 2016. With respect to school characteristics, ELA teachers in high-FRPL schools tended to be located in more urban areas, teaching a substantially higher percentage of minority students or, relatedly, in majority minority schools, with higher incidence of students with LEP than in low-FRPL schools. These relationships remained stable from 2000 to 2016. In terms of patterns that have changed, we find that a higher percentage of ELA teachers had graduate degrees in low-FRPL schools than in high-FRPL schools in 2012 and 2016, but not in 2000. Relative to the 1999–2000 wave, ELA teachers in high-FRPL schools reported more issues with student disciplinary problems and less teacher cooperation than those in low-FRPL schools. In short, this analysis reveals that ELA teachers in high-FRPL schools tend to be more diverse, are less likely to have graduate degrees, come from less selective colleges, receive less pay, work in more urban areas, teach more minority students, have more students with LEP, and report more student disciplinary problems and less teacher cooperation than their counterparts in low-FRPL schools. Due to these systematic differences for ELA teachers in low- and high-FRPL schools, we would expect the attrition rates would differ between these two groups, which we examine next.


ELA TEACHER ATTRITION RATES


In Table 3, we examine the attrition rates for non-ELA teachers and ELA teachers in Panels A and B, respectively, by wave and pooled together. Pooling across four waves, we observe the attrition rates are about the same. In general, we observe 6.7% of ELA teachers switch from one school to another whereas 7.6% leave the teaching profession altogether, relative to 7.3% of non-ELA teachers doing those two things. In short, we do not see a substantial difference between ELA and non-ELA teachers. What we are more concerned about, given the descriptive results earlier, is that ELA teachers may turn over at higher rates in high-FRPL schools than in low-FRPL schools, which we examine next.


Table 3. Rate of Attrition


 

(1)

(2)

(3)

(4)

(5)

 

Wave: 2000

Wave: 2004

Wave: 2008

Wave: 2012

Pooled

Panel A: Attrition Rate for non-ELA Teachers

Stayer

84.97

84.31

86.27

86.26

85.45

Switcher

7.37

7.94

7.12

6.61

7.26

Leaver

7.66

7.76

6.61

7.13

7.28

Observations

32,710

33,590

28,180

27,150

121,620

Panel B: Attrition Rate for ELA Teachers

Stayer

85.99

83.61

86.60

86.66

85.75

Switcher

6.93

7.90

6.24

5.84

6.68

Leaver

7.08

8.49

7.16

7.50

7.57

Observations

4,070

4,660

4,210

4,620

17,550

Note. Nationally representative weights are employed. Sample sizes are weighted to the nearest 10 in accordance with NCES nondisclosure rule. NTPS 2015–2016 does not contain attrition data.

Source: U.S. Department of Education, National Center for Education Statistics, Schools and Staffing Survey (SASS) and National Teacher and Principal Survey (NTPS)


In Table 4, we examine whether ELA teachers and teachers in high-FRPL schools turn over at higher rates and whether ELA teachers are turning over at higher rates in high-FRPL schools while accounting for a host of teacher and school characteristics. In Panel A where we examine overall turnover, we find that, as basic descriptive results show in Table 3, even controlling for many factors, we do not find ELA teachers having a higher rate of attrition. We also do not find that teachers in high-FRPL schools turn over at higher rates. Contrary to expectations, we do not find that ELA teachers in high-FRPL schools turn over at higher rates than ELA teachers in low-FRPL schools. These results are consistent in all forms of turnover (Panels B and C). In short, our results show no evidence that ELA teachers versus other kinds of teachers, and ELA teachers at high-FRPL schools versus ELA teachers at low-FRPL schools, are more likely to turn over. Although there may not be any differential attrition between high- and low-FRPL schools, the factors that are associated with turnover may have differential relationships in high- and low-FRPL schools.




Table 4. Logistic and Multinomial Logistic Regression of Teacher Turnover by FRPL School Status


 

(1)

(2)

 

High-FRPL

High-FRPL with ELA Interaction

Panel A: Overall Turnover from Logistic Regression

ELA teacher

1.00

0.99

 

(–0.04)

(–0.12)

High-poverty indicator

0.98

0.98

 

(–0.57)

(–0.63)

ELA teacher*

 

1.01

High-poverty indicator

 

(0.13)

Panel B: Switchers from Multinomial Logistic Regression

ELA teacher

0.95

0.99

 

(–1.02)

(–0.20)

High-poverty indicator

0.98

0.99

 

(–0.25)

(–0.12)

ELA teacher*

 

0.92

High-poverty indicator

 

(–0.79)

Panel C: Leavers from Multinomial Logistic Regression

ELA teacher

1.05

1.00

 

(0.88)

(0.03)

High-poverty indicator

0.97

0.96

 

(–0.70)

(–0.83)

ELA teacher*

 

1.12

High-poverty indicator

 

(0.89)

Observations

102,400

102,400


Note. Nationally representative weights are employed. Sample sizes weighted to the nearest 10 in accordance with NCES nondisclosure rule. Z-statistics from heteroskedastic-robust state-level clustered standard errors are in parentheses. The majority indicator corresponds to the model title. The interaction term is the interaction between ELA teacher and the majority indicator. All models control for teacher race/ethnicity, gender, age, teaching experience, undergraduate selectivity, graduate degree, certification, passing content exam, qualification, salary, urbanicity, school enrollment, school-level characteristics, administrative support, and cooperation among the staff, along with state and year fixed effects.

+ p < 0.10

* p < 0.05

** p < 0.01

Source: U.S. Department of Education, National Center for Education Statistics, Schools and Staffing Survey (SASS) and National Teacher and Principal Survey (NTPS)


Next, we examine teacher characteristics and school conditions predictive of turnover across low- and high-FRPL schools to identify correlates with turnover. The most relevant results are found in Table 5, with models 1 and 2 comparing switchers against stayers and models 3 and 4 comparing leavers against stayers. First, we find no evidence that novice ELA teachers are more at risk of switching schools than more experienced teachers. However, ELA teachers with a graduate degree are at greater risk of switching from low-FRPL schools, holding all else constant. ELA teachers whose undergraduate degrees were from the most selective colleges are substantially more likely to switch from high-FRPL schools, but not from low-FRPL schools. Increase in salary is predicted to decrease turnover for ELA teachers in both high- and low-FRPL schools. In terms of supportive school conditions, we find stronger administrative support and better teacher cooperation are both predictive of decreasing turnover for ELA teachers at high-FRPL schools but not at low-FRPL schools.  


Table 5. Multinomial Logistic Regression of Turnover for ELA Teachers by High- and Low-FRPL School Status


Variables

Switchers

 

Leavers

 

(1)

(2)

 

(3)

(4)

 

Low-FRPL schools

High-FRPL schools

 

Low-FRPL schools

High-FRPL schools

Novice teacher

1.14

0.86

 

2.89**

2.33**

 

(0.73)

(–1.04)

 

(5.76)

(5.68)

Graduate degree

1.94**

1.05

 

1.10

0.87

 

(4.14)

(0.31)

 

(0.77)

(–0.68)

No certification

0.81

2.62+

 

1.44

3.43**

 

(–0.52)

(1.87)

 

(1.02)

(3.66)

Most sel. college

0.93

2.34**

 

0.89

1.82**

 

(–0.28)

(2.81)

 

(–0.69)

(2.73)

Very sel. college

0.92

0.92

 

1.16

1.72*

 

(–0.49)

(-0.33)

 

(0.83)

(2.26)

Salary per $1,000

0.97**

0.98*

 

0.98**

0.97**

 

(–4.44)

(–2.13)

 

(–3.81)

(–4.14)

Union member

0.80

0.72+

 

0.78*

1.15

 

(–1.49)

(–1.94)

 

(–2.35)

(0.47)

Urban school

0.74+

0.98

 

0.93

0.67

 

(–1.94)

(–0.14)

 

(–0.76)

(–1.49)

Student discipline

1.09

0.98

 

1.09*

1.04

 

(1.53)

(–0.36)

 

(2.33)

(0.62)

Admin support

0.92

0.76**

 

0.92+

0.80**

 

(–1.09)

(–2.59)

 

(–1.72)

(–2.94)

Teacher cooperation

0.90+

0.88*

 

0.97

0.96

 

(–1.71)

(–2.10)

 

(–0.40)

(–0.65)

Observations

12,280

5,270

 

12,280

5,270

Note. Nationally representative weights are employed. Sample sizes are weighted to the nearest 10 in accordance with NCES nondisclosure rule. Z-statistics from heteroskedastic-robust state-level clustered standard errors are in parentheses. All models control for other teacher and school characteristics from Table 3 including school size and level and employ year and state fixed effects.

+ p < 0.10

* p < 0.05

** p < 0.01

Source: U.S. Department of Education, National Center for Education Statistics, Schools and Staffing Survey (SASS) and National Teacher and Principal Survey (NTPS)


In comparison, novice ELA teachers are substantially more likely to leave teaching in both low- and high-FRPL environments. ELA teachers without certification are likely to leave the teaching profession in high-FRPL schools. Similar to model 2, we observe teachers from the most selective colleges are more likely to leave teaching, but only in high-FRPL schools. Additionally, we find teachers from selective colleges are also more likely to leave teaching in high-FRPL schools. Union membership seems to reduce the risk of leaving, but only in low-FRPL schools. Similar to our findings in terms of supportive school environment, better administrative support is predictive of decreasing the likelihood of leaving for ELA teachers in high-FRPL schools.


In short, we find several factors that predict teacher turnover, and the patterns diverge for switchers versus leavers as well as in low- versus high-FRPL schools. Next we briefly examine how our overall findings for ELA teachers may contrast with findings for ESL teachers.


ESL TEACHERS


In general, the results for ESL teachers are comparable to those of the larger ELA teacher population (Appendix Tables 3–7). As such, we focus on discussing findings that are substantively different. First, we find a larger portion (29%) of ESL teachers are Hispanic relative to 7% of ELA teachers as a whole (Appendix Table 3). ESL teachers are also more likely to teach in urban settings and more likely to teach in high-FRPL and majority minority schools, 61% and 67%, respectively, which likely reflects the fact that ELs are also heavily concentrated within certain districts, as well as individual schools within districts (U.S. Department of Education, 2020). Reflecting their content area, ESL teachers work with LEP students more frequently than general ELA teachers.


When we examine differences in teacher characteristics between low- and high-FRPL schools, the most substantial difference is the 21% difference in ESL teachers being Hispanic in high-FRPL relative to low-FRPL schools, which is substantially more than the 5% difference for ELA teachers (Appendix Table 4). The other notable difference is that ESL teachers are 17% less likely to have graduate degrees in high-FRPL schools. In terms of attrition rate, ESL teachers are generally more likely to turn over than ELA teachers (Appendix Table 5). Only about 80% of ESL teachers stay in the same school from year to year, with 9% switching schools and 11% leaving the profession. This elevated risk of turning over can be observed even when we account for a rich set of teacher and school characteristics. The results indicate that, although ELA teachers are not more likely to turn over (Table 4), ESL teachers are substantially more likely to turn over, and are especially more likely to leave the profession (Appendix Table 6). Lastly, we find limited evidence of factors associated with ESL teachers switching or leaving, but this is likely due to extremely small sample size (Appendix Table 7). One notable finding is that increases in administrative support are associated with decreased risk of switching schools and leaving the profession. In short, we do find some interesting patterns for ESL teachers relative to ELA teachers. Next we discuss how our findings fit in the larger literature and the implications of our results for policy and practice.


DISCUSSION AND CONCLUSION


This study examines teacher attrition for ELA and ESL teachers, a key component of the K–12 education workforce given the increasing number of EL students in the United States. Overall, we find that attrition rates for ELA teachers are similar to attrition rates of non-ELA teachers, with approximately 8% of ELA teachers leaving the profession and another 7% leaving their current school. However, ESL teachers leave at a higher rate, with approximately 11% of ESL teachers leaving the profession and another 9% leaving their current school.


Despite the fact that ELA teachers are increasingly working at schools with greater rates of student poverty, we find that there is no difference in attrition between ELA teachers working at low-poverty schools and high-poverty schools. Instead, we find that factors that predict retention of ELA teachers include more administrative support and more teacher collaboration, particularly for teachers working in high-poverty schools, and higher salaries. Novice ELA teachers and ELA teachers without certification are more likely to leave. This fits the same pattern we see in the literature on teacher attrition that examines the teaching labor force as a whole.


In contrast, ESL teachers are more likely to leave both their current school and the profession. This is particularly concerning given that the EL population is increasing while at the same time there is a shortage of trained-ESL teachers (Sutcher et al., 2016). However, we found the factors associated with attrition for ESL teachers to be similar to those of the ELA teachers, as well as to other teachers in previous studies (Nguyen et al., 2020). In particular, we found that lack of administrative support is an important factor associated with attrition of ESL teachers. Given that ESL teachers may be facing more challenging work conditions, support and resources specifically targeted to the unique challenges ESL teachers have may be necessary to lessen attrition. In particular, boosting administrator support for challenges such as addressing racism and xenophobia may create a more positive school climate (Crutchfield et al., 2020; Voight et al., 2015). In addition, ESL teachers may be able to play a role in combating the negative perceptions other students and teachers have of language minority students, such as being less likely to complete college (Blanchard & Muller, 2015) and having less involved parents (Ho & Cherng, 2018).


Our study was particularly interested in looking at the attrition rates and factors associated with attrition in high-poverty schools. Although ELA teachers working in high-poverty schools do not turn over at a higher rate than teachers in low-poverty schools, there are several working conditions that are predictive of ELA teacher turnover in high-poverty schools that, in many cases, are not factors in teacher turnover from low-poverty schools. Similar to prior research, working conditions seem to play a more enhanced role in low-income and disadvantaged schools than in more affluent schools (Nguyen, 2021). Relatedly, we find teacher characteristics, such as experience, certification type, and college selectivity, as well as school characteristics, including administrative school support and salary, are important components in reducing teacher turnover. Focusing on recruitment of qualified ELA teachers and improving workplace conditions may be important policy tools in the retention of ELA teachers, specifically for high-poverty schools. This is especially important given that ELA teachers in this sample consistently reported that they received less administrative support and lower levels of teacher cooperation than other types of teachers.


Better administrative support and stronger cooperation also may help retain teachers and may be accomplished in several different ways. For instance, better administrative support includes administrators being encouraging of the faculty, sharing their leadership and decision-making powers with teachers, and being protective of teachers from outside pressures from parents that may interfere with their teaching (Boyd et al., 2011; Djonko-Moore, 2016). Relatedly, districts may consider providing antiracist professional development, which may improve classroom practice as well as how teachers interact with other teachers and their own students (Denson, 2009; King & Castenell, 2001; Lawrence & Tatum, 1997). Moreover, districts should also consider hiring more minority teachers to reflect the demographics of their students, which for many EL students would be Hispanic ELA or ESL teachers, because this race/ethnicity match may improve teacher–student relationships as well as learning (Redding, 2019). In other words, schools need conditions that may boost and improve the interactions between administrators and teachers as well as interactions among teachers themselves and with their students.


 In sum, our study makes several contributions to the scholarly study of ELA teachers, an important group of teachers who have been underexamined in the literature. In addition to examining the changes in ELA teacher characteristics and qualifications over a 28-year period as well as the changes in student characteristics and the school conditions in which ELA teachers work during this time period, we find several factors that predict attrition and retention of ELA teachers in high- versus low-poverty schools and how these factors have differential relationships for the various forms of teacher turnover. These contributions should go a long way to addressing the gap in the sparse literature on ELA teachers and their mobility patterns.


Notes


1.

For an example of this flexibility, see the Ohio Department of Education website (2021).


References


Allington, R. L., & McGill-Franzen, A. (Eds.). (2018). Summer reading: Closing the rich/poor reading achievement gap. Teachers College Press.


Bastian, K. C. & Henry, G. T. (2015). Teachers without borders: Consequences of teacher labor force mobility. Educational Evaluation and Policy Analysis, 37(2), 163–183.


Blanchard, S., & Muller, C. (2015). Gatekeepers of the American dream: How teachers’ perceptions shape the academic outcomes of immigrant and language-minority students. Social Science Research, 51, 262–275.


Borman, G. D., & Dowling, N. M. (2008). Teacher attrition and retention: A meta-analytic and narrative review of the research. Review of Educational Research, 78(3), 367–409.


Boyd, D., Grossman, P., Ing, M., Lankford, H., Loeb, S., & Wyckoff, J. (2011). The influence of school administrators on teacher retention decisions. American Educational Research Journal, 48(2), 303–333.


Boyd, D., Grossman, P., Lankford, H., Loeb, S., & Wyckoff, J. (2008). Who leaves? Teacher attrition and student achievement. NBER Working Paper No. 14022.


Boyd, D., Lankford, H., Loeb, S., Rockoff, J., & Wyckoff, J. (2008). The narrowing gap in New York City teacher qualifications and its implications for student achievement in high-poverty schools. NBER Working Paper No. 14021.


Boyd, D., Lankford, H., Loeb, S., Ronfeldt, M., & Wyckoff, J. (2010). The role of teacher quality in retention and hiring: Using applications to transfer to uncover preferences of teachers and schools. Journal of Policy Analysis and Management, 30(1), 88–110.


Carver, R. P. (2000). The causes of high and low reading achievement. Routledge.


Cohen-Vogel, L., & Smith, T. M. (2007). Qualifications and assignments of alternatively certified teachers: Testing core assumptions. American Educational Research Journal44(3), 732–753.


Crutchfield, J., Phillippo, K. L., & Frey, A. (2020). Structural racism in schools: A view through the lens of the National School Social Work Practice Model. Children & Schools, 42(3), 187–193.


Denson, N. (2009). Do curricular and cocurricular diversity activities influence racial bias? A meta-analysis. Review of Educational Research79(2), 805–838.


Djonko-Moore, C. M. (2016). An exploration of teacher attrition and mobility in high poverty racially segregated schools. Race Ethnicity and Education, 19(5), 1063–1087.


Duran, C. S. (2019). On issues of discrimination and xenophobia: What can TESOL practitioners do to support and advocate for refugee students? TESOL Quarterly, 53(3), 818–827.


Goldhaber, D., Gross, B., & Player, D. (2011). Teacher career paths, teacher quality, and persistence in the classroom: Are public schools keeping their best? Journal of Policy Analysis and Management, 30(1), 57–87.


Goldhaber, D., Krieg, J., Theobald, R, & Liddle, S. (2021). Lost to the system: A descriptive exploration of where teacher candidates find employment and how much they earn. CALDER Working Paper No. 251-0421.


Goldhaber, D., Lavery, L., & Theobald, R. (2015). Uneven playing field? Assessing the teacher quality gap between advantaged and disadvantaged students. Educational Researcher44(5), 293–307.


Guarino, C. M., Santibanez, L., & Daley, G. A. (2006). Teacher recruitment and retention: A review of the recent empirical literature. Review of Educational Research, 76(2), 173–208.


Gray, L., & Taie, S. (2015). Public school teacher attrition and mobility in the first five years: Results from the first through fifth waves of the 2007-08 Beginning Teacher Longitudinal Study. NCES 2015-337. National Center for Education Statistics.


Grossman, P., Loeb, S., Cohen, J., & Wyckoff, J. (2013). Measure for measure: The relationship between measures of instructional practices in middle school English language arts and teachers’ value-added scores. American Journal of Education, 119(3), 445–470.


Hahs-Vaughn, D. L., & Scherff, L. (2008). Beginning English teacher attrition, mobility, and retention. The Journal of Experimental Education77(1), 21–54.


Hancock, C. B., & Scherff, L. (2010). Who will stay and who will leave? Predicting secondary English teacher attrition risk. Journal of Teacher Education61(4), 328–338.


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


Hanushek, E. A., Rivkin, S. G., & Schiman, J. C. (2016). Dynamic effects of teacher turnover on the quality of instruction. Economics of Education Review, 55, 132–148.


Henry, G. T., Fortner, C. K., & Bastian, K. C. (2012). The effects of experience and attrition for novice high-school science and mathematics teachers. Science, 335, 1118–1121.


Hess, F. M., & Hochleitner, T. (2012). College rankings inflation: Are you overpaying for prestige? Education Outlook, 3. American Enterprise Institute for Public Policy Research.


Ho, P., & Cherng, H. S. (2018). How far can the apple fall? Differences in teacher perceptions of minority and immigrant parents and their impact on academic outcomes. Social Science Research, 74, 132–145.


Hochschild, A. (1983). Comment on Kemper’s “Social constructionist and positivist spproaches to the sociology of emotions,” American Journal of Sociology, 89(2), 432–434.


Ingersoll, R. M. (2001). Teacher turnover and teacher shortages: An organizational analysis. American Educational Research Journal38(3), 499–534.


Kelly, S. (2004). An event history analysis of teacher attrition: Salary, teacher tracking, and socially disadvantaged schools. The Journal of Experimental Education, 72(3), 195–220.


Kelly, S., & Northrop, L. (2015). Early career outcomes for the “best and the brightest” selectivity, satisfaction, and attrition in the beginning teacher longitudinal survey. American Educational Research Journal52(4), 624–656.


Kieffer, M. J., & Thompson, K. D. (2018). Hidden progress of multilingual students on NAEP. Educational Researcher47(6), 391–398.


King, S. H., & Castenell, L. A. (2001). Racism and racial inequality: Implications for teacher education. AACTE Publications.


Kraft, M. A., Marinell, W. H., & Yee, D. S-W. (2016). School organizational contexts, teacher turnover, and student achievement: Evidence from panel data. American Educational Research Journal, 53(5), 1411–1449.


Lankford, H., Loeb, S., & Wyckoff, J. (2002). Teacher sorting and the plight of urban schools: A descriptive analysis. Educational Evaluation and Policy Analysis, 24(1), 37–62.


Lawrence, S. M., & Tatum, B. D. (1997). Teachers in transition: The impact of anti-racist professional development on classroom practice. Teachers College Record99, 162–178.


Liz, J. (2020). State racism, social justice, and the COVID-19 pandemic. Public Philosophy Journal, 3(1), 1–11.


McFarland, J., Hussar, B., Zhang, J., Wang, X., Wang, K., Hein, S., Diliberti, M., Forrest Cataldi, E., Bullock Mann, F., & Barmer, A. (2019). The condition of education 2019. NCES 2019-144. National Center for Education Statistics.


National Center for Education Statistics. (2019).The nation’s report card. National Center for Education Statistics. https://www.nationsreportcard.gov


Nguyen, T. D. (2021). Linking school organizational characteristics and teacher retention: Evidence from repeated cross-sectional national data. Teaching and Teacher Education97, 1–14.


Nguyen, T. D., Pham, L. D., Crouch, M., & Springer, M. G. (2020). The correlates of teacher turnover: An updated and expanded meta-analysis of the literature. Educational Research Review31, 1–17.


Nguyen, T. D., & Redding, C. (2018). Changes in the demographics, qualifications, and turnover of American STEM teachers, 1988–2012. AERA Open4(3), 1–13.


Ohio Department of Education. (2021). Characteristics of programs serving EL students in Ohio. https://education.ohio.gov/Topics/Student-Supports/English-Learners/Teaching-English-Learners/Characteristics-of-Programs-Serving-LEP-Students-i


Papay, J. P., Bacher-Hicks, A., Page, L. C., & Marinell, W. H. (2017). The challenge of teacher retention in urban schools: Evidence of variation from a cross-site analysis. Educational Researcher, 46(8), 434–448.


Petrie, G. M., & Darragh, J. J. (2018). “This worry machine”: The impact of executive orders on ELL teachers. TESOL Journal, 9(3), 412–430.


Redding, C. (2019). A teacher like me: A review of the effect of student–teacher racial/ethnic matching on teacher perceptions of students and student academic and behavioral outcomes. Review of Educational Research89(4), 499–535.


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.


Redding, C., & Smith, T. M. (2016). Easy in, easy out: Are alternatively certified teachers turning over at increased rates? American Educational Research Journal, 53(4), 1086–1125.


Roberto, K. J., Johnson, A. F., & Rauhaus, B. M. (2020). Stigmatization and prejudice during the COVID-19 pandemic. Administrative Theory & Praxis, 42(3), 364–378.


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


Schmidt, L. M. (2018). Trauma in English learners: Examining the influence of previous trauma and PTSD on English learners and within the classroom. TESOL Journal, 10(1), 1–10.


Sorensen, L. C., & Ladd, H. F. (2020). The hidden costs of teacher turnover. AERA Open, 6(1), 1–24.


Sutcher, L., Darling-Hammond, L., & Carver-Thomas, D. (2016). A coming crisis in teaching? Teacher supply, demand, and shortages in the U.S. Learning Policy Institute. https://learningpolicyinstitute.org/sites/default/files/product-files/A_Coming_Crisis_in_Teaching_REPORT.pdf


U.S. Department of Education. (2020). Our nation’s English learners: What are their characteristics? https://www2.ed.gov/datastory/el-characteristics/index.html


Voight, A., Hanson, T., O’Malley, M., & Adekanye, L. (2015). The racial school climate gap: Within-school disparities in students’ experiences of safety, support, and connectedness. American Journal of Community Psychology56(3–4), 252–267.


Wayne, A. J., & Youngs, P. (2003). Teacher characteristics and student achievement gains: A review. Review of Educational Research, 73(1), 89–122.


APPENDIX TABLES


Appendix Table 1. Variable Descriptions


Employment Status

Leavers

Switchers

Stayers

Leavers are teachers who left the teaching profession.

Switchers are teachers who switched to a new school.

Stayers are teachers who are currently teaching in the same school.

Teacher Characteristics

Female

A dichotomous variable where 1 = female and 0 = male.

Black

A dichotomous variable where 1 = Black and 0 = non-Black.

Asian

A dichotomous variable where 1 = Asian and 0 = non-Asian.

American Indian

A dichotomous variable where 1 = American Indian and 0 = non-American Indian.

Hispanic

A dichotomous variable where 1 = Hispanic and 0 = non-Hispanic.

White

A dichotomous variable where 1 = White and 0 = non-White.

Novice

A dichotomous variable where 1 = teacher has less than three years of teaching experience and 0 = teacher has three or more years of teaching experience.

Under 30

A dichotomous variable where 1 = teacher is younger than 30 years old and 0 = teacher is 30 or older.

Graduate degree

A dichotomous variable where 1 = teacher has graduate degree and 0 = no graduate degree.

Teaches ELA

A dichotomous variable where 1 = teacher’s subject is English, communications, composition, journalism, language arts, reading, speech, or English as a Second Language and 0 = other subjects.

Teaches ESL

A dichotomous variable where 1 = teacher’s subject is English as a Second Language and 0 = other subjects.

Teaches STEM

A dichotomous variable where 1 = teacher’s subject is math or science and 0 = other subjects.

Teaches SPED

A dichotomous variable where 1 = teacher’s subject is special education and 0 = other subjects.

No certification

A dichotomous variable where 1 = teacher has no certification and 0 = teacher has any certification.

Most selective college

A dichotomous variable where 1 = teacher’s undergraduate college/university has Barron’s classification of most competitive or highly competitive and 0 = Barron’s classification is very competitive, competitive, less competitive, or noncompetitive.

Very selective college

A dichotomous variable where 1 = teacher’s undergraduate college/university has Barron’s classification of very competitive and 0 = Barron’s classification is most selective, highly competitive, competitive, less competitive, or noncompetitive.

Salary ($1,000)

A continuous variable of the base teaching salary for the entire school year, scaled in $1,000s, and in constant 2012 dollars.

Satisfy w/salary (std)

On a scale of 1 = strongly disagree and 4 = strongly agree, teachers report on how satisfied they are with their salary. Measure standardized for each wave.

Union member

A dichotomous variable where 1 = teacher is a union member and 0 = teacher is not a union member.

School Characteristics

Urban school

A dichotomous variable where 1 = school is classified as urban by U.S. census and 0 = non-urban area as classified by U.S. census.

K–12 enrollment

A continuous variable of the size of school where the teacher is teaching in the base year.

Secondary school

A dichotomous variable where 1 = the school is classified as a secondary school and 0 = the school is not classified as a secondary school.

Combined elem-sec

A dichotomous variable where 1 = the school is classified as a combined elementary and secondary (K–8) school and 0 = the school is not classified as a combined elementary and secondary school.

Percent FRPL students

Percentage of students eligible for the federal free or reduced-price lunch program.

Majority FRPL

A dichotomous variable where 1 = the majority of students at the school are eligible for federal free or reduced-price lunch and 0 = the majority of students at the schools are not eligible for federal free or reduced-price lunch.

Percent minority students

Percentage of non-White students enrolled in a school.

Majority minority

A dichotomous variable where 1 = the majority of students at the school are non-White and 0 = the majority of students at the school are White.

Percent LEP

Percentage of students classified as limited English proficient (LEP).

Student discipline (std)

On a scale of 1 = never happens to 5 = happens daily, the principal reports of six kinds of student discipline problems: physical conflict, robbery or theft, vandalism, student use of alcohol, drug use, and possession of weapons.

Admin support (std)

On a scale of 1 = strongly disagree and 4 = strongly agree, teachers report on the school administration’s behavior toward the staff is supportive and encouraging (standardized).

Teacher coop (std)

On a scale of 1 = strongly disagree and 4 = strongly agree, teachers report on the level of cooperative effort among the staff members. Measure standardized for each wave.


Appendix Table 2. Multinomial Logistic Regression of Turnover for ELA Teachers by Bottom and Top Quartile FRPL School Status


Variables

Switchers

 

Leavers

 

(1)

(2)

 

(3)

(4)

 

Bottom quartile FRPL schools

Top quartile FRPL schools

 

Bottom quartile FRPL schools

Top quartile FRPL schools

Novice teacher

1.35

0.90

 

1.75

1.92**

 

(0.97)

(–0.63)

 

(1.23)

(3.44)

Graduate degree

2.98**

0.96

 

0.94

0.84

 

(4.69)

(–0.28)

 

(–0.31)

(–0.87)

No certification

0.17**

3.06*

 

4.44**

3.79**

 

(–2.76)

(2.36)

 

(2.77)

(3.73)

Most sel. college

0.87

2.46**

 

0.64+

1.81**

 

(–0.33)

(2.73)

 

(–1.77)

(2.84)

Very sel. college

0.90

0.91

 

1.48*

1.66*

 

(–0.36)

(–0.35)

 

(2.14)

(2.03)

Salary per $1,000

0.98*

0.99*

 

0.99

0.97**

 

(–2.34)

(–1.99)

 

(–0.98)

(–4.58)

Union member

0.66

0.75+

 

0.95

1.11

 

(–1.52)

(–1.94)

 

(–0.27)

(0.30)

Urban school

1.45+

0.95

 

0.47**

0.70

 

(1.74)

(–0.29)

 

(–3.90)

(–1.32)

Student discipline

0.68

0.91+

 

1.04

1.03

 

(–1.46)

(–1.71)

 

(0.84)

(0.43)

Admin support

1.11

0.79*

 

0.97

0.84*

 

(0.70)

(–2.22)

 

(–0.35)

(–2.20)

Teacher cooperation

0.90

0.87*

 

0.99

0.92

 

(–0.83)

(–2.35)

 

(–0.11)

(–1.40)

Observations

3,570

5,030

 

3,570

5,030

Note. Nationally representative weights are employed. Sample sizes are weighted to the nearest 10 in accordance with NCES nondisclosure rule. Z-statistics from heteroskedastic-robust state-level clustered standard errors are in parentheses. All models control for other teacher and school characteristics from Table 3 including school size and level and employ year and state fixed effects.

+ p < 0.10

* p < 0.05

** p < 0.01

Source: U.S. Department of Education, National Center for Education Statistics, Schools and Staffing Survey (SASS) and National Teacher and Principal Survey (NTPS)


Appendix Table 3. Descriptive Statistics for ESL Teachers


 

(2)

(3)

(4)

(5)

(6)

(7)

(8)

 

Wave: 1991

Wave: 1994

Wave: 2000

Wave: 2004

Wave: 2008

Wave: 2012

Wave: 2016

Teacher Characteristics

       

Female

0.85

0.82

0.85

0.84

0.90

0.88

0.89

Black

0.06

0.02

0.06

0.06

0.02

0.04

0.04

Asian

0.04

0.05

0.05

0.06

0.03

0.03

0.06

American Indian

0.01

0.00

0.01

0.00

0.01

0.03

0.01

Hispanic

0.17

0.16

0.17

0.35

0.30

0.31

0.29

White

0.73

0.77

0.71

0.54

0.65

0.62

0.62

Teacher age

43.27

44.44

44.20

42.37

43.93

44.98

43.39

Novice teacher

0.10

0.09

0.08

0.12

0.14

0.05

0.13

Graduate degree

0.52

0.50

0.56

0.49

0.57

0.57

0.63

No certification

0.16

0.15

0.11

0.02

0.02

0.00

0.01

Most selective college

0.00

0.11

0.07

0.17

0.16

0.10

0.13

Very selective college

0.00

0.21

0.25

0.12

0.16

0.20

0.26

Salary per $1,000

52.05

51.58

55.15

58.68

55.76

58.60

55.83

Union member

-

0.75

0.79

0.74

0.77

0.79

0.71

School Characteristics

       

Urban school

0.45

0.47

0.42

0.49

0.35

0.39

0.39

Percent FRPL

0.37

0.38

0.43

0.54

0.49

0.60

0.59

Majority FRPL school

0.31

0.32

0.39

0.54

0.43

0.69

0.61

Percent minority students

0.49

0.49

0.48

0.62

0.57

0.64

0.63

Majority minority school

0.48

0.51

0.43

0.64

0.54

0.71

0.67

Percent IEP

-

-

0.12

0.16

0.11

0.12

0.13

Percent LEP

-

-

0.15

0.22

0.17

0.20

0.35

Student discipline (std)

0.09

–0.03

–0.22

–0.04

–0.12

0.03

0.06

Admin. support (std)

-

0.11

0.03

–0.10

0.03

–0.10

0.04

Teacher cooperation (std)

-

0.02

0.05

0.02

0.01

0.02

–0.05

Observations

470

430

330

340

320

350

410

Note. Nationally representative weights are employed. Sample sizes are weighted to the nearest 10 in accordance with NCES nondisclosure rule. Salary has been adjusted to constant 2012 dollars.

Source: U.S. Department of Education, National Center for Education Statistics, Schools and Staffing Survey (SASS)


Appendix Table 4. Descriptive Statistics of ESL Teachers by High- and Low-FRPL School Status


 

Wave: 2000

  

Wave: 2012

Wave: 2016

 

Low-FRPL

High-FRPL

Diff

  

Low-FRPL

High-FRPL

Diff

Low-FRPL

High-FRPL

Diff

Teacher Characteristics

Female

0.84

0.86

0.02

  

0.95

0.85

–0.10*

0.92

0.87

-0.05

Black

0.03

0.11

0.08

  

0.04

0.03

–0.01

0.02

0.05

0.02

Asian

0.04

0.06

0.01

  

0.05

0.02

–0.03

0.04

0.07

0.03

American Indian

0.01

0.00

0.00

  

0.01

0.05

0.04

0.02

0.01

–0.01

Hispanic

0.15

0.20

0.05

  

0.24

0.34

0.11

0.16

0.37

0.21**

White

0.77

0.63

–0.14+

  

0.67

0.60

–0.07

0.78

0.52

–0.25**

Teacher age

44.26

44.10

–0.15

  

43.55

45.62

2.07

43.77

43.15

–0.62

Novice teacher

0.09

0.06

–0.03+

  

0.05

0.05

0.01

0.10

0.15

0.06

Graduate degree

0.58

0.54

–0.03

  

0.64

0.54

–0.10

0.73

0.56

–0.17**

No certification

0.11

0.11

0.00

  

0.00

0.00

0.00

0.00

0.01

0.01

Most sel. college

0.08

0.07

–0.01

  

0.17

0.06

–0.11+

0.15

0.11

–0.05

Very sel. college

0.23

0.28

0.05

  

0.20

0.20

0.00

0.31

0.22

–0.10**

Salary per $1,000

55.66

54.37

–1.29

  

56.92

59.34

2.43

57.96

54.44

–3.52

Union member

0.79

0.78

–0.01

  

0.80

0.79

–0.01

0.73

0.70

–0.03

School Characteristics

Urban school

0.30

0.59

0.29*

  

0.24

0.46

0.22+

0.23

0.50

0.27**

Percent FRPL

0.22

0.76

0.54**

  

0.26

0.75

0.5**

0.24

0.82

0.58**

Percent minority

0.37

0.67

0.3**

  

0.39

0.75

0.36**

0.43

0.76

0.33**

Major. minority

0.26

0.70

0.45**

  

0.31

0.89

0.57**

0.39

0.85

0.46**

Percent IEP

0.12

0.11

0.00

  

0.12

0.12

0.00

0.11

0.14

0.03*

Percent LEP

0.08

0.26

0.18**

  

0.08

0.25

0.18**

0.24

0.41

0.17*

Student discip.

–0.21

–0.23

–0.03

  

–0.11

0.10

0.21+

–0.03

0.12

0.15

Admin. support

0.04

0.02

–0.02

  

–0.05

–0.12

–0.07

0.06

0.03

–0.02

Teacher coop

0.11

–0.05

–0.17

  

0.17

–0.05

–0.22

–0.02

–0.08

–0.05

Observations

330

  

350

 

410

 

Note. Nationally representative weights are employed. Sample sizes are weighted to the nearest 10 in accordance with NCES nondisclosure rule. Salary has been adjusted to constant 2012 dollar.

+ p < 0.10

* p < 0.05

** p < 0.01

Source: U.S. Department of Education, National Center for Education Statistics, Schools and Staffing Survey (SASS)


Appendix Table 5. Rate of Attrition for ESL Teachers


 

(1)

(2)

(3)

(4)

(5)

 

Wave: 2000

Wave: 2004

Wave: 2008

Wave: 2012

Pooled

Panel A: Attrition Rate for ESL Teachers

Stayer

73.91

82.58

81.82

80.27

80.45

Switcher

8.22

10.13

6.55

9.62

8.74

Leaver

17.88

7.29

11.63

10.11

10.81

Observations

330

340

320

350

1330

Note. Nationally representative weights are employed. Sample sizes are weighted to the nearest 10 in accordance with NCES nondisclosure rule. NTPS 2015–2016 does not contain attrition data.

Source: U.S. Department of Education, National Center for Education Statistics, Schools and Staffing Survey (SASS)



Appendix Table 6. Logistic and Multinomial Logistic Regression of Teacher Turnover by FRPL School Status


 

(1)

(2)

 

High-FRPL

High-FRPL with ESL interaction

Panel A: Overall Turnover from Logistic Regression

ESL teacher

1.35**

1.64**

 

(2.77)

(2.75)

High-poverty indicator

0.98

0.99

 

(–0.54)

(–0.37)

ESL teacher*

 

0.71

High-poverty indicator

 

(–1.02)

Panel B: Switchers from Multinomial Logistic Regression

ESL teacher

1.37

1.80*

 

(1.49)

(2.49)

High-poverty indicator

0.98

0.99

 

(–0.24)

(–0.10)

ESL teacher*

 

0.61

High-poverty indicator

 

(–0.99)

Panel C: Leavers from Multinomial Logistic Regression

ESL teacher

1.32*

1.50*

 

(2.19)

(2.03)

High-poverty indicator

0.97

0.98

 

(–0.68)

(–0.60)

ESL teacher*

 

0.81

High-poverty indicator

 

(–0.77)

Observations

102,400

102,400

Note. Nationally representative weights are employed. Sample sizes weighted to the nearest 10 in accordance with NCES nondisclosure rule. Z-statistics from heteroskedastic-robust state-level clustered standard errors are in parentheses. The majority indicator corresponds to the model title. The interaction term is the interaction between ESL teacher and the majority indicator. All models control for teacher race/ethnicity, gender, age, teaching experience, undergraduate selectivity, graduate degree, certification, passing content exam, qualification, salary, urbanicity, school enrollment, school-level characteristics, administrative support, and cooperation among the staff, along with state and year fixed effects.

+ p < 0.10

* p < 0.05

** p < 0.01

Source: U.S. Department of Education, National Center for Education Statistics, Schools and Staffing Survey (SASS)



Appendix Table 7. Multinomial Logistic Regression of Turnover for ESL Teachers by High- and Low-FRPL School Status


Variables

Switchers

 

Leavers

 

(1)

(2)

 

(3)

(4)

 

Low-FRPL schools

High-FRPL schools

 

Low-FRPL schools

High-FRPL schools

Novice teacher

1.10

3.41+

 

2.92

1.34

 

(0.19)

(1.89)

 

(1.58)

(0.24)

Graduate degree

1.24

1.38

 

0.89

0.41+

 

(0.55)

(0.62)

 

(–0.40)

(–1.93)

No certification

1.28

0.43

 

2.76

0.55

 

(0.43)

(–0.70)

 

(1.29)

(–0.56)

Most sel. college

0.61

3.30*

 

1.20

0.21

 

(–0.72)

(2.46)

 

(0.33)

(–1.37)

Very sel. college

0.56

2.60*

 

1.97

1.72

 

(–1.33)

(2.09)

 

(1.30)

(0.77)

Salary per $1,000

1.01

0.97

 

1.01

0.99

 

(0.74)

(–1.55)

 

(0.67)

(-0.47)

Union member

0.88

0.73

 

0.34**

1.07

 

(–0.39)

(–0.59)

 

(–2.85)

(0.13)

Urban school

1.05

0.85

 

0.57

1.11

 

(0.11)

(–0.29)

 

(–1.11)

(0.26)

Student discipline

0.85

1.71+

 

1.20

1.53+

 

(–0.94)

(1.78)

 

(1.45)

(1.85)

Admin support

0.81

0.50**

 

1.04

0.60**

 

(–0.86)

(–5.64)

 

(0.21)

(–2.87)

Teacher cooperation

0.93

0.82

 

1.04

0.76

 

(–0.36)

(–1.14)

 

(0.17)

(–1.36)

Observations

740

590

 

740

590

Note. Nationally representative weights are employed. Sample sizes are weighted to the nearest 10 in accordance with NCES nondisclosure rule. Z-statistics from heteroskedastic-robust state-level clustered standard errors are in parentheses. All models control for other teacher and school characteristics from Table 3 including school size and level and employ year fixed effects. State fixed effects are not employed due to small sample size of ESL teachers.

+ p < 0.10

* p < 0.05

** p < 0.01

Source: U.S. Department of Education, National Center for Education Statistics, Schools and Staffing Survey (SASS)







Cite This Article as: Teachers College Record Volume 123 Number 10, 2021, p. -
https://www.tcrecord.org ID Number: 23856, Date Accessed: 12/8/2021 4:57:32 AM

Purchase Reprint Rights for this article or review
 
Article Tools
Related Articles
There are no related articles to display

Related Discussion
 
Post a Comment | Read All

About the Author
  • Tuan D. Nguyen
    Kansas State University
    E-mail Author
    TUAN D. NGUYEN, Ph.D., is an assistant professor in the Department of Curriculum and Instruction at Kansas State University. His main research interests include teacher leadership and school improvement, teacher labor markets, and financial aid and postsecondary success. He has recently published in the American Education Research Journal and Teaching and Teacher Education.
  • Laura Northrop
    Cleveland State University
    E-mail Author
    LAURA NORTHROP, Ph.D., is an assistant professor in the Department of Teacher Education at Cleveland State University. Her research focuses on struggling readers, middle school literacy instruction, and educational policy. Recent publications examining the type of texts assigned eight-grade students have been published in Reading Research Quarterly and The Elementary School Journal.
 
Member Center
In Print
This Month's Issue

Submit
EMAIL

Twitter

RSS