Abstract
This study examines the retention rates of new teachers in Texas given their initial certification type and initial school setting. The 5-year persistence rates of 175,664 teachers from 2000 to 2015 are analyzed using logistic regression to calculate the probability of new teacher persistence at the school, district, and state level. Main findings suggest that teachers’ certification type and initial school setting are statistically significant predictors of new teacher persistence, and traditionally licensed teachers initially placed in traditional public school are more likely to persist when compared with other preparation types or initial placements into charter schools.
Keywords
Introduction
Schools in the United States face persistent difficulty in hiring enough teachers; the exact type of teacher and locations facing shortage are documented in a 200-page list maintained by the U.S. Department of Education spanning 30 years and all 50 states (Cross, 2017). To address this shortage, Texas has initiated broad policy shifts regarding teacher preparation and certification. In doing so, Texas is emblematic of ongoing efforts nationwide to rapidly and sustainably attract, place, and retain teacher candidates. The U.S. Department of Education’s Race to the Top program both recognized the national crisis associated with teacher shortages and offered explicit incentives for states that increased the number of available pathways to teaching. Texas led the nation in deregulating teacher licensure and continues to prioritize alternate routes to certification among its education policies (Feistritzer, Griffin, & Linnajarvi, 2011; Walsh & Jacobs, 2007). Furthermore, Texas has recently signaled ongoing support for alternative certification programs and allows districts to waive alternative certifications in favor of hiring candidates who can be certified on the job by the district.
In an effort to reduce barriers to entering the profession and to place new teachers into classrooms as quickly as possible, Texas has taken a free-market approach to teacher licensure and granted a variety of agencies significant authority to train and deploy new teachers rather than relying on traditional schools of education (Guthery & Richards, 2017). This approach has resulted in a proliferation of alternative pathways, and teacher certification is currently offered by community colleges, online programs, nonprofits, for-profit companies, and school districts. Texas now relies heavily on nontraditionally certified teachers to sufficiently staff schools. For seven of the 10 years represented in this study, the number of alternatively certified novice teachers outnumbered the number of traditionally certified novice teachers (see Appendix A).
Over this same time period, Texas has also introduced policies which permitted the expansion of charter schools. Like traditional and alternative certification, charters are exempt from many of the regulations governing traditional public schools, allowing competition within the public education system. In 2018, the U.S. Department of Education approved the Texas improvement plan filed under the Every Student Succeeds Act (ESSA), which includes special provisions for charter schools. According to a policy review by Fordham, the Texas plan is projected to infuse hundreds of millions of dollars of investment and expansion of charter schools and set a precedent for other states to use their Title I funds on charters (Petrilli, 2018). However, there is limited evidence regarding the interaction of these two deregulatory policies, specifically, how alternately prepared teachers fare after placement in a charter school.
It is evident that an infusion of resources and creativity is necessary to recruit and train new teachers in Texas and nationwide. However, because teacher retention is rarely an incentivized outcome of preparation programs or of placement schools, most policies focus on recruiting new teachers. To understand whether initial preparation paths are equally effective, this article compares the rates of persistence among teachers certified in traditional and alternative pathways. Specifically, this study examines whether pipelines to the teaching profession, which represent broader autonomy on the part of the certifiers, combined with teachers’ initial placements in traditional public or charter schools, result in a corps of teachers who are likely to remain in schools for at least 5 years. We integrate several lines of inquiry, including teacher preparation, teacher turnover, and charter school research to investigate the ways in which teachers’ preparation pathways and initial school placements are associated with staying in or leaving the profession. Much is made of widening teacher pipelines, and advocates suggest that interested candidates should be granted pathways by which they can be trained for and ushered into classrooms cheaply and quickly. However, this article suggests that some combinations of licensure pathway and initial school type result in higher attrition, and teachers are more likely to turn over long before the conventionally recognized 5-year benchmark.
Literature Review
Teacher Turnover
Ingersoll’s seminal 2001 study challenged prevailing theories that attributed high vacancy rates in districts to retirement and increasing student enrollment. To counter these narratives, Ingersoll used multiple regression analysis of the 1991-1992 national Schools and Staffing Survey (SASS) database, inclusive of 6,733 teachers. Based on this study, Ingersoll drew several important conclusions about the teacher labor market. First, the primary source of demand for teachers is neither the result of teacher retirements nor of increasing student enrollment, but it is due, instead, to teacher turnover. Second, when compared with similar professions in terms of education and training, teaching has a significantly higher turnover rate. Third, 42% of all teachers leaving reported their primary reasons as job dissatisfaction or career opportunities outside of education. In high-poverty urban schools, teachers also cited low salaries, lack of administrative support, student discipline problems, lack of student motivation, and lack of influence over decision-making as reasons for leaving. Taken together, “These findings suggest that teacher supply and demand imbalances and attendant school staffing problems are neither synonymous with, nor primarily caused by, teacher shortages in the technical sense of a supply-side deficit of qualified candidates” (Ingersoll, 2001, p. 524). Ingersoll concludes that it is the “revolving door” of the teaching profession which creates excess demand in the labor market, not retirement or new students. This revolving door of the teaching profession is a concern for myriad reasons, but two primary concerns emerge: (a) teacher turnover engenders a significant financial cost and (b) teacher experience can be linked to student achievement.
It is estimated that teacher turnover costs the nation US$7 billion a year (Carroll, 2007). In addition, teacher turnover is concentrated in urban schools, so it is an additional financial burden in districts that already experience economic disadvantages (Watlington, Shockley, Guglielmino, & Felsher, 2010). In Texas, the State Board for Educator Certification commissioned a study on the cost of teacher turnover (Benner, 2000). The agency found the cost of recruiting and training each teacher is significant—some estimates suggest as much as US$5,000 per teacher throughout the state. The hiring district is then responsible for their ongoing professional development and training. When a teacher leaves, he or she takes with them the financial investment a district has made in his or her training and leaves the district with zero return on that investment. The district must then start the recruitment and hiring process again, making the same investment in training over and over. In the year 2000 alone, the total financial loss to Texas due to teacher turnover was estimated to be between US$321 million and US$2.1 billion.
Districts with high turnover face not only the financial cost of new teachers, but also achievement losses which often result from having inexperienced teachers year after year. Teacher experience strongly correlates with student experience in the first 3 years of a teacher’s career (Boyd, Grossman, Lankford, Loeb, & Wyckoff, 2006; Hanushek & Rivkin, 2007). In addition, recent research indicates “large returns to experience for middle school teachers in the form both of higher test scores and improvements in student behavior” (Ladd & Sorenson, 2017). However, the experience of having an inexperienced or ineffective teacher year after year can have a negative cumulative effect. A student may never catch up academically after having an inexperienced teacher, and the end result may be a lifetime of reduced educational and earning potential when compared with students who consistently had experienced or effective teachers (Rivkin, Hanushek, & Kain, 2005).
If teacher turnover is a liability, then teacher consistency can be conceptualized as its own kind of capital or resource in schools and school systems. One recent and important contribution to teacher turnover literature suggests that some schools experience episodes of instability related to their teacher workforces while others experience chronic instability (Holme, Jabbar, Germain, & Dinning, 2017). That study goes on to say that average turnover statistics belie the extremity of the problem experienced by schools with high rates of poverty, low performance, and high proportions of minoritized students. Such schools are more vulnerable to becoming a high turnover school (turnover entry) and have difficulty recovering from high turnover status (turnover exit). We find that such schools are also likely to suffer from multiple types of instability problems at the same time. (p. 70)
If teacher turnover is identified as a chief source of instability, it is imperative that both researchers and practitioners work to understand why teachers move or leave and what can be done to stem the tide.
Recently, Redding and Smith (2016) used four cohorts of SASS data to examine preparation characteristics of newly certified teachers and subsequent patterns of retention. They found that alternatively certified teachers turn over at higher rates than traditionally certified teachers even after controlling for a variety of school characteristics. However, entrants who received new teacher supports such as mentorship were less likely to leave or move, with alternatively certified teachers benefiting most from mentorship. Redding and Smith also examined school setting and found that alternatively and traditionally certified novice teachers were equally likely to work in a charter school setting but did not examine turnover differences based on school context. Using the same SASS data, Smith and Ingersoll (2004) also found positive effects for mentorship. They found that the new teacher induction programs, which resulted in the highest retention rate of new teachers, centered on mentorship from experienced teachers and not reduced course load or an aide. Thus, the value of teacher retention is twofold in that teacher consistency results in experienced mentors for future new teachers, reinforcing retention.
Other research has identified imbalances in teacher turnover between charter and traditional public schools. Stuit and Smith (2012) found that the turnover rate in charter schools was twice that of traditional public schools. Their multistate study identified teacher inexperience and lack of union membership as two significant factors in charter turnover. Miron and Applegate (2007) also found charter school teachers more likely to leave than their public school counterparts. They identified age, lack of certification, and lack of experience as significant factors predicting charter teacher turnover. In addition, they found that new teacher turnover could range from 15% to 40% across states, with new teachers at charter schools the most likely to leave.
Teacher Sorting
Teacher sorting, the name for the phenomenon of teacher distribution and movement among schools, is not a random process, and there are distinct differences between teachers of high and low socioeconomic status (SES) students. While the Elementary and Secondary Education Act (1965) dictates equitable distribution of highly qualified teachers among students with varying needs, that goal has yet to be achieved. Imazeki and Goe (2009) found that poor and non-White students do not have access to as many highly qualified teachers as their wealthier, White peers. New teachers are less effective than their veteran peers and more likely to leave the school or teacher workforce (Walsh & O’Tracy, 2007). Placement patterns continue to indicate that the most novice teachers are placed in the highest need settings, which often precludes the possibility of establishing an expert, sustainable teaching corps among the schools and students with the most need. Importantly, then, issues of teacher sorting and the consequences for schools are at work before teachers commence their first lesson.
In 2004, Hanushek, Kain, and Rivkin examined teacher migration patterns in Texas. They found that 25% of teachers who taught in the lowest achieving quartile of schools leave each year. In the highest achieving quartile, less than 20% leave. The biggest difference between these two populations was that the teachers who left the lowest performing schools left the profession altogether. Furthermore, Hanushek, Kain, and Rivkin (2004) found that teachers may move among schools most often based on student demographics. That is, teachers most often move to a school that matches their own ethnic background. Hanushek et al. (2004) speculate that, because housing tends to be segregated by race, teachers may move to teach closer to home. In a confirming study from 2004, Smith and Ingersoll found that a school with 75% of students eligible for free or reduced-priced lunches (FRL) experience teacher attrition 50% more often than schools with only 25% of the population eligible for FRL. Low-SES districts have a higher rate of turnover and bear the burden of the additional cost that represents in terms of recruitment and replacement. Simultaneously, more affluent teachers with a few years’ experience choose to move closer to their home, which means that students in poverty are most profoundly affected by teacher attrition.
Typically, teacher turnover occurs when the stress or workload becomes untenable, when they seek out better compensation for a similar job elsewhere, or when the compensation simply cannot account for the extent and varieties of labor teachers are asked to undertake. Mental health concerns may also play a role in the voluntary aspects of teacher sorting. Teaching is typically a stressful and cognitively demanding job, and many of a teacher’s responsibilities fall outside of working hours. Studies repeatedly suggest that novice teachers are among the least likely demographic to persist in the profession (Ingersoll, 2003; Ingersoll & Perda, 2008; Miron & Applegate, 2007). Two factors seem to play into the decision to remain in a stressful work environment: the teacher’s individual degree of ideological commitment and the school’s capacity to support, mentor, develop, and sustain teachers despite undesirable circumstances (Cochran-Smith et al., 2012).
Teachers may switch jobs for a variety of personal reasons, as noted previously, but they also move schools to meet their professional goals. As Atteberry, Loeb, and Wyckoff (2017) found, teachers may seek new positions (and new schools) for the possibility of being more effective in another position, the appeal of another grade or subject position, or changes in school enrollment which mandate reassignments throughout a school. Teachers may also be mandated to switch grades or content areas as a result of turnover. When a teacher leaves a school, it is rarely a matter of merely hiring, replacing, or moving a single teacher to make up for the shortfall. In fact, researchers have found that, in the wake of one teacher’s departure, a school must make 3.8 personnel swaps (Atteberry et al., 2017), further destabilizing the workforce and relocating the established expertise within a given school.
Gaps in the Literature
Within the teacher preparation market, there is an array of pathways which place capable teachers in classrooms across the country. However, even though there is substantial extant research on teacher retention, little is known about the interaction between those pathways and teachers’ initial placement schools. This article sheds light on how individuals’ preparation pathways, in combination with their placement schools, promote retention or contribute to increased deterioration of the teacher workforce.
To better understand how the proliferation of alternate routes has influenced the production, preparation, placement, and persistence of teachers, this study builds on the turnover literature and contributes new information regarding the ways in which pathways from preparation to placement are likely to influence teachers’ likelihood of persisting in the profession or leaving it. In addition, this study is able to distinguish between movers (those who switch schools) and leavers (those who leave the profession; Ingersoll 2001). One major contribution to the extant literature is that our analyses count educators who remain in the profession for at least 5 years regardless of the positions they occupy in public education. Previous studies do not distinguish between educators who change roles within the state’s public education system and those who exit completely. Educators who no longer appear in the state teacher data because they have assumed school or district leadership roles appear to have exited the profession. This is colloquially referred to as the “up or out problem” by Ingersoll (2001), and we address that quandary in this study. In addition, we examine the importance of a teacher’s first placement and examine whether a new teacher’s starting school type (charter or traditional public school) has an effect on 5-year teacher persistence rates.
Purpose and Research Questions
As Texas experiments with alternative methods of teacher preparation and innovations in school choice, we investigate whether nontraditional preparation paths retain teachers at the same rate as traditional paths. Specifically, we assess 5-year persistence rates for new teachers following their transitions from preparation program to initial placement, along with the degree to which initial entry into a traditional public or charter school influences their likelihood of staying in the profession. Furthermore, we examine the lowest performing schools, and which teachers persist in the most difficult settings. We define “stayers” as meeting or exceeding the 5-year benchmark established by Ingersoll (2001) and examine the broad category of “leavers” in more detail and upheld in a more recent report from the National Center for Education Statistics (Gray & Taie, 2015). In a profession marked by persistent shortages, it is imperative to understand which preparation paths are associated with higher persistence rates in the field.
Research Questions
Our research questions are as follows:
Method
In this section, we describe the data, contextual factors, and primary analytic method employed to address our research questions.
Data
A single data set was created by merging a host of publicly available data sources, including school-level data from Common Core Data from the National Center for Education Statistics. In addition, teacher-level data from Texas Education Agency were linked using unique identification codes for each teacher and state identification codes for schools. After listwise deletion for nonmatch on teacher records, this data set contains teacher records for 175,664 new teachers in Texas from 2000 to 2015. The data were then stratified into 11 five-year cohorts. We defined a teacher using the state definition, which assigns a particular job code to anyone designated as the class teacher of record (Texas Education Agency, 2017). We define new teachers based on the first time they enter the data set with that job code, even if they were employed as aides or had years of experience prior to their initial teaching year.
As noted previously, a unique feature of this data set is the capacity to distinguish between a non-class reassignment and departure from the profession—the “up or out” dilemma (Ingersoll, 2001). That is, once a teacher is not classified with the same job code, a separate database for non-class assigned district staff is needed to follow the teacher into another role or to determine when she or he exited the profession.
The first cohort began teaching in 2000 and meets the 5-year benchmark in 2005. A new cohort is measured every year from 2000 to 2010 with the last cohort reaching the 5-year benchmark in 2015. Each cohort has approximately 15,000 new teachers with the smallest cohort in 2000 with 13,182 new teachers and the largest cohort in 2008 with 19,164 teachers (for complete cohort counts, see Appendix A).
In Texas, only 27% of all new teachers will persist for 5 years in their original school, 38% in their initial district, and 40% in any Texas public schools. Of the teachers who start in a traditional public school, 39% will persist 5 years in Texas in their district, and 62% will still be teaching in public schools after 5 years. Over the 16-year period of this study, Texas has lost 67,179 newly licensed teachers from Texas public schools, which is 38.24% of all novice teachers.
Confirming other findings, there are demographic differences in the average school setting in which teachers start and the average school setting in which they persist 5 years later (Hanushek et al., 2004). After 5 years’ experience, teachers in Texas are working in schools with fewer economically disadvantaged and more English-speaking students. However, teachers with 5 years’ experience are, on average, working in more ethnically diverse schools than the average initial school. Some of this shift is due to changing demographics in Texas over the 15-year period of the study. More students in Texas public schools are non-White, bilingual, and economically disadvantaged as compared with 2000, the first year of the study.
Contextual Factors
There are three types of school districts in Texas: independent school districts (traditional public school), charter schools, and common school districts. A common district in Texas is a rural district that may share the burden of educating few students with another common district. There were 57 new teachers over 11 years who started in a common school district, and due to the nontraditional arrangements of the rural districts, they were excluded from analysis.
The Texas Education Agency does not impose the same level of regulation on charter and traditional public schools. Traditional public schools in Texas are subject to the full regulation of all state education mandates while charter school districts have only to comply with minimum state education regulations. Charter schools are exempt from the state requirements regarding minimum teacher pay and duty-free lunch and often have required extended hours and extra duties for teachers. When comparing the two settings, the charter setting is the least regulated by the state.
Within this study, we also examine two types of preparation pathways: traditional and alternative. In the United States, traditional preparation and certification are typically linked processes. That is, candidates studying in an institution of higher education receive a bachelor’s degree, pass a series of standardized assessments, and successfully meet any other requirements set forth by individual state boards of education to become certified school teachers. However, definitions vary from state to state so that alternate and traditional pathways in different states may, in fact, resemble one another. Generally, alternative certification is available to individuals who hold a bachelor’s degree (often in an area other than education) and who want to enter the classroom without returning to additional undergraduate coursework (Woods, 2016).
In Texas, traditional preparation programs based at universities are subject to all regulations of the Higher Education Coordinating Board, and the strictest portions of the state education code regarding teacher preparation and licensure. Traditional preparation pathways, then, are those which are most constrained in their autonomy to license or certify new teachers and are typically limited to university-based preparation programs.
There are two ways to classify a teacher’s certification: at the program level and at the teacher level. The majority of the time, both classifications will be the same; however, in the case of a university-based alternative program, the certificate is classified as traditional at the program level and alternative at the teacher level. We define the teacher preparation program at the teacher level, not the institutional level. This most closely captures the entrant’s preparation experience.
In Texas, alternative teacher certification refers to an array of pathways available to individuals who want to enter a K-12 classroom who have not received traditional preparation. Alternative certification programs are designed to produce teachers quickly and as a result are exempt from portions of the traditional certification requirements. The National Education Association recognizes alternate routes as legitimate pathways to teaching and endorses alternate pathways because they allow diverse candidates to enter the profession. The National Education Association does suggest a caveat when discussing the proliferation of alternate certification programs: Every teacher candidate must meet equal requirements, and alternate preparation pathways must be as rigorous as traditional preparation pathways (National Education Association, n.d.). In this study, alternative certification is defined based on the Texas definition for each preparation program degree, which includes the following possibilities: school district training programs, online licensure programs, university-based alternative programs, and licensure programs in coordination with nonprofit and for-profit partners. There are additional pathways to certification in the state of Texas that include out-of-state transfer, aide programs, and licensure by state exam without enrollment in any preparation program. These are not labeled as alternative pathways by the state because they comprise a route to certification without enrollment in any formal Texas preparation program. All pathways not defined as traditional or alternative, we categorize as “other” (Table 1).
Initial School Setting by Teacher Preparation Type.
Primary Analytic Method
This study looks at the various pipelines of teacher preparation as well as the context for entrants’ first- and fifth-year benchmarks in the field. We analyzed the 5-year retention rates of 11 separate cohorts of teachers at the three levels of persistence to understand whether attrition is associated with teacher pathway and school type. We define three categories of educator persistence:
Persistence at the state level: After 5 years, an individual hired as a teacher remains employed in any role by a public district anywhere in the state.
Persistence at the district level: After 5 years, a teacher remains employed as a teacher in the same district in which she or he was hired, but on a different campus (i.e., a different school) than his or her initial placement.
Persistence at the campus level: After 5 years, a teacher remains employed as a teacher in the same campus (school) as initial placement.
We ran a series of logistic regression models to estimate the probability that a new teacher would persist to the 5-year benchmark controlling for certification pathway, school type, and school context. We used the following equation to estimate the outcome variable at the state, district, and campus levels of persistence.
Results
Descriptive analyses show that charter and traditional public schools exhibit markedly different hiring patterns with regard to the preparation of the teachers they employ (Figure 1).
Composition of new teacher preparation by school type.
In charter schools, 65% of the new teachers are alternatively certified, 25% are traditionally certified, and 10% are from other pathways (such as aide programs and teachers who are hired holding licenses issued out of state). In traditional public schools, the percentage of new teachers who are either traditionally or alternatively certified is roughly equal: 43% of the new teachers in traditional public schools are traditionally certified, 43% are alternatively certified. The remaining 14% receive certifications from other paths. Since 2001, 75% of the new teachers hired by charter schools were nontraditionally certified.
When examining the socioeconomic contexts of the initial school placement, we found that of the 5,099 new teachers in charter schools over the 16-year period of the study, 3,626 were initially in charter schools classified as majority non-White and majority FRL (71.11%). Among 170,565 traditional public school teachers, 82,007 were initially hired in majority non-White and high-poverty settings (48.07%). A total of 43% of all traditionally certified teachers persist 5 years in their initial or placement districts, and 31% persist 5 years in their placement schools. The persistence rates for traditionally certified teachers who start in a charter school are very different. Only 12% of traditionally certified teachers persist in the classroom in the same charter school after 5 years, and only 14% are still in the same district. Based on this finding, we included a control for majority non-White and high-poverty settings to estimate the effect of school type separate from the socioeconomic setting.
We calculated the probability that a new teacher persists 5 years in Texas public schools using the model specified in the Method section for each of the outcome variables. The reference group for these models comprises traditionally trained teachers initially hired in traditional public schools classified as majority White and low poverty. The largest beta weights are the initial district type, charter or traditional. Thus, the initial school setting type was a significant covariate in the models of persistence. Table 2 presents the results of the final regression analyses in all three categories of persistence.
Full Model Results Predicting Teacher Persistence to 5 Years.
Note. CI = confidence interval; FRL = free and reduced-priced lunches.
p < .1. **p < .05. ***p < .01.
We turn next to a description of specific results for state-level professional retention, district-level retention, and school-level retention. The probability of persistence was calculated from the model results. Full probability results of state-, district-, and campus- (school building) level analyses are presented in Table 3.
Probability Results.
Note. FRL = free and reduced-priced lunches.
Majority non-White and majority FRL eligible.
Table 3 illustrates a pattern of persistence that holds at all levels and educator roles. The most regulated teacher certification path combined with the most regulated initial school setting has the highest probability of retaining new teachers in Texas public schools. A traditionally prepared teacher’s probability of persisting in a majority White and low-poverty traditional public school is 67.5%, as compared with alternatively certified teachers, whose likelihoods are 60.6%. If a teacher’s initial setting is a charter school, the traditionally prepared teacher’s likelihood of persistence is 55.9%, whereas the likelihood of persistence from an alternate pathway is 48.4%. If the initial school setting comprises a majority of students who qualify for FRL and also a majority of students who are identified as non-White, all educators’ probabilities of persistence in any role decline (traditional: 66.2%; alternative: 59.2%), but the traditionally prepared teacher is still more likely to persist than teachers prepared in any other manner.
Leavers: Losses to Texas Public Education
At the state level, we conceptualize retention most broadly: individuals who stay in any role in public education for at least 5 years. Counting education professionals in this way allows us to include persistent individuals who have moved from classroom instruction to other roles such as a principals or instructional coaches. Individuals who have the highest probability of 5-year persistence were traditionally prepared and initially hired by traditional public schools (67.5%) where the majority of students were White and ineligible for FRL. However, controlling for school composition, the likelihood of retention does not change substantially for traditionally licensed teachers placed in traditional public schools. Conversely, when the initial school setting is a charter school, chances of 5-year retention are significantly reduced, regardless of preparation pathway.
Figure 2 illustrates the 5-year retention rates of entrants by preparation path and school type. A traditionally certified teacher has a 67.5% likelihood of staying in any role in Texas state education when initially placed into a traditional public school without the challenges associated with teaching in high-poverty contexts. Under similarly favorable conditions, a teacher hired first into a charter school has only a 55.9% likelihood of staying in the state’s school system for at least 5 years—little better than a one in two chance. A similar drop-off is evident between placement school types for alternatively certified teachers: The probability of lasting 5 years in any role statewide falls from 60.6% to 48.4% when initially hired in a charter school versus a traditional public school. Based on preparation pathways and initial placement type, a new teacher’s probability of staying in education varies greatly. We further address implications in the “Discussion” section.
Flooding the pipes: The probability of 5-year educator retention.
Movers and Stayers: Retention in the Classroom
When calculating the probability that a new teacher continues to be a classroom teacher in his or her initial school and district for 5 years, we used separate outcome variables to estimate each equation. As illustrated in Table 3, traditionally prepared teachers initially hired in traditional public schools are most likely to remain teaching in their initial districts at the 5-year benchmark (43%). Alternatively, certified teachers initially hired in charter districts are the least likely to still be teaching at the 5-year benchmark (12%).
Table 3 reports the probability that, based on certification type, new teachers are likely to persist in teaching roles in their initial districts for at least 5 years. Specifically, a traditionally prepared teacher initially placed in a majority non-White and majority FRL public school has a likelihood of persistence of 43%, as compared with 38% of alternatively certified teachers. Once again, traditionally prepared teachers are more likely to survive the novice teacher gauntlet as compared with their counterparts trained in alternative pathways.
We then examined persistence within the initial school; that is, a teacher’s likelihood of remaining in his or her initial placement school for 5 years. Probabilities regarding persistence in the placement school are, by comparison with those in district and state contexts, lower because they require the highest level of consistency, a teacher must remain a classroom teacher in the same school for 5 years. At the school level, the probability that a traditionally certified teacher initially placed into a traditional public school persists at that school is 31%, or approximately a 1 in 3 chance. Traditional preparation, then, offers an advantage over alternative preparation (27%).
In majority non-White and high-poverty school settings, the probability of persistence to 5 years is reduced to 29% for a traditionally prepared teacher and falls even further to 26% for an alternatively prepared teacher. Controlling for SES and demographic composition at the school level, the strongest predictors of new teacher persistence are traditional certification and initial placement in a traditional public school setting.
Limitations
By defining persistence at the 5-year benchmark, it is possible that annual patterns within the 5-year period are missed. Persistence is measured by a binary variable which indicates whether the novice teacher remains in a school, a district, or the state’s public education system 5 years after his or her initial placement. This method of counting does not account for teachers who may exit the public schooling system and return for any reason. While this distinction is likely relevant for a small proportion of the population, the results regarding attrition from the profession are only interpretable at the 5-year benchmark of persistence.
The current study is also limited in that it cannot determine whether a teacher left Texas public education to teach in a private school. However, given that between 5,179 and 7,068 novice teachers left Texas’s public system each year between 2000 and 2010, it is unlikely that they are all private school teachers. Thus, while we cannot track the destination of each leaver, we can suppose with some confidence that only a small percent of leavers sought and gained employment in private schools. Whether or not leavers move to private school, it remains the case that they were once willing to teach in public school and within 5 years, they were no longer willing to teach in any public education setting in Texas.
In addition, this study does not make causal claims or establish causal direction. This is primarily because there are a number of variables unaccounted for in the current study. For example, this study does not examine personal and psychological characteristics such as motivation, interest, or pre-teaching employment. For example, teacher mental health as it pertains to turnover is an important consideration for studies of teacher longevity. Stress associated with teaching has increased, coinciding with a precipitous drop in enrollment in teacher education programs and an increase in teacher turnover (Barmby, 2006; Davis & Wilson, 2000; Gold & Roth, 1993; Harvey, 2014; Simon & Johnson, 2015). While one narrow interpretation of the results suggests that alternative pathways produce less dedicated teachers, it may also stand to reason that the ample alternative certificate options available in a state like Texas may, in fact, draw individuals who want a shorter or less expensive route to the classroom and that these individuals simply do not plan to invest their careers in education. This study does not establish causality of pathway. It may be that people who are not planning to stay long in education are drawn to an alternative path. In Texas, there are many alternative certification programs that are significantly shorter and less expensive than the traditional university route. Future researchers, therefore, may wish to design studies that investigate the ways in which individual teacher characteristics predict their choice of preparation pathway and ultimately their persistence in public education.
Discussion
It is often said that teaching is a profession that cannibalizes its young, and the phrase seems apt in Texas. However, not all preparation paths and school settings churn through teachers at equal rates. This study concludes that nontraditionally prepared teachers are less likely to persist to 5 years in both traditional public and charter schools. Alternatively certified teachers have lower 5-year persistent rates than traditionally certified teachers in either setting. When a novice teacher’s initial setting is a charter school, regardless of preparation, the probability she or he remains in the profession is diminished. Not only is that individual more likely to leave the school and the district, but also the field of public education altogether. Furthermore, novice teachers who initially start teaching in charter schools leave public education at higher rates than teachers of all preparation pathways who are first placed into a traditional public school.
Of the new teachers in our study, traditionally certified teachers only made up 25% of new hires in charter schools. Our analyses reveal that the charter setting disproportionately hires new teachers from the certification paths least regulated by the Texas Education Code, which regulates all certification paths and school settings. In addition to hiring a majority of nontraditionally certified new teachers, charters are exempt from many other state regulations that govern traditional public schools. This least regulated combination has the lowest probability of retaining new teachers in the classroom or in the state system of public education.
More research is needed into the interaction of two policies—charter schools and alternative certification—which feature reduced oversight and regulation as the hallmark of innovation. Deregulation of teacher certification has reduced barriers to entering the profession, whereas, simultaneously, deregulation of charters has allowed innovations to improve school choice. Alternative and traditional certification pathways produce nearly equal numbers of new teacher candidates, but educators who are alternatively certified are far more likely to “leak” out of the profession. Furthermore, those initially placed in charter settings are most likely to be lost to public education altogether within 5 years. Taken together, these findings suggest that deregulation may inadvertently contribute to the teacher shortage—and the cost incurred by that shortage—over the early years of new teachers’ careers.
There are ample challenges associated with teacher recruitment, training, and retention, especially in majority non-White and high-poverty schools (Boyd, Grossman, Lankford, Loeb, & Wyckoff, 2009). Our findings suggest that even after controlling for socioeconomic and demographic makeup of schools, teacher preparation and school type are significant predictors of 5-year teacher persistence. This study illustrates the need for further research into teacher retention, controlling for preparation pathways and initial school placement. Future research should examine patterns of migration across schools and districts and to identify the strongest and weakest chains of teacher production and retention. In addition, ongoing investigation is necessary regarding various alternate pathways to certification with low persistence rates and identifying campuses and districts with higher than expected turnover when compared with schools in equivalent settings. While this study is not designed to address the reasons for decreased probability of teacher retention in charters, extant research shows that charters frequently exercise exemption from traditional teacher pay scales and maximum work hours standard in traditional public schools, with school culture a mitigating factor in retention (Brewer & Ahn, 2010; Malloy & Wohlstetter, 2003; Podgursky & Ballou, 2001).
While the majority of research focuses on traditional and alternative pathways, another direction for future investigation includes better understanding of the nontrivial number of new teachers who come from “other” preparation pathways. The other category includes any new teacher who was not in an alternative or traditional preparation program and, instead, was issued a certification by exam, certification based on work experience, transfer certification (including out-of-state and foreign certifications), anyone teaching prior to full certification, and educational aides who work as teachers. Approximately 10% of new teachers since 2001 have entered through this pathway, although the number has decreased over recent years. Despite these pathways showing the lowest probability of retention, these paths continue to garner political support. This study finds that the pattern of retention holds and new teachers, regardless of certification pathway, are more likely to persist in traditional public schools. The combination of teachers certified through other pathways with an initial placement in a charter school setting have the lowest rates of persistence of all the preparation paths and initial settings. It is encouraging that the certification path least successful at retaining teachers is shrinking, yet researchers would do well to attend to this category in the future.
New policies which shift resources to fostering alternate pathways to teaching in Texas have not alleviated teacher shortages, according to a longitudinal listing of shortage (Cross, 2017). In addition, the proliferation of charter schools in the state means the school setting least likely to retain teachers is being widely and rapidly replicated. The combination of increasing teacher licensing options and increasing charter funding is happening without regard to the teacher labor market. As the charter sector continues to expand in Texas and across the country, perhaps the retention rate of teachers should be considered as a metric for the success and sustainability of these programs before replicating the models.
While it may be difficult to reintroduce regulation, it is necessary to examine ways in which deregulation of certifications and settings may have unintended consequences. Further research is needed to understand and replicate what makes traditionally prepared teachers more likely to persist. As new entrants are drawn into the pipeline, we must consider whether all preparation pathways and school types are equally effective methods of deploying limited resources. A substantial number of people are initially willing to teach in public schools, however, the probability that they will stay in the profession varies greatly depending upon their preparation path and initial setting. The results of this study indicate that further public discourse is needed to engineer outcomes that are desirable and beneficial to all stakeholders. Aspiring teachers are a resource that Texas and the nation cannot afford to squander through leaky pipes.
Footnotes
Appendix
New Teacher Cohort Count by Preparation Path.
| Year | Traditional (N) | Alternative (N) | Other (N) | Cohort year (N) |
|---|---|---|---|---|
| 2000 | 4,979 | 3,367 | 4,836 | 13,182 |
| 2001 | 7,851 | 4,185 | 3,148 | 15,184 |
| 2002 | 7,544 | 5,336 | 2,748 | 15,628 |
| 2003 | 7,235 | 6,233 | 2,645 | 16,113 |
| 2004 | 5,896 | 5,914 | 1,657 | 13,467 |
| 2005 | 7,408 | 7,740 | 1,486 | 16,634 |
| 2006 | 7,158 | 7,994 | 1,323 | 16,475 |
| 2007 | 7,905 | 8,685 | 1,470 | 18,060 |
| 2008 | 7,786 | 9,960 | 1,418 | 19,164 |
| 2009 | 6,856 | 9,131 | 1,258 | 17,245 |
| 2010 | 5,699 | 7,724 | 1,089 | 14,512 |
| Total | 76,317 | 76,269 | 23,078 | 175,664 |
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
