Abstract

This article critically reviews the recent empirical literature on teacher recruitment and retention published in the United States. It examines the characteristics of individuals who enter and remain in the teaching profession, the characteristics of schools and districts that successfully recruit and retain teachers, and the types of policies that show evidence of efficacy in recruiting and retaining teachers. The goal of the article is to provide researchers and policymakers with a review that is comprehensive, evaluative, and up to date. The review of the empirical studies selected for discussion is intended to serve not only as a compendium of available recent research on teacher recruitment and retention but also as a guide to the merit and importance of these studies. KEYWORDS: teacher attrition, teacher labor market, teacher recruitment, teacher retention. It is the goal of the elementary and secondary public school system in the United States to provide a high-quality education to every student. To do so requires an adequate supply of competent individuals who are willing and able to serve as teachers. Districts and schools are constantly engaged in activities related to the recruitment and retention of their instructional staff. In the face of a growing schoolaged population, schools and districts must struggle to maintain standards for teaching quality while continuously recruiting bright new teachers and seeking to retain their most effective existing teachers. The dual goals of recruiting and retaining effective teachers are often difficult to realize because of insufficient and sometimes dwindling resources. Current U.S. economic conditions are causing many states to roll back their expenditures on public education. The people who dispense federal, state, and local funds to education will be hard-pressed to determine which programs raise the quality of teaching in the most cost-effective manner. It is, therefore, of particular importance at this time that we turn to reliable empirical research to inform the policy community of the advantages and shortcomings of various policies to recruit and retain teachers. The intent of this article is to assemble and discuss the recent published empirical evidence that helps answer the following question: What strategies promote the recruitment and retention of teachers? Guarino et al. 174 To develop our understanding of the context in which policies affecting the supply of teachers are formed, this broad question was broken down into four main subquestions: 1. What are the characteristics of individuals who enter the teaching profession? 2. What are the characteristics of individuals who remain in teaching? 3. What are the characteristics of schools and districts that successfully recruit and retain teachers? 4. What types of policies show evidence of efficacy in recruiting and retaining teachers? Our goal is to provide researchers and policymakers with a review of the research relating to teacher recruitment and retention that is comprehensive, evaluative, and up to date. Thus our review of the empirical studies selected for discussion is intended to serve not only as a compendium of available research on the topic of teacher recruitment and retention but also as a guide to the merit and importance of these studies. The review is structured as follows. The next section presents a conceptual framework that outlines the factors influencing supply and demand in the teacher labor market. Following that, we describe our methodology for selecting and evaluating research for review. The next section contains our review of the literature pertaining to the four research questions above, and the final section presents our summary, conclusions, and suggestions for future research. Conceptual Framework Teacher recruitment and retention are two aspects of the overall labor market for teachers. From the standpoint of the districts and schools that hire teachers, recruitment and retention policies have a direct impact on their ability to fill the desired numbers of teaching slots. These policies, together with current labor market conditions, have a direct impact on the decisions of teachers or prospective teachers to remain in or enter teaching. The conceptual framework developed for this review is derived directly from the economic labor market theory of supply and demand. The general theory can be found in economics texts, such as We define the demand for teachers as the number of teaching positions offered at a given level of overall compensation and the supply of teachers as the number of qualified individuals willing to teach at a given level of overall compensation. By overall compensation, we mean not only salaries (including bonuses, other forms of monetary compensation, and expected future earnings) and benefits but also any other type of reward derived from teaching that can be encompassed under the heading of "working conditions" or "personal satisfaction." The prevailing or negotiated levels of salaries, benefits, and working conditions in a given school district will determine the number of teachers the district will be willing to employ and the number of qualified teachers who will be willing to teach. These principles of supply and demand and the factors that influence them provide a logical frame- Teacher Recruitment and Retention 175 work within which policies relating to recruitment and retention can be investigated, understood, and evaluated. The basic principle driving the supply of teachers is the following: Individuals will become or remain teachers if teaching represents the most attractive activity to pursue among all activities available to them. By attractive, we mean desirable in terms of ease of entry and overall compensation (salary, benefits, working conditions, and personal satisfaction). These elements of attractiveness are the policy levers that can be manipulated at the school, district, or state levels to bring supply in line with demand. The demand for teachers is driven by student enrollments, class-size targets, teaching-load norms, and budgetary constraints. Teacher shortages occur in a labor market when demand is greater than supply. This can be the result of either increases in demand or decreases in supply or of both simultaneously. Conversely, teacher surpluses result when supply is greater than demand. The extent to which the demand for teachers is either unmet or exceeded will generally determine the motivation for changes in policy. The labor market for teachers is nested within and continuously influenced by a larger labor market that includes the markets for all other occupations requiring roughly similar levels of education or skill. In the supply-and-demand framework, studies that focus on teacher recruitment might investigate factors that affect an individual's decision to choose teaching as a career. In choosing teaching over other available occupations, an individual will lose the opportunity to experience the rewards, in terms of overall compensation, of those other occupations. These lost rewards are considered the "opportunity costs" of teaching. Individuals who would incur high opportunity costs by choosing teaching will be less likely to make this choice. In constructing policies that promote recruitment, the goal of policymakers would be to increase the rewards of teaching relative to those of the competing occupations available to the types of people they wish to attract. Because ease of entry, monetary compensation, working conditions, and personal satisfaction are elements of the attractiveness of teaching that can be affected by policy levers, studies that focus on recruitment might examine, for example, changes in policies related to credentialing and alternative certification requirements, early recruitment strategies, and entry-level teacher compensation. Studies focusing on retention, on the other hand, might identify factors that relate to teacher attrition. Attrition can be either voluntary or involuntary, although most attrition from teaching is voluntary, given widespread tenure rates and the prevalence of unionized grievance policies regarding termination. Again, with respect to voluntary attrition, the notion of opportunity costs comes into play. Individuals whose opportunity costs outweigh the rewards gained from teaching will be more likely to leave the teaching profession. Similar opportunity costs might induce teachers to leave specific schools or districts for others, thus creating school-or district-specific attrition. Policies that promote retention would focus on adjusting the rewards offered by teaching relative to those offered by competing occupations or activities. Studies of retention might discuss factors that determine whether a teacher decides to leave teaching because he or she wishes to retire, take another form of employment, stop out for a period of time, remain unemployed, or switch to another school or district. These studies might, for example, investigate policies that improve working Guarino et al. 176 conditions, raise salaries to reflect a value placed on experience, alter tenure or retirement rules, or promote personal satisfaction through campaigns to augment the prestige of the teaching profession or programs that foster mentoring, professional development, and career advancement opportunities. Because policies that promote recruitment and those that promote retention both focus on mechanisms to adjust the attractiveness of teaching relative to other occupations-that is, mechanisms to create rewards that outweigh the opportunity costs of becoming or remaining a teacher-it is often difficult to separate the two issues. In fact, much of the research cited in this literature review does not fall neatly into one of the two categories of recruitment or retention. Often, a study will provide information that relates to both categories. We, therefore, report on the information provided in such studies in more than one section of our review, when appropriate. The Relationship of Teacher Quality to the Supply and Demand Framework The issue of teacher quality is integrally related to the interplay of supply and demand. Because not all teachers are alike, quality is an important variable that can be adjusted by policymakers in their efforts to bring supply in line with demand. Although schools and districts would prefer teachers of the highest caliber if all else were equal, in reality, trade-offs exist when resources must be allocated among competing needs. Ideally, our literature review would contain only research that focuses on the recruitment and retention of effective teachers, because it is the goal of educational institutions to employ such teachers. Very few research studies exist, however, that combine issues of recruitment and retention with the issue of teacher quality. The primary reasons for the scarcity of this research are that (a) it is difficult to establish an agreed-upon definition of teacher quality, and (b) few sources of data exist that permit researchers to identify effective teachers and examine the factors that promote their recruitment and retention. Recent research suggests that teachers exert an influence on student achievement (e.g., 1 If standardized tests provided unambiguous and reliable indicators of student achievement and if more data existed that linked the performance of students directly to their teachers, we could learn a great deal about observable characteristics of teachers that contribute to effectiveness. If data on hiring and turnover contained a detailed list of teacher characteristics, we could learn a great deal about efforts to recruit or retain teachers with those characteristics. Unfortunately, few data sets are as complete as we would like them to be, and, as a result, few studies are able to focus on the recruitment and retention of effective teachers. Even with complete data sets, it should not always be assumed that standardized tests would provide reliable indicators of student achievement. The fact that teachers can vary in effectiveness, however, raises an important issue regarding the value of particular policies in promoting recruitment or retention. One Teacher Recruitment and Retention 177 cannot assume that an improvement in recruitment or retention rates is, per se, a valuable outcome. If these improvements are achieved at the expense of quality, then students may experience more harm than benefit from such a policy. High rates of recruitment or retention may not, in and of themselves, be desirable outcomes, although there may be negative financial and distributional consequences to turnover and the training of novices. They are desirable only when student learning improves or remains constant as a result. In this review of studies of teacher recruitment and retention, we are careful to point out references to observable characteristics of teachers that are related to effectiveness-such as measured ability, credentials, experience, and subject-matter preparation-when they are used in the research design. In summary, economic labor market theory suggests that the willingness of individuals to obtain the necessary qualifications and work as teachers depends on the desirability of the teaching profession relative to alternative opportunities. Individuals compare the overall compensation-salaries, benefits, working conditions, and various forms of rewards-offered by teaching with that offered by other jobs or activities available to them. Schools and districts can influence elements of overall compensation to bring supply in line with their demand for teachers. In addition, they may adjust their standards of teacher quality according to whether teachers are in short or large supply. Throughout this review, we will return to these principles and discuss their applicability to the evidence presented in the empirical research. Methods Our literature review included all studies that we found that met a set of specific criteria. The selections were made on the basis of the following four general criteria: (a) relevance, (b) scholarship (c), empirical nature, and (d) quality. To assess the relevance of a study, we first determined whether it provided insight into recent issues surrounding the recruitment and retention of teachers in the United States. We limited studies to those performed on teacher labor markets in the United States that were published by the end of 2004 and used data that reached 1990 or later. The latter restriction was imposed to provide readers with access to the set of studies most directly relevant to our research questions at present. Although many valuable studies of teacher labor markets use data that end before 1990, we do not review those studies because we are concerned with highlighting only recent trends. Because of the importance of many older studies, however, we compare the findings of the newer research with those of the older research at the end of each section summary to give the reader a sense of what may or may not have changed. 2 An additional restriction imposed because of limited resources was that we did not review the large body of research that focused exclusively on the recruitment and retention of special education and vocational education teachers. We also restricted our review to research endeavors of a scholarly nature. As indicators of this, we looked to peer-reviewed journals and organizations with wellestablished peer-review processes as sources of publications. In addition, books, book chapters, and monographs that offered empirical evidence and analysis were included. 3 In addition, we considered only studies that were empirical in nature. By this, we mean that they offered evidence-quantitative or qualitative, or both-for conclusions, rather than simply opinion, theory, or principles. Although important theoretical work exists in the education literature, a discussion of this work is outGuarino et al. 178 side the scope of this review. We focused here on empirical work only to highlight and distill the reliable existing evidence relating to teacher recruitment and retention for researchers and policymakers alike. We therefore excluded simple program descriptions that were not analytical or evaluative and publications that offered only opinions, theory, or principles without offering new or original evidence to support conclusions. Thus literature reviews and publications that cited only research performed by others were excluded. Finally, we reviewed only those articles, books, and monographs that we considered to be rigorous in quality, according to generally accepted standards for quality in empirical research. A study was included in our review if its research design and analytic strategy were appropriate to the topic under study, its methodology was applied in a careful manner, its focus was relevant to our research questions, and its interpretation was well supported. Although the preponderance of empirical studies that we found on the subject of teacher recruitment and retention were statistical in nature, we included both quantitative and qualitative studies in our search and applied a similar baseline standard of rigor to each type of research-that is, a determination of the appropriateness and adequacy of the design, methods, and interpretation. Our quality criteria for the selection of quantitative studies were based on satisfactory answers to the following questions: (a) Did the sample adequately support the analyses performed? (b) Did the measures appear to be valid and reliable? (c) Did the researchers choose a statistical approach that appropriately modeled the phenomena under study? (d) Was the researchers' interpretation warranted by the findings? Our criteria for the selection of qualitative studies for review were based on satisfactory answers to the following questions: (a) Was the method justified because the study was characterized by a small sample size, data that were difficult to quantify, or phenomena for which no existing hypotheses had previously been developed? (b) Did the study offer sufficient evidence to support its conclusions? (c) Did the study reveal relationships between carefully identified variables that were of interest to other researchers in the field? If the answers to these questions tended to be in the affirmative, then we included the study in our review. To ensure the reliability of the application of our criteria, all of us looked at the studies under consideration to determine whether they met the quality criteria. Search Strategy and Results We performed electronic database searches of Education Abstracts, Social Science Abstracts, Econlit, ERIC, and JSTOR and table-of-contents searches on widely recognized education journals. We also searched publications indexes of education research institutions such as the National Center for Education Statistics. In addition, we asked a number of scholars for suggestions of relevant works to include in this study. These searches returned 4,919 unduplicated studies. Of these, 3,090 were on topics insufficiently relevant to this project; 1,562 were either nonempirical in nature or located in journals not widely regarded as high-quality scholarly publications even though they might be peer-reviewed; and 221 were of interest but were reviews of other work, early publications superseded by later work that we had included, not quite on target regarding our specific research questions, or rejected under our quality criteria as described in the previous section. Limiting studies to those performed on teacher labor markets in the United States and to those Teacher Recruitment and Retention 179 using data from the 1990s or later (and published by the end of 2004), left us with 46 studies to be included in this review. Caveats Our methodology carries with it some limitations. First, we searched for published documents; thus the findings listed in the review, as in most literature reviews, may be subject to "publication bias" because studies that supported the null hypothesis of "no effect" were less likely to be published or offered for publication. Second, many organizations exist that offer information of interest. The Internet alone offers a plethora of downloadable publications, some of which may contain valuable research findings, but our limited resources did not permit a search for these types of items. Last, in applying our quality criteria, we were able to distinguish studies that should be included from those that should not, but the studies that met the threshold for acceptance were by no means homogeneous in quality. We note important shortcomings in various studies when describing them, but we ask the reader to bear in mind that variations in quality exist among the studies included for review. Findings The Characteristics of Individuals Who Enter Teaching The first question addressed was: Who goes into teaching? Research to determine the characteristics of individuals who enter teaching can take three approaches: (a) It can compare those who choose teaching with those who do not, to ascertain the distinguishing characteristics or motivations of teachers; (b) it can simply examine the characteristics of individuals who choose teaching; or (c) it can examine the characteristics of individuals who choose to teach and are chosen to teach by school districts. The first approach would draw a sample from a population of potential applicants to teaching, the second type would draw from the population of actual applicants to teaching, and the third would sample only those who had applied and been accepted. In the group of studies that were included in our review, we found few that compared entrants to teaching with those who chose not to enter teaching, almost none that surveyed actual applicant pools, and several that focused on individuals who had already self-selected and been selected into teaching. The literature provided fairly consistent information regarding the demographic characteristics of entering teachers and the relationship between ability and the decision to teach. A very small number of studies provided evidence of psychological factors motivating individuals to enter teaching. Gender Women are more likely than men to enter teaching. Henke, Chen, Geis, and 180 hires to teaching were female in Race/Ethnicity Minority student enrollments have increased dramatically recent years. In 1991, 17% of public elementary and secondary school students were classified as Black, 12% as Hispanic, and 3% as Asian (U.S. Department of Education, 1993). In, 2000, 39% of public elementary and secondary school students were members of a minority group. Of these, 17% were classified as Hispanic, 17% were classified as Black, and 5% were members of other racial/ethnic groups (U.S. Department of Education, 2002). Kirby, Berends, and Naftel (1999) pointed out that the recruitment of minority individuals into the teaching profession did not keep pace with enrollment increases in the 1980s and early 1990s. Henke et al. Gitomer, Rong and Preissle (1997) analyzed 1990 census microdata representing 5% of the U.S. population and found that in 1990, Asian Americans constituted 2.8% of the U.S. labor force overall but only 1.2% of elementary and secondary teachers. Hispanics were also underrepresented in teaching-they made up 7.5% of the U.S. labor force overall but only 4.7% of elementary and secondary teachers. In contrast, African Americans constituted 9.8% of the U.S. labor force and 9.6% of elementary and secondary teachers. Gordon (1994) used interview data on 140 minority teachers in California in the early 1990s. These teachers expressed the opinion that students of color were Teacher Recruitment and Retention 181 discouraged from entering the teaching profession because they had had negative experiences in school, were concerned about a lack of student discipline and lack of respect in the classroom, perceived teaching to be low in status and pay, and perceived the image of a teacher to be a White, middle-class female. Because the author provides only limited information on the characteristics of the interviewees, however, it is difficult to assess the generalizability of these findings. Ability Four studies found that college graduates with the highest levels of measured ability tend not to go into teaching, and two of these studies found that this holds primarily for elementary school teachers rather than secondary school teachers. Podgursky, Monroe, and Watson (2004) compared the ACT scores of graduates from 4-year public higher education institutions in Missouri who became Missouri public school teachers in the year following graduation with those of graduates who did not enter teaching in the 1997-1998 and 1998-1999 years. They found that those who entered teaching had significantly lower scores than the nonteachers but that the gap was almost entirely due to the scores of elementary school teachers. They also found that while women at all test score levels were more likely than men to enter teaching, high-scoring women were relatively more reluctant to enter teaching than high-scoring men. Henke et al. The occasionally observed inverse relationship between measured ability and entry into teaching may be partially attributable to public district hiring practices. Although individuals can choose or not choose to enter the teaching profession, it is important to keep in mind that the hiring practices and preferences of districts and schools can influence the characteristics of teachers, particularly in regions or times in which there is a surplus of applicants. Two studies suggest that principals do not regard high academic ability as the main characteristic they look for in a teacher when hiring. Abernathy, Forsyth, and Mitchell Psychological and Family-Related Factors Three studies provided information on self-reported psychological and familyrelated factors that drew individuals into teaching. Teacher Recruitment and Retention 183 without going back to school, 54% said that they would want to teach students who were well behaved and eager to learn, and 47% said they would be more likely to consider teaching if it paid more. Because the research design was retrospective and the questions posed to teachers were different from those posed to nonteachers, it is difficult to draw direct comparisons between the two groups. Shipp (1999), using data from 263 questionnaires answered by African American college students at two universities in 1992, found that education majors placed significantly greater importance on the "contribution to society" factor than did non-education majors. Non-education majors placed more importance on the "salary," "job-security," "prestige," and "advancement opportunities" factors than did education majors. The author suggests that non-education majors were put off by a perceived lack of sufficient compensation and opportunity for promotion in teaching. These findings may not generalize to college students of other races or to African American students in other locations, however. King (1993) surveyed a small group of 41 prospective and beginning African American teachers from one institution of higher education in [1988][1989][1990] and found that intrinsic rewards, such as the opportunity to work with young people, the perception that their abilities were well matched to teaching, the belief that teaching contributed to the betterment of society, and the opportunity to be creative, outweighed considerations related to salary and prestige, according to self-reported perceptions. Salary, prestige, and autonomy, however, were cited as more important motivators for men than for women, whereas vacation time and the ability to couple teaching with child-rearing were cited as more important motivators for women than for men. Almost all respondents said that salary increases and better working conditions would lead to greater recruitment of minority teachers. Because of the small sample size, the fact that all subjects were drawn from one institution, and the absence of a comparison group, it is difficult to generalize these findings to the population of either all education students or all African American education students. Summary of Findings on Characteristics Predicting Entry to Teaching Results that arose fairly consistently in research regarding the characteristics of individuals who enter the teaching profession suggested that females formed greater proportions of new teachers than males and Whites formed greater proportions of new teachers than minorities, although minority participation appeared to have risen in the early 1990s. In addition, four studies suggested that college graduates with higher measured academic ability were less likely to enter teaching than other college graduates. It is possible, however, that such differences are driven by the measured ability of elementary school teachers, who represent the majority of teachers. Two studies suggested that academic ability was only one and not the most important of the characteristics valued by schools and districts in the hiring process. A more tentative finding based on a small number of methodologically weaker studies (e.g., with small sample sizes, no comparison group) is that an altruistic desire to serve society is one of the primary motivations for pursuing teaching. These findings concur, for the most part, with findings in the earlier empirical literature on entry to teaching, suggesting that few major changes have occurred over time. Studies conducted by The labor market supply-and-demand framework that was discussed in our conceptual framework suggested that individuals with higher opportunity costs in the form of attractive alternatives to teaching would be less likely than those with lower opportunity costs to enter the teaching profession. The empirical findings discussed above tend to fall in line with this hypothesis. Males have historically had a wider choice of jobs than women, and this imbalance may persist today. Alternatively, it may be that women continue to bear a greater share of child-rearing responsibilities than men and find teaching to be more compatible with these constraints, thus narrowing their choice set. The documented rise in minority participation may indicate increased accessibility to teaching positions and the attractiveness of these positions relative to other accessible positions. Individuals with higher ability are likely to have a wider range of job choices. High school teaching positions may offer more attractive extrinsic and intrinsic rewards than elementary school positions. The Characteristics of Individuals Who Remain in Teaching The second research question under consideration is the following: Who remains in teaching? The decision to continue teaching shares the same motivating principle that led to entry into teaching-namely, the perception that among all available alternate activities, teaching remains the most attractive in terms of compensation, working conditions, and intrinsic rewards. We found a large number of empirical studies devoted to answering this research question. This is because data on persistence in or attrition from the teaching profession are relatively plentiful. Many states keep extensive records of the movements of their public school teachers in and out of the profession and often make these data available to researchers, complete with the characteristics of the teachers and the schools they work in and identifiers that allow teachers to be followed over time. Some of the most notable of these data sets are collected in the states of New York, Texas, and Michigan. In addition, some nationally representative data sets, such as the Schools and Staffing Survey, contain longitudinal data on teachers. In addition to the relative abundance of data on this issue, there exist welldeveloped statistical models of attrition and persistence, many of which have arisen out of biomedical research that investigated the causes and incidence of mortality and illness. We found several studies that applied various types of hazard and duration models to the study of the duration of teaching careers. Although our conceptual framework emphasizes the process by which individuals form their decision to teach, one caveat to keep in mind is that not all teacher attrition is the result of choices made by the teachers themselves. Singer and Willet (1988) discuss the possibility of bias in studies that assume all attrition to be voluntary. It is also important to keep in mind that teacher turnover is not synonymous with overall attrition from teaching. More than half of all teacher turnover consists of migration from one school to another Teacher Recruitment and Retention 185 teacher turnover by studying movements of teachers across schools or districts. While this type of teacher turnover represents "attrition" from individual schools or districts, it does not represent overall attrition from teaching. We try to note, where appropriate, whether the study in question deals with teacher migration or mobility (movements from one school to another) or teacher attrition (leaving the occupation). In addition, it is important to consider that attrition from teaching need not be permanent and that many individuals return to teaching after a few years. Teaching Versus Other Occupations To place the issue of teacher retention in context, we report on two studies that compared rates of turnover in the teaching profession with those in other occupations. The studies suggested that the teaching profession may have somewhat lower retention rates than other occupations that employ college graduates, but they disagreed on results comparing specific occupations. Ingersoll Henke, Zahn, and Carroll Age and Experience One very stable finding is that attrition is high for young or new teachers and lower for older or more experienced teachers until they reach ages at which retirement is feasible. This phenomenon produces the well documented U-shaped plot of attrition against age or experience. Some of these studies-those using state public school or district-level databases-define entry and attrition as entering and leaving teaching in the particular state or district system under consideration. Although it is likely that teachers who entered or left the public school system left teaching altogether, it is possible that those teachers may have come from or migrated to other states or private schools or gone into other education positions, such as administration. Hanushek, 186 although the rates of early attrition declined somewhat with successive cohorts over the time period studied. In addition, they found that second-year attrition rates were lower for slightly older teachers. Gender Three studies found that women had higher attrition rates than men. Ingersoll (2001a), using the 1990-1991 Schools and Staffing Survey and 1991-1992 Teacher Follow-up Survey, found that male teachers were less likely to quit teaching than female teachers. Race/Ethnicity Another finding that emerged from several studies was that minority teachers tended to have lower attrition rates than White teachers. Ingersoll (2001a), using the 1991-1992 Schools and Staffing Survey and Teacher Follow-up Survey to investigate factors related to teacher attrition, found that minority teachers were less likely to quit teaching than White teachers. Ability, Field, and Qualifications As mentioned in our conceptual framework, attrition is generally costly to schools and may be detrimental to learning. If effective teachers are less likely to leave than less effective teachers, however, then high levels of teacher attrition may improve rather than decrease the overall quality of the teaching workforce. We reviewed several studies that investigated the relationship between attrition and measured academic ability or qualifications. The preponderance of evidence suggests that teachers with higher measured ability have a higher probability of leaving and that retention rates varied by level of education and field, as well. Teacher Recruitment and Retention 187 The following studies found that secondary teachers, particularly science teachers and sometimes math teachers, were more likely to leave than were elementary teachers.