CC BY-NC-ND
81
To Remain as a Teacher:
Factors Influencing Attitudes towards Leaving the Teaching Profession
K. Maslinsky, V. Ivaniushina
Received in June 2016
The research was supported by grants from the Russian Foundation for Humanities (N14-03-00582), the Basic Research Program (2014-2015)and the Academic Fund Program at the National Research University Higher School of Economics (N15-09-0269).
We are very grateful to Amy Krishnevsky for her friendly help with the proofreading of the paper.
Kirill Maslinsky
Research Fellow, Laboratory of Sociology in Education and Science, National Research University Higher School of Economics (Saint Petersburg). E-mail: kmas-linsky@hse.ru Valeria Ivaniushina
Candidate of Sciences, Leading Researcher, Laboratory of Sociology in Education and Science, National Research University Higher School of Economics (Saint Petersburg). E-mail: ivaniushina@ hse.ru
Address: 55 Sedova St, 192171 Saint Petersburg, Russian Federation.
Abstract. The present study examines structural and socio-psychological factors affecting attitudes towards quitting the profession among school teachers. We explore effects of perceived workplace difficulties, employment opportunities, self-ef-
ficacy beliefs, and emotional attachment to the teaching profession. The survey was conducted among public secondary school teachers in Saint Petersburg, Russia (N = 730). The regression analysis revealed that self-efficacy beliefs and professional commitment are the strongest predictors for retention, some work-related stress factors contribute to the likelihood of switching profession, while the number of years of teaching experience and work experience outside of teaching have no effect. The results do not support the hypothesis that early-career teachers are more tolerant to switching professions. The implications for retaining teachers in the profession are discussed. Keywords: teachers, employee attitudes, self efficacy, professional commitment, labor turnover
DOI: 10.17323/1814-9545-2016-4-8-30
Introduction The decision to choose a teaching career is not necessarily taken once and for life; it can be reversible. Giving up a teaching career and making a transition to another occupation is in many countries of the world most often attested among beginner teachers in the first five years of their school employment [Grissmer, Kirby 1997; Liu, Ramsey 2008]. But more experienced teachers sometimes take the decision to leave the profession, too, and economical considerations may be part of their decision [Baugh, Stone 1982; Harris, Adams 2007].
Attrition of qualified and active young teachers is a process that presumably affected the age structure of the population of school teachers in Post-Soviet Russia. In 2013, Russia participated in the OECD Teaching and Learning International Survey (TALIS). In this survey, the proportion of teachers older than 50 turned out to be
much higher (almost 40%) in Russia than in other participating countries, while at the same time the percentage of very young teachers is twice as high in Russia (4.7% of teachers younger than 25) as in other countries [Pinskaya et al. 2015]. Middle age teachers turned out to be the most diminished cohort in comparison with other countries. The question of whether or not young teachers will stay in the profession becomes essential to preventing the aging of teacher population in this situation and to develop a sustainable teacher recruitment and retention policy. The research of individual and organizational factors influencing the attitudes of teachers toward leaving the profession is required to obtain data necessary for the development of evidence-based policy in this area.
Teacher turnover is a topic of long-standing research interest occasionally supported by published data reporting critically high levels of attrition among teachers (see e. g., Ingersoll 2001). A wide range of institutional, economic, and organizational factors have been studied, that can serve as predictors of teacher turnover at schools. A prominent group of predictors that emerged in this research is a group of socio-psychological factors characterizing the attitudes of teachers toward their work and the profession at large: teachers' job satisfaction, individual and collective self-efficacy, and professional commitment [Ashton, Webb 1986; Goddard, Goddard 2001; Ingersoll 2001; Wang, Hall, Rahimi 2015]. However, most of the research on this topic was performed using data from North America and Europe. We were unable to find similar studies based on Russian data.
Our research is based on a sample of teachers of state secondary schools in Saint-Petersburg. Our goal is to study the interplay between the structural and socio-psychological factors in a teacher's decision to stay in school. We evaluate the significance and relative importance of individual-level predictors (type of education, work experience both within and outside of school), job-related stress factors (conflicts with colleagues, conflicts with administration, difficulties with students etc.), and two important socio-psychological constructs: self-efficacy and commitment to teaching. Policy implications are discussed along with the research results.
Turnover The act of quitting a job is generally treated as the final step ofa long process models psychological process for an employee involving job-related attitudes and decisions. The existing models explaining employee turnover are far from reaching any consensus on the details of the process, but agree on a general sequence of stages [Steel, Lounsbury 2009]. The commonly acknowledged path leading to resignation is structured along the lines of ecological reasoning from distal to proximal conditions. Factors external to an individual, such as the job market and organizational or family situation, constitute distal conditions forming the initial stage for the turnover process. The next stages of the pro-
cess are characterized by changes in job attitudes, eventually forming a psychological condition that was recently conceptualized as a state of withdrawal from the job. On the last stages of the process explicit intentions of quitting can be attested followed by eventual resignation [Hom et al. 2012].
In empirical studies intentions of quitting proved to be the most reliable predictor for the actual turnover [Carsten, Spector 1987]. The downside of this predictor is its rather weak explanatory power. The more specific intentions are measured (like, 'quit in December'), the more trivial their predictions become [Hanisch, Hulin, Roznowski 1998]. Thus to explore causes for turnover, concepts pertaining to earlier stages of turnover process models are more useful.
In a classical formulation by March and Simon (1958), two crucial conditions for turnover were given as an individual's perceived ease of movement out of the job and the perceived desirability of such movement. Since then, the attitudinal sphere is commonly seen as a key mediator between distal factors and behavior leading to actual turnover. Empirical studies have identified a substantial list of attitudinal constructs associated with the likelihood of quitting a job: burnout levels, work motivation, job satisfaction, self-efficacy, commitment to the workplace and the profession, and withdrawal cognitions, to name a few [Steel, Ovalle 1984; Griffeth, Hom, Gaertner 2000]. These constructs are usually seen as interrelated and sometimes overlap in terms of survey items used to measure them. The causal path leading from negative attitudes toward work to turnover need not be direct and uniform for all employees. Some theorists advocate a non-linear interpretation for the decision taking process leading to quitting a job [Lee, Mitchell 1994; Steel 2002]. For our work, we find Lee and Mitchell's notion of a script to be of most conceptual utility in this domain. A script refers to a pre-defined plan of action that becomes a potential behavioral response whenever some external events disturb the stability of one's attitudes to her job and trigger the turnover process [Lee, Mitchell 1994]. We regard the idea of switching one's occupation for higher pay to be appropriately described as a script in a sense, and close to Lee and Mitchell's. Evaluating attitudes to such a script among respondents is a good way to find the subgroup that is prone to switching professions due to experiencing difficulties.
Empirical data on The causes of teacher attrition is a long-respected topic in education-teacher turnover al studies [Guarino, Santibanez, Daley 2006]. In this section, we briefly summarize available data on teacher turnover in general, with a particular focus on individual predictors for teacher attrition.
Public interest in predicting teacher turnover is explained by the high practical importance of the problem. US data published by Ingersoll (2001) showed high rates of yearly loss of teachers by school districts, either to other districts or out of the profession [Ingersoll 2001].
Some other studies have shown that the overall teacher turnover rate actually is not higher [Harris, Adams 2007], or is even lower [Henke, Zahn 2001; Stinebrickner 2002] than in other comparable occupations. Still, much empirical research on teacher turnover has been published since 2000 both inside and outside the US.
At an individual level, turnover rates for teachers are strongly associated with years of teaching experience. In the empirical studies of US teachers, a characteristic U-shaped pattern of teacher attrition emerged. The highest turnover is attested among beginner teachers in their first years of service and among the pre-retirement age group [Grissmer, Kirby 1997; Liu, Ramsey 2008]. The high turnover among beginner teachers is explained mostly in terms of (un)successful adaptation to the profession, often with reference to professional identity formation [Cochran-Smith et al. 2012]. The high turnover in the pre-retirement group is best explained by the high pension-to-salary ratio for US teachers, which makes early retirement attractive [Harris, Adams 2007].
Many of those who left teaching at some point in their career return to the profession several years later. Grissmer et al. (1992) found that during the 1980s, 40% of teacher hires were comprised of teachers returning to the profession. Longitudinal data show that of all the teachers who quit, approximately every third teacher returns to teaching at some point during the next five years. This figure is the same both for teachers who left the work force for child-care and for those who left for other occupations [Stinebrickner 2002].
The wage level in alternative occupations available for teachers is an obvious predictor for turnover. Indeed, in making decisions about quitting, teachers were found to be at least as responsive as other workers to wage differentials between teaching and other occupations [Baugh, Stone 1982]. An absolute wage level within and outside of teaching is also significant for career decisions. Higher opportunity wages outside teaching make teachers more likely to switch professions, and higher wages in teaching make a teacher less likely to quit either for career or family reasons [Dolton, Van der Klaauw 1999]. Paradoxically, this doesn't mean that former teachers earn more in other occupations. UK data show that those teachers who actually leave for other occupations in fact work for a 22% lower hourly-wage, work on average two hours longer, and work mainly in the nonprofessional occupations within the public sector [Frijters, Shields, Price 2004].
Individual characteristics that are linked to higher opportunities in the labor market like ability, qualification, and field of specialization may predict resignations from the profession. Teachers who score higher on standardized tests, graduated from more selective universities and specialize in chemistry or physics stay in teaching for a shorter length of time [Murnane, Olsen 1990]. That means that in the case of a large enough wage differential, the teaching profession is bound to lose the most qualified individuals to other occupations. But that
doesn't necessarily mean that it will lose its best teachers. A teacher's performance is usually measured by the educational attainment of her students on standardized tests. Judging by this measure, it is the least effective teachers who tend to exit the school district after a couple of years of teaching [Murnane 1984]. A more recent study confirmed that better female teachers tend to stay in the profession but there was no association between teaching performance and turnover for male teachers [Krieg 2006].
Attitudinal variables were also studied as predictors for teacher turnover. Here we consider the two psychological constructs most closely related to the idea of leaving the teaching profession: self-efficacy beliefs and occupational commitment. Both these constructs were extensively studied in application to teachers.
For teachers, professional self-efficacy means a belief by a teacher in her capability to manage her students and teach them something. The concept of self-efficacy was introduced by Bandura (1977) and was quickly shown to be a useful predictor of a teacher's classroom performance, academic performance of her students and overall job satisfaction [Ashton, Webb 1986; Pajares 1997; Canrinus et al. 2012]. Besides that, in some studies it was demonstrated that teachers with higher self-efficacy are more likely to stay in teaching [Glickman, Tamashiro 1982; Swanson 2012; Vieluf, Kunter, van der Vijver 2013].
Occupational commitment is a construct that characterizes the degree of one's attachment to a career role (in our case, a teacher) and willingness to work in it [Chapman 1983]. It is theoretically and empirically distinguishable from commitment to a particular organization [Hackett, Lapierre, Hausdorf 2001]. Commitment has received much attention in personnel turnover studies because it has been found to strongly predict intentions to quit [Martin 1982]. For teachers, occupational commitment is a predictor not only of teacher attrition, but also of teacher performance and burnout [Tsui, Cheng 1999; Rots, Aelterman 2008]. Notably, there exists a strong positive association between a teacher's self-efficacy and occupational commitment [Klassen, Chiu 2011; Chan et al. 2008; Canrinus et al. 2012].
In line with the seminal model of March and Simon (1958), we may structure the evidence on teacher turnover predictors along the lines of the ease and desirability of a move. Higher ability and better qualifications contribute to the mobility ease while making more desirable alternatives available. Low self-efficacy and low commitment levels contribute to the desirability of leaving the profession. Job satisfaction which is related to workplace difficulties and an individual's reaction to these difficulties makes an additional contribution to the desirability of resigning.
Russian context Retention of teachers in the profession was a pressing social issue in Russia in the 1990s. The fast transition to the market economy accom-
panied with the relative neglect of educational policy on the part of the state was a scene for a growing teacher turnover rate and decreasing professional status. The situation was discussed by the teachers themselves, researchers and the public at large primarily in economic terms [Gimpelson, Treisman 2002]. At that moment, a severe cut in the state funding of education led to a drop in teacher salaries to the point where many teachers were forced to leave for better paid skilled and unskilled occupations. The teacher turnover rate of the 1990s had a lasting impact on the demography of Russian school teachers, forming what is known to be the effect of 'aging' the teacher population, since most leavers were early- and mid-career teachers [Rzhanitsyna 2000; Zajda 2003]. The mean age of Russian school teachers in 2013 was still higher than the cross-national average [Pinskaya et al. 2015].
The past ten years have seen changes in the economic situation along with significant policy modifications and finally an increase in the proportion of young teachers currently employed in schools [Pinskaya et al. 2015]. Following the economic line of reasoning, we may suggest that a more competitive salary was a factor in recovering the attractiveness of the teaching profession. Nonetheless, the issue of teachers' salary levels remains a publicly discussed problem [Zaba-turina, Kovaliova 2010; Zijatdinova 2010].
Optimism about the current trend in teacher recruitment should be constrained with the evaluation of the prospects to retain the most able younger teachers at their posts in the long-term perspective. Some surveys show that the proportion of teachers who were considering leaving the profession as a career option is rather large. In a mid-1990s survey of teachers in Saint Petersburg, 22% of respondents reported that they could quit the job [Tumalev 1995]. In the latest TALIS survey with a nationally representative sample of teachers, 33% of respondents younger than 40 and 18% of those older than 40 reported that they wouldn't choose this occupation if they started their career anew [Pinskaya et al. 2015]. We are unaware of any empirical studies that have investigated factors influencing such a sentiment toward the profession among teachers in the Russian context.
The high rate of job leaving intentions (even as hypothetical as an answer to a questionnaire) is more noteworthy given the changes in the labor market and the career strategies that have occurred since the 1990s. Evidence exists that the younger generation of workers tend to perceive their career not as a life-long choice but as a more fluid experience, and are more eager to change occupations [Wise, Mill-ward 2005], and teachers are no exception [Smethem 2007].
Research The goal of our work is to explore the factors that make a teacher more questions tolerant to the idea of leaving teaching for an occupation with a higher wage. We test for the significance and relative importance of three groups of factors:
1. Factors pertaining to the individual's employment opportunities outside of teaching. Here we consider type of education (higher education degree not in pedagogy or specialized pedagogical education), as well as work experience within and outside of teaching.
2. Job-related stress factors. In this group we consider workplace difficulties that respondents associate with the idea of leaving the profession: conflicts with colleagues, conflicts with administration, difficulties with students, excessive work load, inadequate salary.
3. Attitudinal factors. We consider teacher's self-efficacy (a perceived ability to manage students and to teach them something) and an individual's emotional attachment to the teaching profession. We measure emotional attachment to the profession using a scale of occupational commitment.
Data and The empirical results are based on data collected in 2014 on a sam-method ple of Saint Petersburg schools. Two city districts were chosen for the study, one in the central part of the city and another on the periphery. Random sampling of schools was performed in each district. The proportions of gymnasiums1 and schools with specializations in certain subjects matched the proportions in the general population2. Altogether, we surveyed 39 schools and 769 teachers. All the schools in the sample agreed to participate in the study. The survey procedure was conducted as follows. The research goal was explained to the school administration, after that each school received 20-30 questionnaires, depending on the school size, for teachers to complete themselves. These questionnaires were anonymous. In each school 11 to 30 teachers were surveyed (20 per school on average). Teachers filled in the written questionnaires after a professional meeting at school; completion of the survey was voluntary. Only teachers who work in secondary/high school were included in the study. 91% of participants were female, with a mean age of 45 years old and mean professional experience of 19 years.
The comparison of gender and age distribution in our sample with the data on the general population of Russian urban teachers for the year 2013 published by the Ministry of Education and Science [Ministry of education 2016] shows no significant differences in the distribution of gender and age categories used in the Ministry's data, whereas the comparison of the number of years of working experience shows
1 Gymnasium—is a type of secondary school in modern Russia with an emphasis on academic subjects. The name 'gymnasium' refers to a name of type of secondary schools in Russian Empire the diploma of which allowed the admission to the university.
2 More details on the sample and the descriptive statistics can be found in [Ivaniushina, Alexandrov 2016].
that our sample has somewhat more teachers with less than 20 years of professional experience. This may be explained by our sampling strategy, as we did not include in the study teachers of special education schools, boarding and elementary schools. Besides that, it is possible that young and less experienced teachers are less likely to opt out of the survey. Hence, our sample may be considered generally representative of the chosen segment of Saint-Petersburg schools. Possible over-representation of younger teachers in our sample is consistent with the aim of the study—to analyze the risk of teacher attrition, which is more prominent in the case of younger teachers.
Measures Dependent variable. The dependent variable for regression models was constructed from the answer to the question: 'If you were offered a job with a more attractive salary, but in a field unrelated to school, how difficult would it be for you to quit the teaching profession?', with a response scale from 1 (very easy) to 5 (very difficult).
Characteristics of the employment situation. We included several characteristics that might contribute to teachers' perception of the ease of leaving the profession. The first is a special type of education. All the teachers in our sample had higher education. But we distinguished teachers who graduated from a pedagogical institute/university from others (a binary variable). The second is the duration of job experience measured as the number of years teaching. The third variable was job experience outside of school, that is whether a respondent had ever had a non-school related job (a binary variable). The fourth variable was whether a respondent was engaged in private tutoring (a binary variable).
Professional Difficulties. The next group of measures was constructed from the replies to the question 'What problems and difficulties could make you leave your job?' The following possible answers were offered: 1) conflicts with administration; 2) conflicts with colleagues; 3) difficulties with students; 4) too much work load; 5) not enough payment; 6) lack of opportunity for personal growth and development. Respondents had to mark 'yes' or 'no' for every answer; thus six binary variables were constructed.
Self-efficacy. This construct was measured using a 5-item teacher efficacy scale (Gibson, Dembo 1984), translated into Russian and slightly adapted. Sample items are 'If I try hard, I can motivate even the most uninvolved students' and 'If a student in my class becomes disruptive and noisy, I feel assured that I know some techniques to redirect him.' A Likert-type response scale with four categories was used, ranging from 1 ('completely disagree') to 4 ('completely agree'). Previous research confirmed adequate internal consistency and a single-factor structure of the scale (R. D. Goddard and Goddard 2001). On our sample Cronbach's alpha was 0.72. The items were averaged, with higher scores indicating a greater level of self-efficacy.
Occupational Commitment. For measuring this construct we designed our own 5-item scale. Sample items are 'I would not like any other profession except school teacher for myself' and 'despite all the difficulties and problems, I like my profession'. Items were rated on a 4-point metric, ranging from 1 ('completely disagree') to 4 ('completely agree'). Analysis has shown that all items are combined in a single factor. The reliability coefficient (Cronbach's alpha) is 0.74. The items were averaged, with appropriate reverse scoring. Higher scores indicate greater commitment.
The gender of the respondent was included in all models as a control variable.
Analytical strategy For the analysis of relations between the dependent variable—how hard it would be for a teacher to quit her job—and a number of explanatory variables, we used a series of multiple regression models3, built using SPSS19.0 software. The logic of the analysis was as follows. The first model included explanatory variables describing education, job experience and gender. The second model included variables describing professional difficulties and self-efficacy. In the third model, we combined variables from the first and second models. Our last step (model 4) was to include professional commitment as an explanatory variable.
Results Descriptive statistics for the variables are presented in Table 1. Most of the teachers in our sample had special pedagogical education, meaning they graduated from a pedagogical institute or university. Nevertheless, almost 60% of the respondents at some point of their lives had held a job outside of school and teaching. We might suggest that this happened mostly in 1990-2000, when a school teacher's salary was quite low, and teachers either had to supplement it by having an additional job, or had to leave school altogether for a while. Of the total sample, 20% are young teachers with experience of 5 years or less. Almost the same proportion are teachers with more than 30 years of experience; that is, persons close to retirement or already over retirement age. Private tutoring is quite common: almost every 6th teacher earns additional money by giving private lessons.
Regression models results are presented in Table 2. In the 1st model we introduced only the variables related to job experience and education. There is a positive relationship between the length of teaching experience and the difficulty in leaving the profession, but this relationship is very weak. If a teacher had ever hadanother job
3 Though the dependent variable is negatively skewed, residuals of the regression with the original variable are normally distributed; this means that variable transformation is not necessary.
Table 1. Descriptive statistics
Variable name % Mean (SD) Min-Max
Women 92%
Teaching experience (in years) 18,8 (12,1)
Higher pedagogical education 67%
Ever had a job not in school 57%
Engaged in private tutoring 15%
Conflicts with administration 051 (0,50)
Conflicts with colleagues 0,27 (0,44)
Difficulties with students 0,17 (0,37)
Too much work load 0,22 (0,41)
Not enough payment 0,57 (0,50)
No possibilities for personal growth and development 0,40 (0,49)
Self-efficacy 3,00 (0,44) 1,4-4,0
Commitment to teaching 2,84 (0,65) 1-4
How hard to leave teaching profession 3,85 (1,20) 1-5
outside of school, or if she was engaged in private tutoring, it would be easier for her to quit the teaching profession. This model has very small explanatory power and explains only 3.6% of the variance in the dependent variable.
In the second model we analyze how different problems and difficulties that teachers experience in their work affect their unwillingness to leave the teaching profession. We also include self-efficacy in this model as a concept reflecting teachers' perceived expectation of succeeding in their everyday tasks. As it turned out, not all factors that are perceived by teachers as work difficulties are related to their attitude toward leaving the profession. Out of six possible stress factors, only problems with students and low payment, and (to a lesser degree) excessive work load increase the perceived ease of leaving the teaching profession. On the contrary, conflicts with colleagues and administration and lack of opportunities for personal growth and development do not make it easier for a teacher to quit her job. Self-efficacy is positively related to the dependent variable, that is, the higher one evaluates herself on this parameter, the more difficult it would be for her to leave the teaching profession. It is worth noting that R-squared for Model 2 is much larger than for Model 1 and equals 16.4%.
The third model combined Model 1 and Model 2. Combining two models gives an additional slight increasein R-squared—18.3%. Signs of all the coefficients remain unchanged, while the significance of
Table 2. Regression modeling result.
Dependet variable: how hard it would be to leave teaching profession
Model 1 b (S.E.) Model 2 Model 3 Model 4
Constant 3,53 (0,20)*** 2,82 (0,30)*** 4,09 (0,21)*** 1,14 (0,37)**
Higher pedagogical education 0,17 (0,10) 0,22 (0,10)* 0,04 (0,09)
Teaching experience (in years) 0,010 (0,004)* 0,004 (0,004) 0,000 (0,004)
Ever had a job not in school -0,19 (0,10) -0,20 (0,10)* -0,03 (0,09)
Engaged in private tutoring -0,30 (0,13)* -0,19 (0,13) -0,13 (0,11)
Gender 0,20 (0,17) 0,21 (0,18) 0,27 (0,16) 0,17 (0,15)
Conflicts with administration -0,17 (0,10) -0,17 (0,10) -0,10 (0,09)
Conflicts with colleagues -0,04 (0,11) -0,05 (0,11) -0,10 (0,11)
Difficulties with students -0,62 (0,12)*** -0,61 (0,12)*** -0,31 (0,11)**
Excessive work load -0,19 (0,10) -0,23 (0,11)* -0,06 (0,10)
Not enough payment -0,54 (0,09)*** -0,52 (0,09)*** -0,39 (0,09)***
No possibilities for development -0,02 (0,09) -0,06 (0,10) -0,03 (0,09)
Self-efficacy 0,52 (0,10)*** 0,51 (0,10)*** 0,26 (0,10)**
Commitment to profession 0,77 (0,07)***
R-square 3,6% 16,4% 18,3% 31,7%
*** p < 0,001; ** p < 0,01; * p < 0,05.
some coefficients change (when it approaches the threshold value of p = 0.05).
Our final model (Model 4) includes one additional variable—commitment to the teaching profession. The model has much more explanatory power and explains almost 32% of the variance in the dependent variable. Comparison of the coefficients between the third and fourth models shows that while self-efficacy and job-related difficulties that were significant in Model 3 retain their significance, all variables related to education and job experience become insignificant. The final model demonstrates that commitment to the profession is the strongest predictor that a teacher finds it extremely difficult to quit her job, when other demographic and socio-psychological factors are controlled for.
Discussion We estimated the relationship of the three groups of factors (individual employment opportunities outside of school, job-related stress factorsand attitudinal factors) with the perceived ease of leaving the profession. Our results show that most teachers find it hard to leave
the profession even when considering a hypothetical alternative with a higher wage. We found the two psychological constructs—occupational commitment and self-efficacy beliefs—to have the most prominent effect. The higher the efficacy and commitment, the harder it is for a teacher to consider leaving the profession. When these attitudinal factors are taken into account, other factors that we see as indicative of the opportunity for an individual's mobility in the labor market lose their significance. At the same time, some job-related stress factors retain their significance even when self-efficacy and professional commitment are taken into account. These factors are: difficulties with students, not enough payment and excessive workload, though their relationship with the attitude toward leaving profession is much weaker.
The design of our study does not allow us to draw conclusions about the direction of causality between self-efficacy and professional commitment on one side, and the ease in the attitude toward switching professions. Most probably, there is an interactional relationship among these attitudes during the teaching career.
Since there is no similar research based on Russian data, we can only compare our results to broader international literature. In a recent study, the internal structure of the Dutch teacher's professional identity was explored [Canrinus et al. 2012]. One of the factors identified by the authors, 'responsibility to remain in teaching,' is conceptually very close to our outcome variable. Among the most significant direct and indirect effects on it were classroom self-efficacy and emotional commitment to the profession, which directly corresponds to our findings.
One of our negative results is the most interesting from a theoretical and also a practical point of view. Our data do not support the hypothesis that early career teachers are more easily inclined toward the idea of switching professions. Some previous research has persuasively shown that the early years of teaching are crucial in making the decision to quit the profession, with a tipping point of 5 years of teaching experience [Struyven, Vanthournout 2014]. Authors argue that in the first years of teaching a teacher is still evaluating the prospects of an alternative career and can more easily decide to switch professions. Out results demonstrate that attitudes toward quitting the profession are independent of teaching experience4.
Our data confirm that effects of the work experience are mediated by the self-efficacy (which grows with experience) and occupational commitment. Such a mediational relationship is in line with the findings in similar studies on personnel attrition [Karsh, Booske, Sainfort 2005]. In theoretical models of the personnel turnover process, at-
4 We have built models (not shown in this article) that compared young teachers with less than five years of teaching experience with more experienced teachers. These models also show no difference in attitude toward leaving the profession among teachers with different experience.
titude toward the job is also treated as a mediator between external variables and events and turnover intentions [Steel, Lounsbury 2009].
An important limitation of our study is that our sample represents only urban teachers, moreover teachers from urban schools in a megalopolis. Thus our findings may not be easily generalized to the entire population of Russian teachers, many of whom work in small towns and in village schools. It should also be noted that despite proven correlation between intentions and actual quitting behavior, populations of employees who intend to quit and those who actually quit the job may differ in demographic characteristics [Cho, Lewis 2012]. Hence, our results could not be generalized as an assessment of the group of teachers who will eventually leave the occupation, but it is still a useful estimate for the parameters of that population.
Summing up the contribution of this study, we wish to highlight two points. First, we present empirical data on predictors of teacher attrition in a population of Russian teachers, previously unstudied in this respect. Our findings indicate that those who regard higher wages as a motivator to quit teaching are likely to have low professional self-efficacy and low occupational commitment. Second, our results do not confirm that early career teachers are more tolerant to the idea of switching professions, contrary to expectations based on high turnover rates among young teachers evidenced in research on US teachers. This means that there are good prospects of retaining young teachers in the long run, provided that they keep a positive attitude toward the profession. Although occupational commitment is hardly likely to be influenced by policy measures directly, self-efficacy can be. Thus we suggest that programs aimed at improving the classroom self-efficacy of young teachers may be an effective policy measure for teacher retention.
References Ashton P. T., Webb R. B. (1986) Making a Difference: Teachers' Sense of Efficacy and Student Achievement. New York: Longman. Bandura A. (1977) Self-Efficacy: Toward a Unifying Theory of Behavioral Change.
Psychological Review, vol. 84, no 2, pp. 191-215. Baron R. M., Kenny D. A. (1986) The Moderator-Mediator Variable Distinction in Social Psychological Research: Conceptual, Strategic, and Statistical Considerations. Journal of Personality and Social Psychology, vol. 51, no 6, pp. 1173-1182.
Baugh W. H., Stone J. A. (1982) Mobility and Wage Equilibration in the Educator Labor Market. Economics of Education Review, vol. 2, no 3, pp. 253-274. Canrinus E. T., Helms-Lorenz M., Beijaard D., Buitink J., Hofman A. (2012) Self-Efficacy, Job Satisfaction, Motivation and Commitment: Exploring the Relationships between Indicators of Teachers' Professional Identity. European Journal of Psychology of Education, vol. 27, no 1, pp. 115-132. Carsten J. M., Spector P. E. (1987) Unemployment, Job Satisfaction, and Employee Turnover: A Meta-Analytic Test of the Muchinsky Model. Journal of Applied Psychology, vol. 72, no 3, pp. 374-381. Chapman D. W. (1983) A Model of the Influences on Teacher Retention. Journal of Teacher Education, vol. 34, no 5, pp. 43-49.
Chan W.—Y., Lau S., Nie Y., Lim S., Hogan D. (2008) Organizational and Personal Predictors of Teacher Commitment: The Mediating Role of Teacher Efficacy and Identification with School. American Educational Research Journal, vol. 45, no 3, pp. 597-630.
Cho Y. J., Lewis G. B. (2012) Turnover Intention and Turnover Behavior Implications for Retaining Federal Employees. Review of Public Personnel Administration, vol. 32, no 1, pp. 4-23.
Cochran-Smith M. et al. (2012) A Longitudinal Study of Teaching Practice and Early Career Decisions A Cautionary Tale. American Educational Research Journal, vol. 49, no 5, pp. 844-880.
Dolton P., van der Klaauw W. (1999) The Turnover of Teachers: A Competing Risks Explanation. Review of Economics and Statistics, vol. 81, no 3, pp. 543-550.
Frijters P., Shields M. A., Price S. W. (2004) To Teach or not to Teach? Panel Data Evidence on the Quitting Decision. IZA Discussion Paper No 1164.
Gibson S., Dembo M. H. (1984) Teacher Efficacy: A Construct Validation. Journal of Educational Psychology, vol. 76, no 4, pp. 569-582.
Gimpelson V., Treisman D. (2002) Fiscal Games and Public Employment: A Theory with Evidence from Russia. World Politics, vol. 54, no 2, pp. 145-183.
Glickman C. D., Tamashiro R. T. (1982) A Comparison of First-Year, Fifth-Year, and Former Teachers on Efficacy, Ego Development, and Problem Solving. Psychology in the Schools, vol. 19, no 4, pp. 558-562.
Goddard R. D., Goddard Y. L. (2001) A Multilevel Analysis of the Relationship between Teacher and Collective Efficacy in Urban Schools. Teaching and Teacher Education, vol. 17, no 7, pp. 807-818.
Griffeth R. W., Hom P. W., Gaertner S. (2000) A Meta-Analysis of Antecedents and Correlates of Employee Turnover: Update, Moderator Tests, and Research Implications for the Next Millennium. Journal of Management, vol. 26, no 3, pp. 463-488.
Grissmer D., Kirby S. (1997) Teacher Turnover and Teacher Quality. The Teachers College Record, vol. 99, no 1, pp. 45-56.
Grissmer D., Kirby S. N., Schlegel P. M., Young R. (1992) Patterns of Attrition among Indiana Teachers, 1965-1987 Endowment. Santa Monica: Rand.
Guarino C. M., Santibanez L., Daley G. A. (2006) Teacher Recruitment and Retention: A Review of the Recent Empirical Literature. Review of Educational Research, vol. 76, no 2, pp. 173-208.
Hackett R. D., Lapierre L. M., Hausdorf P. A. (2001) Understanding the Links between Work Commitment Constructs. Journal of Vocational Behavior, vol. 58, no 3, pp. 392-413.
Hanisch K. A., Hulin C. L., Roznowski M. (1998) The Importance of Individuals' Repertoires of Behaviors: The Scientific Appropriateness of Studying Multiple Behaviors and General Attitudes. Journal of Organizational Behavior, vol. 19, no 5, pp. 463-480.
Harris D. N., Adams S. J. (2007) Understanding the Level and Causes of Teacher Turnover: A Comparison with Other Professions. Economics of Education Review, vol. 26, no 3, pp. 325-337.
Henke R. R., Zahn L. (2001) Attrition of New Teachers among Recent College Graduates: Comparing Occupational Stability among 1992-93 Graduates Who Taught and Those Who Worked in Other Occupations. Education Statistics Quarterly, vol. 3, no 2, pp. 69-76.
Hom P. W., Mitchell T. R., Lee T. W., Griffeth R. W. (2012) Reviewing Employee Turnover: Focusing on Proximal Withdrawal States and an Expanded Criterion. Psychological Bulletin, vol. 138, no 5, pp. 831-858.
Ingersoll R. M. (2001) Teacher Turnover and Teacher Shortages: An Organizational Analysis. American Educational Research Journal, vol. 38, no 3, pp. 499-534.
Ivaniushina V. A., Alexandrov D. A. (2016) Sushchestvuet li differentsiatsiya uchiteley v rossiyskikh shkolakh? [Is There a Differentiation of Teachers in Russian Schools?]. Sociological Studies, no 9, pp. 59-65.
Karsh B., Booske B. C., Sainfort F. (2005) Job and Organizational Determinants of Nursing Home Employee Commitment, Job Satisfaction and Intent to Turnover. Ergonomics, vol. 48, no 10, pp. 1260-1281.
Klassen R. M., Chiu M. M. (2011) The Occupational Commitment and Intention to Quit of Practicing and Preservice Teachers: Influence of Self-Efficacy, Job Stress, and Teaching Context. Contemporary Educational Psychology, vol. 36, no 2, pp. 114-129.
Krieg J. M. (2006) Teacher Quality and Attrition. Economics of Education Review, vol. 25, no 1, pp. 13-27.
Lee T. W., Mitchell T. R. (1994) An Alternative Approach: The Unfolding Model of Voluntary Employee Turnover. Academy of Management Review, vol. 19, no 1, pp. 51-89.
Liu X. S., Ramsey J. (2008) Teachers' Job Satisfaction: Analyses of the Teacher Follow-Up Survey in the United States for 2000-2001. Teaching and Teacher Education, vol. 24, no 5, pp. 1173-1184.
March J. G., Simon H. A. (1958) Organizations. New York: Wiley.
Martin T. N. (1982) Commitment Predictors of Nursing Personnel's Intent to Leave. Medical Care, vol. 20, no 11, pp. 1147-1153.
Murnane R. J. (1984) Selection and Survival in the Teacher Labor Market. The Review of Economics and Statistics, vol. 66, no 3, pp. 513-518.
Murnane R. J., Olsen R. J. (1990) The Effects of Salaries and Opportunity Costs on Length of Stay in Teaching: Evidence from North Carolina. Journal of Human Resources, vol. 25, no 1, pp. 106-124.
Pajares F. (1997) Current Directions in Self-Efficacy Research. Advances in Motivation and Achievement, vol. 10, no 149, pp. 1-49.
Pinskaya M., Kosaretskiy S., Ponomareva A., Brun I., Savelyeva M., Lenskaya E., Zavertiaev K. (2015) Rossiyskie pedagogi v zerkale mezhdunarodnogo sravnitelnogo issledovaniya pedagogicheskogo korpusa (TALIS2013) [Russian Pedagogues in the Mirror of the International Comparative Research of Pedagogical Corpus (TALIS2013)]. Moscow: National Research University Higher School of Economics.
Rots I., Aelterman A. (2008) Two Profiles of Teacher Education Graduates: A Discriminant Analysis of Teaching Commitment. European Educational Research Journal, vol. 7, no 4, pp. 523-534.
Rzhanitsyna L. (2000) Working Women in Russia at the End of the 1990s. Problems of Economic Transition, vol. 43, no 7, pp. 56-67.
Smethem L. (2007) Retention and Intention in Teaching Careers: Will the New Generation Stay? Teachers and Teaching: Theory and Practice, vol. 13, no 5, pp. 465-480.
Steel R. P. (2002) Turnover Theory at the Empirical Interface: Problems of Fit and Function. Academy of Management Review, vol. 27, no 3, pp. 346-360.
Steel R. P., Lounsbury J. W. (2009) Turnover Process Models: Review and Synthesis of a Conceptual Literature. Human Resource Management Review, vol. 19, no 4, pp. 271-282.
Steel R. P., Ovalle N. K. (1984) A Review and Meta-Analysis of Research on the Relationship between Behavioral Intentions and Employee Turnover. Journal of Applied Psychology, vol. 69, no 4, pp. 673-686.
Stinebrickner T. R. (2002) An Analysis of Occupational Change and Departure from the Labor Force: Evidence of the Reasons that Teachers Leave. Journal of Human Resources, vol. 37, no 1, pp. 192-216.
Struyven K., Vanthournout G. (2014) Teachers' Exit Decisions: An Investigation into the Reasons why Newly Qualified Teachers Fail to Enter the Teaching
Profession or Why Those Who Do Enter Do not Continue Teaching. Teaching and Teacher Education, vol. 43, pp. 37-45.
Swanson P. (2012) Second/Foreign Language Teacher Efficacy and Its Relationship to Professional Attrition. Canadian Modern Language Review, vol. 68, no 1, pp. 78-101.
The Ministry of Education and Science of the Russian Federation (2016) Sve-deniya o chislennosti i sostave rabotnikov uchrezhdeniy, realizuyushchikh programmy obshchego obrazovaniya, v gorodskikh poseleniyakh i selskoy mestnosti [Data on Number and Composition of Workers of Institutions Implementing Programs of General Education in Urban and Rural Areas]. Available at: http://opendata.mon.gov.ru/opendata/7710539135-83rikA (accessed 10 October 2016).
Tsui K. T., Cheng Y. C. (1999) School Organizational Health and Teacher Commitment: A Contingency Study with Multi-Level Analysis. Educational Research and Evaluation, vol. 5, no 3, pp. 249-268.
Tumalev V. V. (1995) Uchitelstvo v situatsii sotsialno-politicheskikh peremen [Teachers in a Situation of Socio-Political Change]. Vol. 1-5. Saint Petersburg: Saint Petersburg University of Economics and Finance.
Vieluf S., Kunter M., van de Vijver F. J. (2013) Teacher Self-Efficacy in Cross-National Perspective. Teaching and Teacher Education, vol. 35, pp. 92-103.
Wang H., Hall N. C., Rahimi S. (2015) Self-Efficacy and Causal Attributions in Teachers: Effects on Burnout, Job Satisfaction, Illness, and Quitting Intentions. Teaching and Teacher Education, vol. 47, pp. 120-130.
Wise A. J., Millward L. J. (2005) The Experiences of Voluntary Career Change in 30-Somethings and Implications for Guidance. Career Development International, vol. 10, no 5, pp. 400-417.
Zabaturina I. Y., Kovaliova N. V. (2010) Chto my znaem o rossiyskom uchitele? [What Do We Know of a Russian Teacher?] Narodnoe obrazovanie, no 10, pp. 36-42.
Zajda J. (2003) Educational Reform and Transformation in Russia: Why Education Reforms Fail. European Education, vol. 35, no 1, pp. 58-88.
Zijatdinova F. G. (2010) Sotsialnoe polozhenie uchiteley: ozhidaniya i realii [Social Status of Teachers: Expectations and Realities]. Sociological Studies, no 10, pp. 100-107.
