Research paper
Leaving the teaching profession: The role of teacher stress and educational accountability policies on turnover intent

https://doi.org/10.1016/j.tate.2017.03.016Get rights and content

Highlights

  • There are two distinct kinds of teacher movement: Migration and attrition.

  • Accountability policies may affect teacher stress, which predicts teacher turnover.

  • Greater teacher experience may be related to lower migration between schools.

  • Young and experienced teachers display similar intent to leave the profession.

Abstract

This study examined the relationship between test-based accountability policy at the state level, teacher test stress, teacher burnout, and teacher turnover intentions, while controlling for years of teacher experience. Structural equation modeling of data from 1866 teachers across three states identified that state-specific accountability significantly predicted higher rates of test-stress, burnout, and turnover intent. Greater teacher experience was significantly related to a lower likelihood of teacher migration between schools. This study provides evidence across multiple states that test-based accountability policies may predict greater teacher turnover intent, as well as higher levels of teacher stress. Implications for future research are discussed.

Introduction

Over the past three decades, the United States has made dramatic changes in the way it measures and evaluates teacher effectiveness. In response to growing concern about the quality of education in the United States (cf. A Nation at Risk; National Commission on Excellence in Education, 1983), landmark educational legislation, including the No Child Left Behind Act (NCLB, 2002) and most recently the Every Student Succeeds Act (ESSA, 2015), has firmly established an ‘accountability era’ wherein student test scores on statewide and national educational assessments have become a national yardstick for evaluating schools, teaching quality, and school effectiveness. The practice of measuring teaching quality through student performance on standardized assessments has become standard practice in school districts across the United States (Baker et al., 2010). As a result of this practice and increasing use of student test scores as a factor in teachers' annual evaluations, the stakes for teachers have greatly increased in the current educational climate (Valli & Buese, 2007). Around the world, individual countries are implementing policies to evaluate teachers based on student test score performance. While these policies are too numerous to review individually, England may be a useful example. England passed the Education Reform Act in 1988 (ERA, 1998), and with it, implemented a national curriculum, including Ofsted inspections and national standardized assessments to evaluate student progress (Gipps, 2003). As in the United States, these assessments have been linked with a rise in reported teacher stress, student test anxiety, and school climate (Denscombe, 2000, Putwain and Symes, 2011, Putwain, 2008). Moreover, in a survey from the OECD Teaching and Learning International Survey which sampled 23 countries, 65% of teachers reported that student test scores formed a major component of their teacher evaluations (OECD, 2013). Among the countries surveyed, a majority of teachers within the several countries, including Mexico, Chile, Poland, Turkey and Bulgaria, reported that student test scores were not only incorporated, but of moderate or high importance in their teacher evaluations. Given that numerous countries around the world have begun implementing test-based accountability policies similar to the United States, it is important to look at one country's test-based accountability policies closely and examine how these policies influence teacher outcomes overall.

In recent years, some states in the United States have passed and adopted individual test-based accountability policies that punish or reward teachers based on the extent to which they raise student test scores. While practices vary based on state, they may include the use of student test scores in teacher bonus pay decisions, in the decisions to award tenure, and in professional evaluation scores. Further, within states, local educational agencies differ in how they enact teacher evaluation guidelines. Finally, teachers that instruct in subjects on the state test are often evaluated differently from teachers who teach untested subject areas. Thus, due to differences in accountability policies in each state, in how policies are implemented at a district level, and in what subject teachers instruct, each teacher may experience somewhat different consequences for low student test scores (von der Embse, Pendergast, Segool, Saeki, & Ryan, 2016).

State-specific systems vary according to state, but many share common features related to teacher evaluations and teacher tenure. In Connecticut, Public Act 12–116 (2012) required the use of an approved teacher evaluation system to measure teacher effectiveness. The state-specific system Smarter Balance was implemented in 2013–2014. Based on this legislation, teachers with tenure may be dismissed for ineffectiveness. Additionally, a provision was passed which required 22.5% of the teacher evaluation to be based on student test scores (Pryor, 2012). Although intended to be implemented in 2013–2014, this was delayed for two years until 2015–2016. In Maryland, the Partnership of Assessment for Readiness for College and Careers (PARCC) was implemented in 2014–2015. In that system, 20% of teacher evaluations are based on student test scores, and another 10% is based on school-wide test scores or other measurement criteria, including attendance and graduation rates. The recent passage of HB 1167 (2014) in Maryland prohibits the use of student standardized test scores in teacher personnel decisions until 2016–2017.

In Pennsylvania, the School Performance Profile was created in 2013 to measure student progress across time (PA Department of Education). The SPP includes student test scores and school criteria, such as attendance and graduation, to measure school quality. Student test scores are weighted heavily in this system. For example, within an annual evaluation, 15% of the teacher evaluation is based on student test score growth, and another 15% is based on school-wide test scores, graduation rates, and attendance. Despite variation in how student test scores are used to evaluate teachers, all three states have linked student test scores to an evaluation of teacher performance. If a school has low test scores, they may face restrictions in access of state funds.

Given the similarities in policies across states, there is a need to examine how the implementation of high-stakes accountability policies influences teachers across schools, districts, and states. Currently, most research on teacher attrition is limited to within-state research and may occur post-onset of NCLB, but prior to the onset of Common Core (e.g., Clotfelter et al., 2004, Jones and Egley, 2007). Little is known about how teachers across the United States are affected personally and professionally by these policies as a whole, many of which have been introduced in recent years and gone unexamined (Ysseldyke et al., 2004). Moreover, given that teacher attrition and teacher stress are global phenomenon, studying how test-based accountability may impact teacher stress and turnover in the United States may act as a case study for the influence of policy on teacher well-being around the world. Understanding the relationship between policy and teacher outcomes is particularly important given the high rates of teacher attrition currently plaguing the field of education.

Teaching is known to be a highly stressful profession (Kyriacou, 2001). Teacher stress has been linked with adverse professional outcomes, including burnout, absenteeism, stress, and attrition (von der Embse et al., 2015a, Menken, 2006, Yoon, 2002). It may be that changes in educational accountability policies at both the federal and state level trickle down to contribute to increased stress and adverse outcomes for teachers. Recent research suggests that the use of test-based accountability in performance evaluations, merit pay, and tenure decisions results in increased test-related stress in the environment, increased stress related to the teaching curriculum, and increased teacher stress in general and specific to testing (von der Embse et al., 2016).

Historically, teacher attrition has been linked to stress, burnout, salary, and job dissatisfaction (Betoret, 2006, Darling-Hammond, 2000, Sass et al., 2011). Teacher attrition is a widely acknowledged challenge facing the teaching profession in the United States, and some studies estimate that as many as 40–50% of new teachers depart in their first five years of teaching (Darling-Hammond, 2010, Ingersoll, 2001). Teacher longevity, or the amount of time a teacher remains in the teaching profession, is an important issue for school climate and school resource allocation. Teachers grow more effective the longer they teach and they typically require significantly more investment in training and development at the start of their careers (Berry, 2010). Indeed, newer teachers typically are less effective in their first year of teaching (Rice, 2003). Therefore, when schools train teachers who do not remain in the field, this represents a discrepancy between economic resources invested and the professional output that the teacher contributes back to the school and education profession. Understanding if accountability policy pressures related to test-based accountability are associated with teachers being less likely to remain in the field of teaching would be important as teacher attrition has significant financial impacts on districts in training, professional development, hiring expenses, and instructional quality impacts, as more experienced teachers are more effective.

Of the 250,000 teachers newly hired in the United States each year, half are entering into the profession for the first time (Darling-Hammond, 2010); the remaining hires are the result of teacher migration between schools or are teachers returning to the profession. Ingersoll (2001) estimated that 13.2% of teachers change schools yearly. During the time of No Child Left Behind, it was found that higher rates of attrition and migration occurred in under-resourced and under-funded districts (Darling-Hammond, 2000). The negative impact of this turnover is even greater when seen in the context that over 200,000 fewer teachers entered teacher training programs in 2014 than in 2008 (U.S. Department of Education, Office of Postsecondary Education, 2015). In England, almost half of the state teachers polled said they planned to leave the profession in the next five years, and almost 8 out of 10 schools said they struggled to recruit teachers to fill their schools (Lightfoot, 2016). This turnover impacts schools financially through the administrative costs and time required for the hiring and training of new teachers, as well as the sunk cost of investing in teachers who do not remain in the field.

Teacher attrition represents a potentially devastating loss of personnel and resources at the school level. Schools must expend time and resources to rebuild their staff when they lose personnel (Barnes, Crowe, & Schaefer, 2007), which is especially concerning for high-turnover schools. Rather than allocating the money to students and classroom teachers, high-turnover schools must face costs in recruitment, re-training, and retention efforts for new teachers. For example, the National Commission on Teaching and America's Future estimates the cost of a teacher leaving to be as high as $17,862 per teacher. In addition to fiscal costs, turnover further impairs school climate by reducing the number of experienced teachers in schools and disrupting the consistency of classroom instruction.

While it is clear that teacher turnover, or teachers leaving their positions, is expensive and problematic, the research on where teachers go after leaving a position is sparse. As indicated by Darling-Hammond (2010), teachers typically leave a school for two reasons: to leave the profession (“attrition” from the profession) or to leave one school for another (“migration” within the profession). Most studies on teacher attrition do not distinguish between these two kinds of movement, although they may have different causes and different effects. Teacher migration may reflect young teachers changing schools in an establishment of their professional identity early in their careers (Huberman, 1995). While it does negatively impact the school that they leave, teacher migration does provide continuity in the profession. Unfortunately, few teacher attrition studies have clarified whether teachers who were leaving were planning to leave the profession entirely or were moving to another school. This study addresses this gap by predicting two kinds of teacher movement: teacher attrition and teacher migration, and clarifying the relationship between changes in educational policy at the federal and state level, two well-established key predictors of teacher longevity, stress, burnout, attrition, and migration.

Stress is one of the clearest predictors of teacher attrition (Borman and Dowling, 2008, Brownell et al., 1997). Stress is typically defined as a physiological and psychological response to a situation perceived as threatening to one's resources (Betoret, 2006, Lazarus and Folkman, 1984). Burnout, which is defined as an extreme form of teacher stress, is characterized by “emotional fatigue, disengagement, irritability, and apathy related to the work environment” (Butler & Constantine, 2005, p. 5). Causes of burnout vary widely, including pressure from administration, student discipline, inadequate training, student discipline, pay or working hours, and the teacher role (Boyd et al., 2011, Butler and Constantine, 2005, Klassen and Chiu, 2010, Kukla-Acevedo, 2009). While teacher stress has been widely studied in relation to teacher attrition, measures of stress have been highly variable across studies, without a unified conceptualization or definition (von der Embse, Kilgus, et al., 2015). Because recent accountability policies have shifted how states use testing data to evaluate and make professional determinations about teachers, including retention decisions and merit pay, there may be significant influences on teacher stress which are yet undetermined. Unfortunately, widely used measures of teacher stress, such as the Teacher Stress Inventory (Fimian, 1988), were developed prior to NCLB and the onset of high-stakes testing. Therefore, despite its wide use as a measure of teacher stress, it may not capture the multi-dimensional nature of high-stakes testing stress that was introduced into the contemporary educational climate. Since most research on teacher attrition has used measures of teacher stress that examine stress broadly, it may be that the specific stress experienced in relation to high-stakes testing and accountability has gone unexamined. This study addresses this limitation by using a narrow measure of educator testing stress in order to specifically examine the association between test-based accountability policies, burnout, and high-stakes testing stress. There is preliminary evidence indicating that test-based accountability policies may heighten teacher stress and the development of burnout. Numerous studies have documented that high-stakes testing may be stressful for teachers (Saeki et al., 2015, von der Embse et al., 2016, Berryhill et al., 2009). Further, school climate (Grayson & Alvarez, 2008) and instructional practice (Pendergast and Kaplan, 2015, Putwain and Roberts, 2009) have been found to be influenced by accountability changes at the federal and state level that then impact schools at the organization, teacher, and student level.

Within the school, new teachers are a particularly vulnerable subgroup for attrition (Darling-Hammond, 2000). Therefore, it is important to consider the relationship between attrition, teacher test stress, and teacher experience, or whether accountability differentially influences new teachers within their first five years of teaching.

Although the link between teacher stress and teacher attrition is well-established (Borman & Dowling, 2008), it is important to consider whether this link functions equally for all teachers across their teaching career and whether or not teacher test stress uniquely contributes to teacher attrition. Some studies suggest that teachers are more vulnerable to stress in certain points of their careers (Huberman, 1993). For example, teachers within the first five years of teaching experience competing developmental demands relative to the stabilization of their identities, such as marriage and child-rearing (Ingersoll, 2001). The discrepancy between stated expectations of the role during a training program and the reality of the job may also lead to shock and abrupt departure (Huberman, 1995). However, it is currently unknown whether new teachers are more vulnerable to attrition than existing teachers as a result of test-based accountability policies. Because many of the seminal studies on teacher attrition occurred around the onset of No Child Left Behind (e.g., Ingersoll, 2001), it is important to reconsider teacher attrition within the context of the contemporary policy shift.

It is important to know if new teachers are uniquely vulnerable to departure because they represent a special investment economically for school districts. Early career teachers who leave the profession are lost prior to the development of teaching expertise (Darling-Hammond, 2010), which is when teachers become the most effective (Rice, 2003). Therefore, when districts expend resources training teachers who do not remain, they do not benefit from the realized potential of experienced teachers. Teacher experience is one of the main indicators of teacher quality (Rice, 2003); consequently, it is not surprising that experienced teachers typically have students who attain higher test scores than new teachers (Greenwald et al., 1996, Rivkin et al., 2005). Moreover, it is possible that accountability pressure based on standardized test scores may uniquely affect new teachers, who may be more vulnerable to administrative pressure regarding increasing student test scores than experienced teachers. By examining teacher experience, this study specifically examines the relationship between test-based accountability policies, burnout, and high-stakes testing stress with respect to both new and experienced teachers.

When considering teacher attrition, it is also important to consider those factors that might promote professional continuity and limit teacher departures. School climate, or the “quality and character of a school life,” is a measure of the strength of the relationships within the school community (Bear, Gaskins, Blank, & Chen, 2011). A positive school climate leads to less stress in teachers (Schwab, 2001); whereas, negative school climate is associated with negative outcomes including burnout, stress around high-stakes testing, and likelihood of leaving (Collie, Shapka, & Perry, 2012). Grayson and Alvarez (2008) found a relationship between burnout and teacher-student relationships, with teacher-student relationships explaining 15% of the variance in burnout. Therefore, school climate may be an important mediator of the relationship between accountability pressures, stress, and attrition and migration in teachers. If school climate explains a portion of the variance in teacher test stress, it may be that positive school climate may reduce the development of test stress symptoms and attrition/migration in school setting. This is the first study to date that considers the mediational link between school climate, test stress, burnout, and attrition and migration while also considering test-based accountability policies.

In the current high-stakes testing era around the globe, teachers are increasingly pressured to increase student test performance (Valli & Buese, 2007) and are experiencing greater stress (Berryhill et al., 2009). Although states in the United States have implemented sweeping test-based accountability reforms, few studies have examined how these accountability changes have influenced teacher outcomes. This is especially important to consider given that countries around the world have implemented similar test-based accountability policies around teacher evaluations, performance pay decisions, and tenure. As such, there is a need to examine how accountability policies influence teacher attrition and migration, especially since this has been widely cited as a major problem in the profession of teaching and as significant revenue loss and fiscal liability for schools (Darling-Hammond, 2010). Moreover, more needs to be understood about how school climate, high-stakes testing stress, and burnout mediate the relationship between accountability and attrition and migration (von der Embse, Schultz, & Draughn, 2015). This study adds to the extant literature on teacher attrition by addressing limits in previous research through considering the unique impact of accountability pressure, testing stress, and burnout on the likelihood of leaving the profession of teaching or moving to another school. Additionally, this study controls for teacher experience, thereby examining whether teacher attrition and migration varies for new teachers as compared to more experienced ones. This study may therefore serve as a paradigm for how national accountability policy can influence teacher professional outcomes around the world.

In this study, test-based accountability is defined as the consequences attached to student standardized test performance for teachers, including tenure decisions, provision of performance or merit pay, and influence on performance evaluations. High-stakes testing stress is defined as the manifestation (expression of anxious symptoms) and sources of stress (pressure from colleagues or administrators) related to high-stakes achievement testing for students. Burnout is defined as the extreme stress related to the role of teaching, characterized as exhaustion, depersonalization, and absenteeism. Teacher attrition is defined as the likelihood of leaving the profession entirely. Teacher migration is defined as the likelihood of leaving the school for another teaching position. This is meant to capture the multidimensional nature of teacher turnover, which may be movement within the field or out of the profession entirely. Four research questions were examined:

  • 1)

    Does test-based accountability policy influence teacher test stress, burnout, attrition, and migration?

  • 2)

    How does high-stakes testing stress predict teacher attrition and teacher migration?

  • 3)

    Does teacher years' experience (i.e., number of years taught) predict teacher attrition differently than it predicts teacher migration?

  • 4)

    Does school climate mediate high-stakes test stress, teacher burnout, attrition, and migration?

Section snippets

Participants

Public school teachers were recruited from Pennsylvania, Maryland, and Connecticut to participate in a survey on State Achievement Testing and Curricular Standards via the online survey platform Qualtrics. Because accountability measures vary widely by state, three different states were sampled to provide an array of accountability procedures across the United States. State specific accountability policies are summarized in Table 1, including use of scores in tenure decisions, merit pay, and

Preliminary analyses

Descriptive statistics (means, standard deviations, skewness, and kurtosis) are reported in Table 2. Only one variable, “Student performance on state standardized tests is included in my merit/performance pay decisions,” exceeded the acceptable limit for skewness and kurtosis ( ±2; George & Mallery, 2009). Due to the highly varied nature of accountability policy across states and districts, it is unsurprising that this data is skewed. This variable was expected to reflect non-normal

Discussion

This study represents an important first step in understanding how accountability pressure at the state level influences teacher stress and teacher professional outcomes. Previous studies have found that accountability pressure predicts teacher stress, burnout, and negative instructional practices (Saeki et al., 2015, von der Embse et al., 2015b). This study, which uses a robust sample across three states, identifies test-based accountability policy as a predictor of teacher attrition. Given

Conclusion

Encouraged by mandates through federal legislation like No Child Left Behind (NCLB, 2002), states have begun implementation of accountability pressure which has raised the stakes for teachers. This study has several important implications for policymakers considering test-based accountability policies in the United States. All aspects of test-based accountability influenced teachers’ decisions to migrate between schools or leave the profession entirely, both of which drain resources and damage

References (65)

  • J. Berryhill et al.

    The effects of educational accountability on teachers: Are policies too stress provoking for their own good?

    International Journal of Education Policy and Leadership

    (2009)
  • F.D. Betoret

    Stressors, self-efficacy, coping resources, and burnout among secondary school teachers in Spain

    Educational Psychology

    (2006)
  • G.D. Borman et al.

    Teacher attrition and retention: A meta-analytic and narrative review of the research

    Review of Educational Research

    (2008)
  • D.J. Boyd et al.

    The influence of school administrators on teacher retention decisions

    American Educational Research Journal

    (2011)
  • M.T. Brownell et al.

    Attrition in special education: Why teachers leave the classroom and where they go

    Exceptionality

    (1997)
  • S. Butler et al.

    Collective self-esteem and burnout in professional school counselors

    Professional School Counseling

    (2005)
  • C.T. Clotfelter et al.

    Do school accountability systems make it more difficult for low-performing schools to attract and retain high- quality teachers?

    Journal of Policy Analysis and Management

    (2004)
  • R.J. Collie et al.

    School climate and social–emotional learning: Predicting teacher stress, job satisfaction, and teaching efficacy

    Journal of Educational Psychology

    (2012)
  • L. Darling-Hammond

    Teacher quality and student achievement

    Education Policy Analysis Archives

    (2000)
  • L. Darling-Hammond

    Recruiting and retaining teachers: Turning around the race to the bottom in high-need schools

    Journal of Curriculum and Instruction

    (2010)
  • M. Denscombe

    Social conditions for stress: Young people's experience of doing GCSEs

    British Educational Research Journal

    (2000)
  • Education Reform Act 1988 (UK). Retrieved on 25 January, 2016 from...
  • Every Student Succeeds Act of 2015, Pub. L. No. 114-95, § 114, Stat. 1177...
  • M. Fimian

    Teacher stress inventory

    (1988)
  • D. George et al.

    SPSS for windows step by step: A simple guide and reference, 17.0 update

    (2009)
  • Gipps, C. V.. (2003, April). Educational accountability in England: The role of assessment. Paper presented at the...
  • R. Greenwald et al.

    The effect of school resources on student achievement

    Review of Educational Research

    (1996)
  • L.T. Hu et al.

    Evaluating model fit

  • M. Huberman

    Burnout in teaching careers

    European Education

    (1993)
  • M. Huberman

    Professional careers and professional development: Some intersections

  • R.M. Ingersoll

    Teacher turnover and teacher shortages: An organizational analysis

    American Educational Research Journal

    (2001)
  • R.M. Ingersoll et al.

    The wrong solution to the teacher shortage

    Educational Leadership

    (2003)
  • Cited by (0)

    View full text