 

 

 

 
 

 
 

 
 

 

 

'IIIIIIIIIIIIIBIII

 

I?“ U ‘1

Ni
9.:E

This is to certify that the
dissertation entitled

WHERE TO TEACH? DEVELOPING A MORE
COMPREHENSIVE FRAMEWORK TO UNDERSTAND
TEACHERS” CAREER DECISIONS

presented by

MARISA CANNATA

has been accepted towards fulﬁllment
of the requirements for the

PhD. degree in Educational Policy

 

 

 

Major Professor’s Signature
m M} I) 200‘}

 

Date

MSU is an afﬁnnative-action, equal-opportunity employer

-.-----.-.-—.----.--.--a------—.-.—.----.-.-------.----c-----.-n-o-o-.-------.-n-n-.-i—.—._.-

 

 

M

LIBRARY

ici’iigan State

University

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

 

 

 

DATE DUE DATE DUE DATE DUE
JUN 2 0 2009
tin: ; {3:110
IOIEOQ SWZIL

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p'lClRC/DateDue.indd-p.1

 

WHERE TO TEACH? DEVELOPING A MORE COMPREHENSIVE FRAMEWORK
TO UNDERSTAND TEACHERS’ CAREER DECISIONS

By

Marisa Cannata

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Educational Policy

2007

ABSTRACT

WHERE TO TEACH? DEVELOPING A MORE COMPREHENSIVE F RAMWORK
TO UNDERSTAND TEACHERS’ CAREER DECISIONS

By

Marisa Cannata

The distribution of teachers is highly uneven, with schools in urban areas—those
likely to have high concentrations of poor and minority students—being the least likely to
have qualiﬁed teachers. Most of this inequitable distribution of qualiﬁed teachers is due
to the initial choices teachers make about where to work. This dissertation is a study of
the processes by which teachers come to work in particular schools and how these career
decisions relate to their social and cultural background. It uses Bourdieu’s cultural
reproduction model and the ideas of ﬁeld, capital, and habitus to understand the practical
sense teachers bring to the job search and how this inﬂuences how they perceive and
respond to available positions in schools. This dissertation uses a mixed-methods design;
interviews with 27 prospective elementary teachers during their job search are combined
with surveys of 289 prospective elementary teachers about their perceptions of the
teacher labor market, preferences for schools, and how they approach the job search.

Overall, this study shows a tendency for teachers to self-sort into schools that
serve students from similar social and cultural backgrounds as themselves. Some of these
forces are deliberate while others are more tacit—such as what teachers think it takes to
get a job, how they identify open positions, and how they perceive schools during
interviews. The ways that teachers approach the job search sometimes are inconsistent

with their espoused preferences for particular school characteristics. In particular, the

processes that teachers use when making decisions indicate that teachers seek out schools
where they feel comfortable and think they ﬁt. A sense of ﬁt is important in where
teachers think they can get a job and what they look for in a potential workplace.

This dissertation also presents evidence that teachers become sorted across
schools based upon both teacher candidate behavior and school and district hiring
practices. Teacher candidates’ race and socioeconomic background are related to the
student characteristics of the districts to which they apply. Schools and districts also have
varying preferences for the candidates they hire. Together, these forces lead to different
labor market outcomes for African American and White teacher candidates, teacher
candidates from different socioeconomic backgrounds, and mid-career changers and new
college graduates. In contrast to existing work on teachers’ career decisions, a cultural
reproduction framework illuminates this adjustment process as teachers end up in schools
where they ﬁt. A cultural reproduction framework thus expands our understanding of

teacher labor markets to include teachers as active participants in the process.

Copyright by
MARISA CANNATA
2007

ACKNOWLEDGMENTS

This dissertation would not have been possible without the help of the prospective
teachers who participated in this study. I want to thank them for their generosity in giving
me their time and thoughts about their job search. I hope I portrayed their experiences
faithﬁllly. I also want to thank the college and university instructors and personnel who
provided me with access to their elementary student teachers. Even if I take nothing else
away from this experience, I have learned that the best theoretical ideas cannot create an
empirical study. Without the help of these individuals, this study would have no data.

This study would also not be possible without the help of my dissertation director,
Dorothea Anagnostopoulos. While I can’t detail all that I have learned from you, I will
limit myself to a list: how to interview individuals in ways that elicit detailed responses
without leading them to particular responses, how to navigate the political waters of
academia, how to code and analyze qualitative data, the importance of positioning oneself
as a scholar, and how to prepare for a career in academia. Somewhere along the line, I
also learned a bit about cultural reproduction theory. Thank you for going above and
beyond what is expected of a teacher and mentor.

I also want to thank my advisor, Gary Sykes, for your guidance throughout this
process. You challenged me when I was complacent and supported me when I struggled.
You have served as an example of a scholar and mentor. I wanted to thank you in
particular for helping me secure access to some colleges to ﬁnish data collection.

I also want to thank my committee, David Arson, Peter Youngs, and Susanna

Loeb. Your feedback throughout the process pushed me empirically and theoretically by

forcing me to delve more deeply into the validity of my argument and more clearly
delineate how theory helps explains what is going on.

Writing a dissertation is usually a lonely process, but I was lucky enough to have
my writing group, Troy Hicks, Yongmei Ni, and Tara Donahue accompany me on this
journey. Sometimes I thought you knew what I was trying to say better than I did. Thank
you for helping me clarify my ideas and telling me when I didn’t really have any.

I want to thank my husband, Matt, for the countless ways you helped me
throughout the process: from reading drafts of my writing and coding interviews to
cooking me dinner so I could work and restoring my conﬁdence when it was failing.
Somehow you knew when I needed to work and when I needed to get away from my
work.

Finally, I want to thank the Spencer Foundation for supporting this research with
a dissertation fellowship. Perhaps even more than the ﬁnancial support, I value the

intellectual community and support provided through the workshops.

vi

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................. ix
LIST OF FIGURES ......................................................................................................... xiii
CHAPTER 1: INTRODUCTION ....................................................................................... 1
What We Know About Teachers’ Career Decisions .......................................... 4
Limitations of Existing Research on Teachers’ Career Decisions .................... 20
Cultural Reproduction Framework ................................................................... 27
Organization of the Dissertation ....................................................................... 34
CHAPTER 2: METHODOLOGY .................................................................................... 36
Survey Component ............................................................................................ 37
Interview Component ........................................................................................ 53
Advantages and Disadvantages of Research Design ........................................ 59
Teacher Supply and Demand in the Smithton Metropolitan Area .................... 61
CHAPTER 3: TEACHER CANDIDATES’ PERCEPTIONS OF THE STRUCTURE OF
THE TEACHING FIELD ................................................................................................. 71
Field and Capital ............................................................................................... 74
Previous Research on Teacher Hiring ............................................................... 75
Teacher Candidates’ Perceptions of the Teaching Field ................................... 79
Teacher Candidates’ Perceptions of the Field of Schools .............................. 101
Teacher Preference for Schools Occupying Weak Positions on the Field ...... 106
Discussion and Conclusion ............................................................................. 111

CHAPTER 4: UNDERSTANDING THE PRACTICAL LOGIC OF THE TEACHER

JOB SEARCH: ESPOUSED PREFERENCES AND PREFERENCES IN USE .......... 116
Theoretical Framework ................................................................................... 118
Espoused Preferences ...................................................................................... 123
Activating Practical Logic and Preferences in Use ......................................... 126
Discussion and Conclusion ............................................................................. 149

CHAPTER 5: SORTING AND SELF-SORTING: HOW PREFERENCES AND

HIRING PRACTICES RELATE TO THE DISTRIBUTION OF TEACHERS ............ 155
Steps in the Job Search ................................................................................... 159
Number of Applications .................................................................................. 161
Progression in the Job Search and the Distribution of Teachers .................... 164
Timing of Job Offers ....................................................................................... 184
Preferences and Labor Market Outcomes ....................................................... 186
Discussion and Conclusion ............................................................................. 196

vii

CHAPTER 6: THE DEMAND SIDE OF TEACHER HIRING .................................... 203

Overall Labor Market Outcomes .................................................................... 204
Number of Interviews ..................................................................................... 206
Which Teacher Candidates Received Job Offers? .......................................... 211
Hiring Decisions and Teacher Sorting ............................................................ 217
Candidates Who Did Not Get Job Offers ....................................................... 219
Discussion and Conclusion ............................................................................. 220
CHAPTER 7: CONCLUSION ....................................................................................... 226
Summary ......................................................................................................... 226
Limitations ...................................................................................................... 23 1
Implications for Policy and Practice ............................................................... 233
Extending Frameworks on Teachers’ Career Decisions ................................. 235
Future Research .............................................................................................. 240
APPENDIX A: INITIAL SURVEY INSTRUMENT .................................................... 244
APPENDIX B: FOLLOW-UP SURVEY ....................................................................... 256
APPENDIX C: INTERVIEW PROTOCOLS ................................................................ 262

APPENDIX D: ANALYSIS TABLES EXCLUDING OAK STATE UNIVERSITY

RESPONDENTS ............................................................................................................ 265
APPENDIX E: DISTRIBUTION OF TEACHER CANDIDATES BY GENDER ....... 270
REFERENCES ............................................................................................................... 273

viii

LIST OF TABLES

Table l - Descriptive Information on the Colleges Used for Data Collection .................. 40
Table 2 - Number of Responses and Response Rates by College .................................... 41
Table 3 - Demographic Characteristics of Survey Respondents and All Elementary
Teachers in the State ............................................................................................. 43
Table 4 - Items With Highest Nonresponse Rate .............................................................. 45

Table 5 - Number of Responses to Follow-Up Survey and Response Rates by College .48

Table 6 — Number of Responses to F ollow—Up Survey and Response Rates by
Demographic Characteristics ................................................................................ 49

Table 7 - Overall Item-Level Nonresponse Rate and Items With Highest Nonresponse for
the F ollow-Up Survey ........................................................................................... 50

Table 8 - Descriptive Information on the Interview Participants ..................................... 55
Table 9 - Demographic Characteristics of Teacher Candidates From Each Institution 68

Table 10 - Factor Analysis Results and Item Loadings From Teacher Candidates’
Perceptions of Impact of Various Characteristics on the Likelihood of Getting a
Job Offer ............................................................................................................... 82

Table 11 - Mean and Standard Deviation for Each Measured Factor ............................... 83

Table 12 - Percentage of Teacher Candidates Who Named Various Characteristics as
What Schools Look for When Hiring Teachers .................................................... 84

Table 13 - Regression Analyses of the Inﬂuence of Being White and Coming From a
Racial Minority Background in Teacher Hiring ................................................... 91

Table 14 - Percentage of Teacher Candidates Who Preferred to Teach in Various Types
of Schools ............................................................................................................ 107

Table 15 - Logistic Regression Results Estimating the Likelihood That Teacher
Candidates Preferred to Teach in an Urban or Low-Income Suburban School.. 110

Table 16 - Teacher Candidates’ Espoused Preferences: Characteristics Named as One of
the Three Most Important Features in Job Search .............................................. 124

Table 17 - Percentage of Teacher Candidates Who Named a Speciﬁc School, Named a
District but Not a School, or Did Not Name Any District as Their Most Desired
and Least Desired Places to Teach ...................................................................... 128

ix

Table 18 - Percentage of Teacher Candidates Who Planned to Apply to Various Schools
............................................................................................................................. 135

Table 19 — Sources of Information on Teacher Candidates’ Most Desired and Least
Desired Schools, as Indicated on the Initial Survey ........................................... 140

Table 20 - Percentage of Teacher Candidates Who Said They Know About Schools
Through Various Sources of Information, as Indicated Through Interviews ..... 141

Table 21 - Percentage of Teacher Candidates Who Did Not Know Whether Their Most
Desired and Least Desired Schools Had the Characteristics They Named as Most
Important ............................................................................................................. 145

Table 22 - Descriptive Statistics for Chapter 5 ............................................................... 162

Table 23 - Percent of Potential Teachers Who Removed Themselves From the Teacher
Labor Market and the Number of Applications .................................................. 163

Table 24 - Percentage of Students Eligible for Free or Reduced-Price Lunch (FRL)
Where Teacher Candidates Applied, Received Offers, and Accepted Jobs ....... 167

Table 25 - Percentage of Minority Students Where Teacher Candidates Applied,
Received Offers, and Accepted Jobs .................................................................. 170

Table 26 - Mean Pupil-Teacher Ratios Where Teacher Candidates Applied, Received
Offers and Accepted Jobs ................................................................................... 172

Table 27 - Mean Per-Pupil Instructional Expenditures Where Teacher Candidates
Applied, Received Offers, and Accepted Jobs .................................................... 173

Table 28 - Mean Salary Where Teacher Candidates Applied, Received Offers, and
Accepted Jobs ...... ' ............................................................................................... 175

Table 29 - Percentage of Teacher Candidates Who Applied, Received Offers, and
Accepted a Job in Urban Schools ....................................................................... 177

Table 30 - Percentage of Teacher Candidates Who Applied, Received Offers, and
Accepted Jobs in Other States and Average Distance From Home for

Applications, Offers, and Jobs ............................................................................ 179
Table 31 - Correlation Between Teacher Candidate Characteristics and the Districts to

Which They Applied ........................................................................................... 183
Table 32 - Month in Which Teacher Candidates Received Their First Job Offer .......... 185

Table 33 - Correlations Between School Characteristics of Job Offers and Month Offer
Was Received ...................................................................................................... 186

Table 34 - Correlations Between Characteristics of Schools Where Teaching and Initial

Top Choice School .............................................................................................. 188
Table 35 - Regression Analyses Predicting the Percentage of Minority Students in School
Where Teaching .................................................................................................. 191
Table 36 - Regression Analyses Predicting the Percentage of FRL Students in School
Where Teaching .................................................................................................. 193
Table 37 - Logistic Regression Analyses Predicting Whether a Teacher Candidate Ended
Up Teaching in an Urban School ........................................................................ 195
Table 38 - Descriptive Statistics for Chapter 6 ............................................................... 205
Table 39 - Labor Market Outcomes ................................................................................ 206
Table 40 - Regression Analyses Predicting the Number of Job Interviews ................... 209
Table 41 - Logistic Regression Analyses Predicting Whether a Teacher Candidate Was
Offered a Teaching Job ....................................................................................... 213
Table 42 - Logistic Regression Analyses Predicting Whether an Application Resulted in a
Job offer .............................................................................................................. 215
Table 43 - Regression Analyses Exploring Whether Districts Varied in the Teacher
Candidates They Offered Jobs ............................................................................ 218
Table 44 - Future Plans of Teacher Candidates Who Did Not Get Job Offers ............... 220

Table A 1 - Regression Analyses of the Inﬂuence of Being White in Teacher Hiring,
Excluding Oak State Respondents ...................................................................... 265

Table A 2 - Percentage of Students Eligible for Free or Reduced-Price Lunch (F RL)
Where Teacher Candidates Applied, Received Offers, and Accepted Jobs,
Excluding Oak State Respondents ...................................................................... 266

Table A 3 - Percentage of Teacher Candidates Who Applied, Received Offers, and
Accepted Jobs in Other States and Average Distance From Home for

Applications, Offers, and Jobs, Excluding Oak State Respondents ................... 267
Table A 4 - Regression Analyses Predicting the Percentage of FRL Students in School
Where Teaching, Excluding Oak State Respondents ......................................... 269
Table A 5 - District Characteristics Where Male and Female Teacher Candidates Applied
............................................................................................................................. 270
Table A 6 - District and School Characteristics Where Male and Female Teacher
Candidates Received Job Offers ......................................................................... 271

xi

Table A 7 - District and School Characteristics Where Male and Female Teacher
Candidates Accepted Jobs ................................................................................... 272

xii

LIST OF FIGURES

Figure l - Scree Plot of Factors in Teaching Field ........................................................... 80

Figure 2 - Steps in the Job Search ................................................................................... 160

xiii

CHAPTER 1: INTRODUCTION

Teachers are a school’s most important resource. Yet the distribution of teachers
is highly uneven with schools in urban areas—those likely to have high concentrations of
poor and minority students—being the least likely to have qualiﬁed teachers (Betts,
Rueben, & Danenberg, 2000; Lankford, Loeb, & Wyckoff, 2002). The inequitable
distribution of qualiﬁed teachers is due, in part, to the initial matching of new teachers to
their ﬁrst teaching jobs, with many candidates choosing more afﬂuent, less diverse
schools located close to where they grew up (Boyd, Lankford, Loeb, & Wyckoff, 2002,
2005a). Teachers’ subsequent decisions to transfer from their initial schools further
exacerbates this inequitable teacher sorting, as teachers tend to move from schools
serving low-income, minority students to schools serving more afﬂuent areas with fewer
racial minority students (Hanushek, Kain, & Rivkin, 2004).

This inequity in access to high quality teachers has implications for student
learning. High quality teachers can have signiﬁcant and substantial effects on student
achievement (Ferguson & Ladd, 1996; Goldhaber & Brewer, 2000; Rivkin, Hanushek, &
Kain, 2005; Sanders & Rivers, 1996). While there are few systematic relationships
between speciﬁc teacher qualiﬁcations and student achievement, teacher quality does
matter (Rivkin et al., 2005). Schools with persistent staffing difﬁculties are oﬁen forced
to hire lower quality teachers and thus reduce the learning opportunities for students—
often the students most in need of good teachers.

Researchers and policymakers have not ignored this important policy area. People

care about the inequities in access to high quality teachers and there is a proliferation of

policies to support the recruitment and retention of new teachers in high-need schools. No
Child Left Behind requires states to make sure poor and minority students are not more
likely to have inexperienced or unqualiﬁed teachers. Seventeen states offer salary
bonuses, scholarships, loan forgiveness, or housing assistance for teachers who agree to
work in hard-to-staff schools (Education Commission of the States, 2005). Arizona State
University partnered with a state senator and local districts to offer scholarships to
student teachers in high-need schools, with the hope they will stay to teach in those
schools. Alternative certiﬁcation programs help districts “grow their own” teachers to
meet stafﬁng needs.

Researchers have also focused attention on the sorting of teachers among schools.
The bulk of this research has examined teacher career patterns and the characteristics of
schools that are associated with these patterns to further our understanding of teachers’
preferences for schools. What characteristics of schools and teaching positions matter
most to teachers? What school features are important in teachers’ career decisions?
Often, this research was designed to answer questions like these so that policies could be
designed to meet teachers’ preferences. For example, how much do we need to increase
salaries to entice teachers to hard-to-staff schools (Hanushek et al., 2004)? What impact
do mentoring or induction policies have on attrition decisions of new teachers (Smith &
Ingersoll, 2004)? Overall, this body of research has focused on identifying and describing
teachers’ preferences, hoping that knowledge of these preferences could inform policy
and practice.

This dissertation also began as an investigation into teachers’ preferences for

schools. By situating teachers in their social and cultural contexts, the goal was to

understand how preferences varied among teachers and how these preferences developed.
Yet it is about more than preferences. I now view this dissertation as a study of the
processes by which teachers come to work in particular schools. This difference between
preferences and processes is not semantic. It reﬂects a different set of assumptions about
how teachers approach the task of ﬁnding a teaching job.

Focusing on preferences posits teachers as rational actors that seek to maximize
their career satisfaction by getting the job that best meets their criteria. Preferences are
about what teachers want in a school and what gives them satisfaction in their work-life.
Yet how teachers gather information or make assessments about which school would
provide the most satisfaction is left unspeciﬁed. Focusing on processes, however,
emphasizes prospective teachers’ experiences on the job market and how they approach
the task of ﬁnding a job. This dissertation tries to understand how teachers make use of
their preferences during the job search. It focuses on the contexts in which schools
become options for teachers and how they decide if a school is a desirable place to work.
It highlights ways in which teachers’ behavior, while reasonable, reduces the options
available to them without assuming this behavior is the result of rational calculation. I
explore these ideas through a longitudinal mixed-methods study of the career decisions of
prospective elementary teachers from six colleges in one metropolitan area.

This dissertation introduces a new framework through which to view teachers’
career decisions. Bourdieu’s cultural reproduction theory has been used to examine
teachers’ perceptions of students and parents as well as individuals’ decisions in school
and the labor market (Lareau, 1987; Lareau & Horvat, 1999; MacLeod, 1995), but has

not yet been applied to a study of teachers’ decisions in the teacher labor market. I use

Bourdieu’s ideas of habitus, ﬁeld, and capital to explore how prospective teachers
perceive and navigate the teacher job search. Prospective teachers may take many actions
throughout the job search; they may research various schools, look for available jobs,
contact individuals in their social network, send out applications, go on interviews, gather
information, and, ultimately, accept (or, in some cases, reject) a position. These behaviors
are inﬂuenced by their perceptions of what makes a good school, their sense of what it
takes to get a teaching job, the criteria they use to evaluate potential teaching jobs, their
experiences on interviews, and how they conceive their job alternatives. These
perceptions and experiences are, in turn, shaped by the socio-cultural contexts of teacher
and schools. By focusing on the processes rather than preferences, I attempt to open up
the black box of a teacher’s job search.
What We Know About Teachers ’ Career Decisions

Three main strands of research add to our knowledge of how teachers make career
decisions. These research strands come from different disciplinary backgrounds: labor
economics, organizational sociology, and psychology. Due to these various disciplinary
backgrounds, each strand relies on different methodologies and assumptions to explain
teachers’ career decisions. Yet, all three seek to understand teachers’ preferences for
schools as workplaces. As a whole, these research traditions identify three important
features of jobs that guide teachers’ decision-making: economic incentives,
organizational characteristics, and position-speciﬁc features. As each research tradition
provides some insight into teachers’ preferences for each of these features, this section

will ﬁrst describe the methods and assumptions behind each research strand and then

describe what they tell us—as a whole—about teachers’ preferences and decisions on the
job market.

Economic studies of the teacher labor market posit teachers as rational actors that
make decisions based upon objective and observable features of schools. Through this
perspective, teachers have preferences for particular work environments and weigh
potential jobs against these criteria. It is assumed that teachers have full information
about the satisfaction they will receive from different jobs, resulting from both salary and
working conditions, and choose the position that will give them the most satisfaction.
Schools with some undesirable characteristics may compensate by having other, more
amenable, features, or by offering a higher salary (Rosen, 1986). Other economic studies
examine the impact of various school characteristics on teachers’ exit decisions, length of
time spent teaching, or mobility between jobs. These studies assume that teachers’
preferences are guiding their turnover decisions; therefore, identifying school
characteristics associated with lower turnover rates provides information on teacher
preferences. Economic studies of the teacher labor market have identiﬁed a number of
features of schools that appear to inﬂuence teachers’ satisfaction with working in a
school, such as salary, class size, the racial ethnic, income, and achievement background
of the school’s students, and the geographic location (Antos & Rosen, 1975; Baugh &
Stone, 1982; Boyd et al., 2005a; Boyd, Lankford, Loeb, & Wyckoff, 2005b; Brewer,
1996; Dolton & van der Klaauw, 1999; Hanushek, Kain, & Rivkin, 1999; Kirby,
Berends, & Naﬁel, 1999; Mumane & Olsen, 1990; Scaﬁdi, Sjoquist, & Stinebrickner,

2007; Stinebrickner, 2001; Theobald, 1990).

A different line of research about teachers’ career decisions draws on ideas from
organizational sociology. This literature focuses on organizational features of schools that
affect teachers’ careers. Like the economic studies, these studies may explore the effect
that different school features have on teacher turnover, although they may also focus
more broadly on factors that inﬂuence teacher satisfaction. They also draw on
occupational and organizational norms, as well as experiences of new teachers. This
framework highlights how individuals are brought into general occupations and speciﬁc
organizations, and how organization-wide features shape an individual’s sense of job
satisfaction and organization-wide turnover. Research in this tradition has emphasized
teachers’ need to feel successful on the job and the administrative or organizational
characteristics of schools that contribute to this sense of success, such as the level of
conﬂict and strife in an organization, support from administrators or more experienced
colleagues, new teacher mentoring and induction, and inﬂuence over schoolwide policies
(Ingersoll, 2001; Johnson & Birkland, 2003; Johnson, Kardos, Kauffman, Liu, &
Donaldson, 2004; Johnson & Landman, 2000; Kardos, Johnson, Peske, Kaufﬁnan, & Liu,
2001; Smith & Ingersoll, 2004).

Finally, a third research tradition uses experimental psychology methods to test
hypotheses about factors that inﬂuence teachers’ evaluations of teaching positions. This
literature draws heavily from more general human resource functions in organizations. It
focuses on factors that inﬂuence teacher recruitment and how attractive particular jobs
appear to teachers. As experiments, these studies are often conducted in laboratory ,
settings and vary one or two components in a recruitment situation to isolate effects. The

studies might choose to vary characteristics of the teacher candidates, characteristics of

the position available, or characteristics of the job announcement or organizational
representative. For example, some studies have varied the race and gender of
organizational representatives talking about an available position and found that teachers
ﬁnd jobs more attractive if they are presented by someone of the same race and gender as
themselves (Young, Place, Rinehart, Jury, & Baits, 1997). Other studies have varied the
recruitment message, or the description of the available position, and found that job
announcements that emphasize extrinsic factors—such as salary—are considered less
attractive to teachers than job announcements that emphasize the work tasks or work
environment (Young, Rinehart, & Heneman, 1993).
Economic Incentives

Together, all three strands of research on teachers’ career decisions have
identiﬁed several school characteristics that appear to inﬂuence teacher preferences for
schools. One feature of teaching jobs that is of central importance to policymakers is the
extent to which teachers are sensitive to salary differences between schools. Studies in
this tradition have found that teachers respond to both wage and non-wage characteristics
of schools. Salaries affect both whether an individual chooses to teach at all, and, if they
do, where that individual takes a teaching job (Brewer, 1996; Dolton & Makepeace,
1993; Stinebrickner, 2001). Teachers also are less likely to quit when teaching wages are
relatively high (Dolton & van der Klaauw, 1999; Grissmer & Kirby, 1997; Kirby &
Grissmer, 1993; Mumane & Olsen, 1990; Theobald, 1990). Indeed, economic studies
have found teachers react to wages at least as much as do workers in other occupations.

Baugh and Stone (1982), for example, found that teachers respond both to wage

differentials between teaching and non-teaching jobs, as well as to pay differences
between teaching jobs.

Yet salary is still a limited tool by which to attract and retain teachers in particular
teaching positions. Mumane and Olsen (1989, 1990) noted that ﬁxed effects methods
reduce the relative impact of salary on teachers’ decisions to quit, suggesting that some
school attributes also affect teachers’ decisions. Evidence from experimental psychology
studies suggest that teachers consider jobs more attractive if they emphasize the work
environment and work tasks rather than economic incentives (Young et al., 1997; Young
et al., 1993), although male teachers give more attention to salary than women when
evaluating jobs (Winter, 1996). Sociologists focus on the importance of the student-
teacher relationship as the psychic rewards of teaching (Lortie, 1975). If teachers cannot
attain these intrinsic rewards and feel successful with their students, ﬁnancial incentives
may have limited effectiveness in retaining teachers (Liu, Johnson, & Peske, 2004). For
example, the Massachusetts Signing Bonus Program promised new teachers $20,000 if
they taught in a public school for four years. Yet many participants left teaching before
earning all of the bonus because they did not feel they were achieving success with their
students (Liu et al., 2004). In an experimental design study, teachers did not rate jobs
more attractive if they had a 10% signing bonus (Winter & Melloy, 2005). Likewise,
Hanushek, Kain, and Rivkin (1999) found that differences in school characteristics are
better predictors of teachers’ movement between schools than salary.

Organizational Characteristics
As school characteristics and working conditions inﬂuence teachers’ career

decisions independently of salary, it is important to know which features of schools or

districts appear to matter in these choices. All three research traditions have identiﬁed
several features of schools that appear to inﬂuence teachers’ career decisions. Teachers
care about class size and workload (Antos & Rosen, 1975; Eberts & Stone, 1985;
Hanushek et al., 1999; Homg, 2005; Kirby et al., 1999; Levinson, 1988; Mont & Rees,
1996; Stockard & Lehman, 2004). For example, teachers in North Carolina are paid a
premium to teach in larger schools, suggesting that teachers prefer smaller schools and
districts (Antos & Rosen, 1975; Levinson, 1988; Walden & Sogutlu, 2001). A teacher’s
probability of leaving a particular school increased by 3 percent for each additional
student in her classroom (Kirby et al., 1999). Other measures of teacher workload have
been used to examine teachers’ preferences for job attributes. For example, teachers
prefer to work in schools with more professional support staff and teacher aides (Kirby et
al., 1999; Walden & Sogutlu, 2001). Qualitative studies of new teachers’ experiences
indicated that the teaching assignment or workload contributed to teachers’
dissatisfaction and likelihood of leaving (Johnson & Birkland, 2003). Teachers are wary
of schools in which the scope and deﬁnition of their responsibilities are not clearly
delineated as they may foster unrealistic work expectations and burnout (Johnson &
Landman, 2000).

Researchers have also investigated the extent to which teachers care about the
resources and facilities in their schools. Loeb, Darling-Hammond, and Luczak (2005)
found no direct relationship between the adequacy of textbooks, availability of
computers, or professional development opportunities on teacher turnover or difﬁculty in
ﬁlling teaching vacancies. Yet, schools that used non-classroom space for instruction or

had dirty bathrooms had more serious problems with turnover and difﬁculty ﬁlling

vacancies. Likewise, Horng (2005) found that having clean and safe facilities was the
most important attribute of schools for teachers. Stockard and Lehman (2004) found that
teachers were more satisﬁed in schools with sufﬁcient supplies. F alch and Strom (2005)
found that teachers are less likely to leave a school if there is a library.

Student discipline also appears to inﬂuence teachers’ career choices. Studies of
teacher turnover and retention have found that teachers are more likely to stay in schools
with lower levels of student discipline problems; teachers moving to a new school or
leaving teaching altogether often cite student discipline as a cause of their dissatisfaction
(Ingersoll, 2001; Knapp, Elfers, Plecki, Loeb, & Zahir, 2005). A study of new teachers’
experiences on the job found that discipline problems and student disrespect led to higher
levels of teacher dissatisfaction (Johnson & Birkland, 2003). Indeed, teachers’
perceptions of student discipline problems contributed to teachers’ decisions to leave
teaching, despite a $4000 salary bonus for staying an additional year (Liu et al., 2004).

Studies from economics, organizational sociology, and experimental psychology
have all identiﬁed the importance of administrative and support structures in teachers’
career decisions. For example, one experiment found that teachers care more about
features of their work environment—such as shared decision-making and interactions
with colleagues—than factors related to economic incentives or the work itself (Young,
Rinehart, & Place, 1989). Although administrative support can mean many things to
teachers, teachers’ perceptions of how effective the school and principal are in supporting
new teachers lead to a reduced likelihood of exit (Ingersoll, 2001) and higher teacher
satisfaction (Stockard & Lehman, 2004). Interaction with more experienced colleagues

and opportunities for instructional growth also led to higher levels of teacher satisfaction

10

(Johnson & Birkland, 2003; Kardos et al., 2001). Indeed, for these teachers, achieving
success in the classroom depended on the contributions of other colleagues and the
principal. Similar to student discipline, the amount of curriculum support available to
new teachers contributed to their decision to leave public school teaching despite an
increased salary (Liu et al., 2004).

Teachers also prefer schools where they can inﬂuence decision-making (Ingersoll,
2001). Teachers are more satisﬁed when they have a role in designing the school plan and
in school governance (Johnson & Landman, 2000). Similarly, the principal leadership
style matters to teachers. Teachers are more likely to transfer schools if they have a
principal with an authoritarian leadership style than one who is more democratic
(Bempah, Kaylen, Osburn, & Birkenholz, 1994).

Another important attribute of schools is the extent to which they have structures
in place to support new teachers. Teachers are most likely to move schools or exit the
profession within their ﬁrst few years of teaching (Mumane, Singer, Willett, Kemple, &
Olsen, 1991). Further, teachers who think they get support from colleagues and a mentor
report higher levels of satisfaction (Stockard & Lehman, 2004). Thus, school
characteristics that help new teachers may be particularly beneﬁcial in reducing turnover.
New teachers are less likely to quit teaching if they are assigned a mentor teacher in the
same teaching ﬁeld, have collaboration or planning time with other teachers, or have
extra classroom assistance (Smith & Ingersoll, 2004). Yet, teachers in low-income
schools are less likely to have a mentor in the same ﬁeld or sufficient curriculum

guidance and ﬂexibility in the classroom (Johnson et al., 2004).

ll

Position-Specific Characteristics

In addition to school organizational characteristics and the economic incentives
associated with various teaching positions, teachers also care about some aspects of the
job itself. While districts and schools may control organizational structures and the salary
offered to teachers, some features of teaching positions are less amenable to policy
intervention. Nevertheless, these features do affect teachers’ career decisions and may
have an important impact on the nature of the work involved in various teaching
positions. While the nature of the work might appear to be uniform across elementary
schools, differences in grade level, subject matter specialties, and student population may
inﬂuence the speciﬁc work tasks involved in teaching.

The ﬁndings with respect to location are particularly noteworthy as teachers care
about the geographical location of teaching positions (Boyd et al., 2005a; De Tray &
Greenberg, 1977; Rynes & Lawlor, 1983; Toder, 1972). Boyd, Lankford, Loeb, and
Wyckoff (2005a) found that teachers prefer to teach close to where they grew up. Sixty-
one percent of new teachers had their ﬁrst teaching job within 15 miles of where they
went to high school. This hometown preference also existed for teachers who went to
college at least 100 miles from their high school location. Teachers have a preference not
only for proximity to their hometown, but also for communities similar in type to where
they grew up. Teachers who grew up in urban areas were more likely to take a teaching
job in an urban area. Similarly, Rynes and Lawlor (1983) found that most teachers ﬁnd
inner city schools less attractive than other types of schools. This preference for location
also affects teacher turnover. For example, teachers who lived outside of New York City

prior to accepting a teaching job within the city were ﬁve times more likely to transfer to

12

a school outside of the city than city residents (Boyd et al., 2005b). The ﬁnding in many
studies that teachers prefer to teach in suburban areas over urban areas (De Tray &
Greenberg, 1977; Kirby et al., 1999; Toder, 1972) may reﬂect the fact that most teachers
grew up in the suburbs.

Teachers also have preferences for teaching a particular grade level or subject
area, even among elementary grades. Teaching a self-contained ﬁrst grade class can be
quite different from teaching a self-contained ﬁfth grade class. Also, it is relatively
common for an upper elementary grade teacher to focus on math and science or language
arts and social studies, trading students with one another for part of the day. As a speciﬁc
example of this, Stockard and Lehman (2004) found that elementary level teachers were
more satisﬁed with their jobs if they taught in their preferred grade and did not teach in a
middle school. Teaching in one’s area of certiﬁcation also leads to greater satisfaction
(Stockard & Lehman, 2004).Teachers in Washington reported that their speciﬁc teaching
assignment was the most important reason to stay in their school (Knapp et al., 2005).

These studies also suggested that teachers care about the demographic
characteristics of their students. In particular, teachers appear to make career decisions
based on the academic performance, socioeconomic status, and racial background of
student populations served by various schools. Teachers tend to leave schools with low
performing students and move to schools serving higher achieving students (Boyd et al.,
2005b; Clotfelter, Ladd, & Vigdor, 2004; Hanushek et al., 1999; Scaﬁdi et al., 2007).
Teachers appear to be offered a positive compensating pay differential for teaching in
schools with lower student verbal scores (Antos & Rosen, 1975) and report lower levels

of satisfaction if they teach low achievers (Stockard & Lehman, 2004). Both experienced

13

and inexperienced teachers ﬁnd jobs less attractive if they are located in schools labeled
in need of attention by state accountability programs and more attractive if they are
labeled as meeting these goals (Winter & Melloy, 2005). Teachers also care about other
student outcomes, including graduation rates and college attendance (Antos & Rosen,
1975; Eberts & Stone, 1985).

Teachers are also sorted across schools by the percent of students who come from
low-income families, with more qualiﬁed teachers in schools in low poverty areas (Betts
et al., 2000; Lankford et al., 2002). While this sorting is evident in the ﬁrst job teachers
take (Boyd et al., 2002), it is exacerbated as teachers move from less afﬂuent to more
afﬂuent schools (Boyd et al., 2005b; Clotfelter et al., 2004; Hanushek et al., 1999;
Levinson, 1988; Scaﬁdi et al., 2007). Among public schools, schools in high poverty
areas have higher rates of teacher turnover and lower levels of job satisfaction than those
in more advantaged areas (Boyd et al., 2005b; Ingersoll, 2001; Levinson, 1988; Scaﬁdi et
al., 2007; Smith & Ingersoll, 2004; Stockard & Lehman, 2004). Teachers also appear to
care about other features of a school community’s socioeconomic status. Both White and
non-White teachers prefer to teach in communities where there is a higher percentage of
professional and technical workers (Levinson, 1988; Mumane & Olsen, 1990), although
the proportion of the population that is poor did not signiﬁcantly affect teachers’
decisions (Levinson, 1988). Teachers appear to be paid a compensating pay differential
when there is a relatively high percentage of students who are eligible for free or
reduced-price lunch, a common indicator of school poverty (Antos & Rosen, 1975;
Levinson, 1988). Teachers are less likely to leave when there is a high median household

income in the area around the school (Mont & Rees, 1996).

14

Finally, the racial and ethnic composition of the student body also appears to
inﬂuence teachers’ career decisions. A number of studies have found that, overall,
teachers prefer to teach in schools with fewer racial and ethnic minority students (Antos
& Rosen, 1975; Boyd et al., 2002, 2005b; Clotfelter et al., 2004; Falch & Strom, 2005 ;
Hanushek et al., 1999; Scaﬁdi et al., 2007; Strunk & Robinson, 2006). Teachers in these
studies are paid compensating differentials for teaching in schools with greater numbers
of minority students (Antos & Rosen, 1975; Eberts & Stone, 1985; Levinson, 1988;
Toder, 1972). They are also more likely to quit teaching or transfer schools if they teach
in a school with a high proportion of non-White students (Boyd et al., 2005b; Mumane &
Olsen, 1989; Scaﬁdi et al., 2007; Theobald, 1990). Schools which have a majority of
African Americans or Latino students are more likely to have a serious problem with
teacher turnover (Loeb et al., 2005).

How Teacher Characteristics Influence Teacher Preferences

One important question when looking at teachers’ preferences for schools is
whether preferences are uniform or varied. A uniform set of preferences would indicate
that all teachers prefer to teach in similar types of schools and that it would be relatively
easy to rank schools in order of attractiveness to teachers. More likely, preferences vary
across teachers, with different teachers preferring to teach in different types of schools.
This variation may be due to teachers having preferences for particular characteristics but
weighting the various characteristics differently. For example, two teachers may both
want to teach in a school with a small class size and a well-stocked library, but one
teacher thinks class size is more important while another gives more weight to the school

library. Or, it may be due to teachers preferring different features. For example, some

15

teachers may avoid schools with many English as a Second Language students, while
other teachers seek out schools with linguistic diversity.

Understanding the nature of the variation in teachers’ preferences for schools is
also important. This variation may be random or systematic. Systematic variation in
teachers’ preferences suggests that we can identify teacher characteristics that are
associated with particular types of preferences for schools. Perhaps, for instance, male
teachers make decisions between schools differently than female teachers. Or maybe
individuals who switch into teaching as a second career have different preferences than
those just starting out. Random variation in teacher preferences implies that while
different teachers may prefer different schools, we cannot predict which teachers will
prefer which schools. As shown below, there is evidence that preferences are systematic.
By better understanding why teachers make the choices that they do, we could then more
effectively address policies to recruit particular types of teachers.

The three research strands on teachers’ career decisions explore variation in
teacher preferences, although they do so in different ways and to different extents. Of the
three, research on human resource decisions in schools using experimental psychology
methods is most likely to explore possible systematic variation in teachers’ preferences.
Studies in this tradition often use a teacher background characteristic as an experimental
control variable in designing studies. For example, one study found that extrinsic features
of available jobs are relatively more important to male teachers than they are to female
teachers (Winter, 1996) while others found that inexperienced teachers rate all jobs as
more attractive than do experienced teachers (Winter & Melloy, 2005; Young et al.,

1993). Further, Young, Place, Rinehart, Jury, and Baits (1997) found that teachers are

16

more likely to accept a job if it is presented to them by someone of the same race,
suggesting that teachers’ evaluations of a particular school varies by their racial and
ethnic background. As noted above, these studies are designed to test speciﬁc hypotheses,
so they do provide some explanation for why teacher preferences vary in these ways. For
example, the study by Young et al (1997) tests the attraction-similarity hypothesis—that
teachers are attracted to jobs in which they match the demographic characteristics of the
organizational representative presenting the job. Their ﬁnding that teachers prefer jobs in
which there is a match on racial and ethnic background but not necessarily on gender
provides some support for the attraction-similarity hypothesis.

Economic studies also pay some attention to how teachers’ preferences for
schools vary by looking at the interaction of teacher and student race. These studies have
explored whether White and non-White teachers have the same preferences for student
racial and ethnic characteristics. For example, as noted above, Boyd, Lankford, Loeb, and
Wyckoff (2005b) found that White and Hispanic teachers were more likely to leave a
school if it served a high number of African American students. African American
teachers’ decisions to leave a school, however, appeared relatively insensitive to the
racial composition of the student body. Similarly, Scaﬁdi, Sjoquist, and Stinebrickner
(2002) found that, Aﬁican American teachers’ likelihood of exiting a school was similar
for schools serving mostly White or mostly Aﬁican American students.

Other studies, however, have found that African American teachers do care about
student racial ethnic composition. For example, while White teachers were more likely to
leave a school if it had large numbers of minority students, the opposite was true for

African American and Hispanic teachers, who were more likely to leave a school if it had

17

large numbers of White students (Hanushek et al., 1999). One study noted the marked
degree of teacher sorting by racial and ethnic group in California schools (Betts et al.,
2000). This segregation of teachers and students by race and ethnicity was present despite
policies to ensure that the racial and ethnic distribution of teachers in a particular school
matched the overall racial and ethnic distribution of teachers in the district (Betts et al.,
2000). Antos and Rosen (1975) found that teachers tended to segregate themselves by
race. Likewise, Strunk and Robinson (2006) found that teachers prefer to teach in schools
with more students of their same race and, for some teachers, more teachers from similar
racial backgrounds. Further, there is heterogeneity among teachers in terms of their
preferences for student achievement attributes. While some teachers seem insensitive to
levels of student achievement when making decisions to quit teaching or transfer to
another school, other teachers are much more likely to leave when achievement is low
(Boyd et al., 2005b).

Studies coming from organizational sociology, in contrast, often overlook issues
of race and class in analyzing teachers’ sense of satisfaction or reasons behind choosing
particular schools. Johnson and colleagues (2004) conducted a longitudinal study of new
teachers’ experiences. This group, called The Project on the Next Generation of
Teachers, published several research articles on teachers’ career decisions. However,
very few of these studies mention how teachers’ race or class may have inﬂuenced these
decisions. The researchers did try to obtain a sample that was racially diverse, but did not
take advantage of this sample variation in their analyses. For example, Kardos, Johnson,
Peske, Kaufﬁnan, and Liu (2001) did provide some demographic information on the

teachers, including their racial and ethnic background, but did not explore how teachers’

18

perceptions of their school’s professional culture may have differed by race. Likewise,
Liu, Johnson, and Peske (2004) provided descriptive information on their sample,
including demographic characteristics of the teachers and their schools, but failed to
explore how these sample characteristics may have affected decisions to continue
teaching or how teachers from different racial or class backgrounds may have reacted to a
$20,000 signing bonus differently. A smaller study (Horng, 2005) did note that Latino
teachers appear to have higher utility levels when they teach in schools with more
minority students.
Recent Trends in Research on Teachers’ Career Decisions

More recent work on teachers’ career decisions tends to take two forms. One
strand attempts to identify elements of induction and mentoring that inﬂuence turnover
decisions of new teachers (for example, see Smith & Ingersoll, 2004). Another focuses on
distinguishing between various school and student characteristics in these decisions.
Teachers prefer particular types of working environments, yet they are also distributed in
ways that disadvantage poor and minority students. Schools serving low-income, racial
minority students, however, are also those that tend to have larger classes, greater
discipline problems, and few supports for teachers. Thus, distinguishing between these
different features of schools and their students is important for establishing effective
education policy. Do we see the unequal patterns of the distribution of qualiﬁed teachers
because teachers seek out schools with high salaries and supportive environments, which
tend to be those serving more afﬂuent students, or do these patterns emerge as teachers
exercise their preferences for teaching students with particular characteristics? For

example, Horng (2005) and Loeb, Darling-Hammond, and Luczak (2005) use surveys of

19

current teachers’ perceptions to conclude that it is measures of the work environment,
rather than student characteristics, that inﬂuence teachers’ career decisions. Yet, Scaﬁdi,
Sjoquist and Stinebrickner (2007) use administrative databases of teachers to track them
over time and found that teachers do care about student characteristics, and, in particular,
their behavior can be explained by preferences for students from a particular racial and
ethnic background, rather than the ability or socioeconomic status of students.

As this dissertation explores the importance of student race and class composition
on teachers’ career decisions, it is in line with current trends in teacher labor market
research. Moreover, it explores the extent to which teachers’ preferences and their
subsequent decisions can be explained by objective features of the school work
environment or on socio-cultural understandings of students and teachers’ places within
the teaching ﬁeld. This dissertation’s focus on teachers’ job search processes extends the
current work on teachers’ career decisions and addresses some of the limitations in the
current literature, as outlined below.

Limitations of Existing Research on Teachers ’ Career Decisions

Despite the valuable knowledge about teachers’ preferences for schools that has
already been produced, studies based on economics, organizational sociology, and
experimental psychology each have methodological faults. Economic studies often use
administrative datasets that may not contain important features of either schools or
teachers that affect how these career decisions are made. Some economic studies use
survey data, although in these cases the data may not have been collected for the purpose
of exploring teachers’ job search processes. While the size of these datasets allow for

complex models, they cannot overcome the lack of important intervening variables.

20

Further, the quantitative nature of these studies results in limited attention to why
teachers prefer the schools that they do and their experiences with the job search.

Organizational sociology studies are more likely to delve into the reasons behind
teachers’ decisions with rich qualitative data, but they sometimes fail to take advantage of
the diversity in their samples to explore how preferences or perceptions might be related
to teacher characteristics. In addition, organizational sociology research is often
retrospective, asking teachers to reﬂect back, on why they moved schools or stayed in
their school after the decision has been made. Relying on retrospective accounts may
overlook some factors that guide teachers’ actions. Further, the factors behind teachers’
decisions to quit or transfer schools may not explain their initial choice of a school, yet
both the initial job and exit decisions contribute to the inequitable distribution of teachers
(Boyd et al., 2002).

Finally, human resource psychology research relies on experimental designs.
Studies in this tradition ask teachers to rate the attractiveness of various jobs that are
similar in most respects, but the researchers vary one or two features of the job for each
participant in the study. While this method is able to isolate the effect of particular
features, it creates a contrived situation in which to understand how teachers make
decisions about where to apply and where to work. Real career decisions are not made in
this isolated manner. Indeed, some decision-making research suggests candidates
evaluate jobs on an holistic basis, rather than attending to variations in particular features
(Rynes & Lawlor, 1983; Soelberg, 1967).

In addition to methodological limits in each of the three existing research strands

on teachers’ preferences, each ﬁeld relies on particular assumptions that limit their

21

usefulness. While these three frameworks have identiﬁed some of the features and
characteristics of schools that attract or repel teachers, they share an assumption that
teachers can and do objectively interpret these features of schools. Though some teachers,
through their location in particular social networks or experiences as a student teacher or
substitute, may have a sense of a school’s culture and speciﬁc policies, it is quite likely
that most beginning teachers will have a limited knowledge of such features. Indeed, the
hiring process often does not provide teachers with the amount and type of information
necessary to form an accurate picture of work-life at the school (Liu & Johnson, 2003).
Thus, teachers may be making decisions about where to teach—and sorting themselves
across schools in ways that disadvantage poor and minority students—based upon
assumptions about what it would be like to teach in certain schools or communities.
These beliefs about students and school communities may likely exist before
teachers enter schools. Guin (2004) used data from a study of schools with chronic
teacher turnover and found that low-income schools are not only more likely to
experience chronic turnover, but also receive substantially fewer applications for every
open teaching position. For example, she described how a school in one of the poorest
areas of a district received just 3-5 applications for each available position, while an
afﬂuent school in the same district had 150 applications for each opening. Potential
teachers around this district appeared to form opinions about which schools are desirable
before even applying. The hiring process may provide only limited information for
potential teachers (Liu and Johnson, 2003), but many teachers appear to select themselves

out of particular schools without even the beneﬁt of this limited information. This

22

underscores the necessity of examining how teachers interpret characteristics of schools
and how they form perceptions of schools as workplaces.

Even decisions to leave a school may be inﬂuenced by culturally—based
perceptions of whether difﬁculties experienced are endemic to the school. While teachers
appear to have full information about a school when they make the decision to quit or
transfer, that information is not necessarily accurate as it includes the biased, partial
understandings that teachers choose to acknowledge (Weiss, 1995). This knowledge is
especially open to being shaped and/or ﬁltered through teacher’s cultural models and
social prejudices. Teachers’ perceptions of students are ﬁltered through social and
cultural understandings and teachers deﬁne both school success and failure as linked to
students’ home environment or social background (Lareau & Horvat, 1999; Lipman,
1998; McDermott, 1974). The extent to which teachers prefer particular work
environments or feel successful with students thus depends, in part, on the meaning to
which teachers ascribe to school and student characteristics.

Teachers are social and cultural beings, and their perceptions of student behavior
as “disrespectful” depends on their culturally informed ideas about what motivates
student actions and what counts as “respect” in particular contexts, especially the
classroom. Schools can and do have effects on student behavior and many schools
serving low-income or African American students would not be perceived by most
teachers as having discipline problems. Yet, using a cultural reproduction framework,
“respectful” students may be little more than students whose communication styles match
those of the teachers, who are often White and come from middle class backgrounds.

Understanding how teachers make sense of these different characteristics of schools and

23

their students through their social and cultural position is important for understanding
teachers’ choices among schools. The organizational context of the school and
instructional support provided by the principal may help teachers deal more effectively
with student disruptions and increase their sense of efﬁcacy in the classroom. Yet, no
matter what the organization provides, teachers also come to schools with cultural ideas
about appropriate behavior and assumptions about how successful particular types of
students can be.

Likewise, teachers’ perceptions of whether they can be successful with students
may be ﬁltered through a socio-cultural lens. Johnson and Birkland (2003) emphasized
the importance of teachers’ ability to achieve a sense of success with students in their
decisions to leave teaching or transfer schools. They argue that teachers who moved
schools were searching for schools that had “stable faculties and the capacity to initiate
and sustain improvement efforts” (pp. 599). Yet, these teachers also all moved to schools
serving more afﬂuent students. While it is likely that inequalities in school funding result
in schools with greater concentrations of poor students having fewer resources for school
improvement efforts, it is also likely that teachers’ perceptions of students and reasons for
school failure result in the inability to engage in meaningful educational reform and the
lack of a sense of success with students.

Indeed, Lipman (1998) described how teachers have ideologies about student
failure that guide their interactions with students. Most teachers in her study explained
educational failure as due to some deﬁcit in the student’s family or home life, or because
students lacked social support for education. She used evidence from schools undergoing

signiﬁcant restructuring activities which created many of the organizational conditions

24

and collegial interactions that some’studies have linked with improved teacher working
conditions, such as creating substantial time for teachers to collaborate and talk about
instruction and encouraging teachers to develop new curriculum materials together. Yet,
these opportunities for collaboration often reinforced teachers’ ideas that students fail due
to a deﬁcit in students or their family life, as teachers realized they were not alone in
experiencing difﬁculties with students in the school. The restructuring efforts in these
schools led to lower expectations for African American students and teachers’
expressions of frustration that they could not use the same instructional strategies or
materials that they could with students in the honors track (who were predominantly
White students).

While Lipman (1998) did not explore teachers’ preferences for schools as
workplaces or their decisions about where to teach, her research is important because it
highlights how teachers’ perceptions of students are ﬁltered through social and cultural
understandings. Teachers make sense of their interactions with students through a socio-
cultural lens that includes ideas about why students fail and norms for student-teacher
interactions. This lens is, in turn, shaped by teachers’ own position within the social
structure. Johnson and colleagues (2004) argued that the student-teacher relationship is a
key component of whether teachers will feel successful and that teachers feel particularly
uncertain about this relationship when they come from a different cultural background
than their students. While schools may be more or less successful in helping teachers
overcome these cultural differences, Lipman’s research suggested that efforts to increase

teacher collaboration or facilitate teacher teaming are limited approaches if they do not

25

also address teachers’ ideological beliefs about students, their families, and the causes of
their educational success and failure.

Further, some attributes of schools may not be universally deﬁned. Teachers may
have different ideas about what it means for a school to be safe or have good student
discipline. Understanding what it means to teachers for principals to be supportive is
crucial, as teachers may evaluate their principal through a social and cultural lens about
the types of behavior and communication styles that are expected. For example, the same
principal who gives a great deal of discretion to teachers may be regarded as either
“ﬂexible” or “lacking clear standards” to teachers, depending on what they expect from a
principal. How teachers make sense of their principal and the amount of support they
perceive coming from the school leadership depends, in part, on their cultural
expectations for interaction with individuals in authority positions.

Finally, economic models often focus on the effects of marginal changes in school
characteristics, implying that teachers are willing to make tradeoffs between different
features of schools and their students at the margin. It is assumed that teachers may
sacriﬁce some desirable features of their work environment for a higher salary or a
shorter commute. Yet teachers may not evaluate positions in this way. Although teachers
may prefer to teach students of their own race, is it something they are willing to trade for
a higher salary or smaller class size? Indeed, research on decision-making behavior
suggests career decisions are made more holistically, using very few criteria and giving
little consideration before rejecting some jobs (Rynes & Lawlor, 1983; Soelberg, 1967;

Stevens, 1998).

26

Cultural Reproduction Framework

This study uses a cultural reproduction framework and Bourdieu’s notions of
ﬁeld, capital and habitus to identify the socio-cultural contexts of teachers’ job search
experiences. Bourdieu’s (1984; 1991) work illuminates both the socio-culturall processes
through which people perceive and construct distinctions between themselves and others,
and their relationship to the distribution of resources and opportunities—~including
occupational positions—across various social groups. The structure of the ﬁeld
inﬂuences people’s habitus and thus guides individuals in every day cultural practices.
This framework highlights how teachers’ career decisions are grounded in particular
social and cultural contexts that structure how they approach the task of ﬁnding a
teaching job and how they perceive various schools.
Field and Capital

According to Bourdieu, ﬁelds are spheres of activities that have, over time, come
to be autonomous from other activity spheres. That is, they have a particular logic that
both structures how people and groups are positioned in relation to each other and to the
work of the ﬁeld, and how actors who occupy these positions interact with one another. A
central part of this logic, or the “rules of the game” as Bourdieu refers to it, requires the
identiﬁcation of the types and combinations of capital as valued resources that provide a
person with status in a ﬁeld. For example, Bourdieu has argued that within the
educational ﬁeld in France, cultural capital, derived from the prestige associated with
where a person attended school, one’s knowledge of “high” culture, and one’s

communication styles and repertoires, is more important to one’s position within that

 

’ The term socio-cultural has more nuanced meanings in some research literatures. I am using the term to
capture the relationship between both social and cultural inﬂuences on the teacherjob search processes.

27

ﬁeld than is economic capital, or the possession of material wealth (Bourdieu, 1977a,
1984). Conversely, within the industrial ﬁeld, economic and social capital, or the
information and resources available through the people one knows or comes into contact
with, are more valued than cultural capital (Bourdieu, 1977a, 1984).

Recognizing that teachers occupy different positions within the teaching ﬁeld is
important for understanding teachers’ decisions. People typically obtain a position on this
ﬁeld depending, ﬁrst, on their possession of a requisite state-sanctioned credential. While
individuals who do not possess these credentials can obtain teaching jobs, their options
tend to be limited either to the private sector (which tends to offer lower salaries) or to
urban areas (which oﬁen face severe teacher shortages and inadequate resources). A
teaching credential represents a type of cultural capital. Within the ﬁeld, other types of
capital, including other types of cultural capital, may be salient. For example, the status
of the institution an individual graduated from, the knowledge of particular types of
pedagogical philosophies, and where an individual grew up may all serve, to varying
degrees, as forms of cultural capital. In addition, social capital in the form of the
information that teachers gather about job openings through formal and informal contacts
may be highly valued and, thus, highly relevant to where teachers end up teaching.

As players on the teaching ﬁeld, teachers interact with each other and with
schools. Through this interaction, teachers get a sense of their position on the ﬁeld.
Teachers occupy positions on this ﬁeld based upon the amount and distribution of capital
they hold. The positions are related to each other according to how the ﬁeld values the
different types of capital. For example, economic capital may not be highly valued in the

teaching ﬁeld and more weight may be given to social and cultural capital. Thus, teachers

28

from wealthy backgrounds may not be successful in converting their economic capital to
a high status position on the teaching ﬁeld, but those with a large social network
(signaling high social capital) or a degree from a highly ranked teacher preparation
program (signaling high cultural capital) may be able to convert those forms of capital
into a position in a highly ranked and/or desirable school. Although the teaching ﬁeld is
relatively autonomous, it is also inﬂuenced by what Bourdieu has called the broader ﬁeld
of power, or structural effects across ﬁelds that seek to conserve or transform relations
between types of capital (Bourdieu, 1998). Thus, teachers’ places in the ﬁeld are partly
shaped by their gender, social class, and racial and ethnic background.
Habitus and Practical Logic

According to Bourdieu, the habitus is the central mediating mechanism between
the structure of a ﬁeld and people’s preferences and everyday practices. Bourdieu deﬁned
the habitus as the enduring systems of dispositions, perceptions, and appreciations that
shape how individuals make sense of and act within a particular ﬁeld (Bourdieu &
Wacquant, 1992). Habitus structures the way individuals perceive objects, events and
actions, and it organizes the social world for them. It also represents the social structure
embodied within the individual agent (Bourdieu, 1984). As individuals interact on a ﬁeld
and experience a social life structured by the relations deﬁned by the ﬁeld, these relations
layer onto them and become embodied within their tastes, behaviors and perceptions.
This creates a correspondence between the objective social structure and individuals’
mental schemas for perceiving the social world (Bourdieu, 1984; Bourdieu & Wacquant,
1992). In this sense, agents internalize the external social structures (the logics of the

ﬁelds)-which they inhabit and act within. As we interact within the social world, we both

29

have and make sense of speciﬁc experiences that are particular to the position we occupy.
Through a sedimentary process (Bourdieu & Wacquant, 1992), the experiences of our
position deposit themselves in our mental map of the world, leading us to view the world
through those perceptions.

A central component of habitus is that individual behavior is guided by a practical
logic, or an understanding of what is appropriate in a given situation (Bourdieu, 1980,
1998). According to Bourdieu, our actions are not based on rational calculus, but on our
sense of “the game” and what action should come next. Agents on a ﬁeld act in ways that
are consistent with their embodied understanding of the ﬁeld, even though they are not
deliberately following rules (Bourdieu, 1998; Lamaison & Bourdieu, 1986). Bourdieu
called these actions strategies, although by strategy he did not mean a conscious choice or
strategic intention made through rational calculation (Bourdieu, 1998). He distinguished
strategies from “acting strategically” because individuals do not usually act with future
outcomes in mind (Bourdieu, 1980). Practical logic is neither unconscious—agents could
explain why they chose particular actions if asked—nor conscious—agents are not
necessarily thinking through possible actions and their ramiﬁcations before acting
(Lamaison & Bourdieu, 1986). In this way, Bourdieu distinguished practical logic from
both rational decision-making and socially deterministic behavior (Bourdieu, 1980). In
describing choices of marriage partners, for instance, Bourdieu describes matrimonial
strategies as “the product not of compliance with rules but of a sense of the game that
leads one to ‘choose’ the best possible match, in view of the hand that one has been
dealt” (Lamaison & Bourdieu, 1986, pp. 113). Actions are thus spontaneous results of

individuals responding to a social speciﬁc situation according to their practical logic.

3O

Yet to call actions spontaneous does not mean that they are random. An
individual’s practical sense has coherence because the habitus serves as a generating
principle that brings a unity to actions (Bourdieu, 1980). Our cultural schemas guide how
we view the world around us and how we act within it. The notion of habitus is useful for
understanding teachers’ career decisions because it focuses attention on how teachers
perceive available jobs and the cognitive schemas that generate their behavior. The
actions of teachers during the job search are generated their habitus. In this way, the
spontaneous actions appear coherent because they are the result of applying the same
generating principle to diverse situations (Bourdieu, 1980). As habitus is structured by
one’s social position, teachers from different cultural and social backgrounds behave
differently because their habitus have contrasting generating principles.

If the habitus inﬂuences one’s perceptions of the social world, it will inﬂuence
how teachers perceive different schools, communities and the teaching positions
available in different contexts, as well as provide a unifying principle that generates how
speciﬁc teachers respond to speciﬁc job offers. Determining the generating principle of
the habitus that creates one’s tastes and preferences is key to understanding how he or she
makes decisions. Teachers have preferences for certain types of schools because their
habitus leads them to interpret school characteristics and conditions in particular ways
and also has an underlying generating principle which produces a taste for particular
features and, by extension, schools. Therefore, teachers occupying different positions
within the teaching ﬁeld may perceive the same school in contrasting ways. Each may
ascribe quite different meanings to these features to create distinctions between schools.

One’s habitus may make a teacher notice differences between the number of poor

31

students in schools and that perception may lead teachers to make a judgment about the
value of that school. Yet one teacher’s generating principle could lead her to avoid
schools with high numbers of low-income students while another teacher may prefer that
school.

Bourdieu’s work suggests that, in making decisions about where to work, teachers
may rely on their perceptions of their position in the social structure and the opportunities
they perceive as open to them in the teacher labor pool. A teacher’s system of perception
and apperception may unconsciously shape their choices in ways that are aligned with
their position in the social structure. According to Bourdieu, our social and cultural
background inﬂuences how we perceive and react to the world, thus teachers may make
employment decisions not on a deliberate accounting of the working conditions and
economic beneﬁts of different schools, but on a subjective sense of their “proper” place
within the occupational ﬁeld, as well as a sense of their limits. Seen in this light, the
sorting of teachers across schools can be seen as a function of teachers’ understanding of
their place in the occupational structure and their self-selection into schools they think
match their position.

A cultural reproduction framework emphasizes how individuals adjust their
preferences to ﬁt one’s position on a ﬁeld. Economic sociology has moved closer to such
a conceptualization of how individual preferences operate. Akerlof and Kranton (2005)
argued that one’s identity describes their social category and norms about how people
should behave in particular situations. Individuals lose utility if their actions, or the

actions of those around them, do not match ideals for expected behavior. Teachers, then,

32

are dissatisﬁed if their students, or fellow teachers, do not match the teacher’s ideal for
student or teacher behavior.

Cultural reproduction theories likewise emphasize how teachers feel most at case
when their habitus, or set of cultural dispositions, match those with whom they interact.
Yet, the framework goes beyond this by highlighting how individuals’ social and cultural
position in the ﬁeld inﬂuences their perceptions of the world around them. Individuals get
a sense of their position in the ﬁeld and their preferences are adjusted to ﬁt what they
think they can expect from their position. Our habitus then not only shapes our
preferences, but it also shapes our system of perceptions and ways we approach the task
of ﬁnding a job. Teachers not only have different preferences for schools, but have
different strategies for getting a job and make different types of distinctions between
potential teaching jobs. A cultural reproduction framework is important to understanding
teachers’ career decisions as labor market studies traditionally look at a limited set of
school and student characteristics, and often do not explain why members of different
social groups behave differently (Akerlof & Kranton, 2005). Bourdieu’s framework
allows us to explore how teachers’ cultural and social background inﬂuences their
perceptions and preferences for schools and to understand how teachers approach the task
of ﬁnding a teaching job.

Bourdieu’s cultural reproduction theory may be particularly useful to deepen our
understanding of teachers’ self-selection into schools. Research on teachers’ school
selection has consistently found that the overwhelming majority of teachers choose to
teach in schools that are both similar to the ones they attended in terms of student

demographics and working conditions, and that are located in or very close to the

33

communities in which they grew up (Boyd et al., 2005a). Further, experimental evidence
suggested that teachers give higher ratings to job recruiters, such as principals, when they
are of the same racial and ethnic background (Young et al., 1997). Although they found
that teachers were receptive to recruitment messages that focused on the work
environment, they also found that teachers reacted more positively to vacant job positions
when they were presented by individuals of the same racial background. In short, the
research suggests that teachers’ school preferences are highly related to their own social
class and racial backgrounds. Teachers sort themselves into schools in ways that
reproduce existing social divisions.

Organization of the Dissertation

This dissertation is an initial attempt to use ideas from cultural reproduction
theory to understand teachers’ career decisions and experiences on the job search. In
doing so, it explores the nature of the teaching ﬁeld, teacher candidates’ habitus and
practical logic in the job search, and the implications of these processes for the overall
distribution of teachers across schools. The chapters are organized accordingly.

Chapter 2 describes the research design and data collection. The data come from
my longitudinal, mixed-methods study of 289 prospective elementary teachers who
responded to a survey, as well as interviews with 27 of these teacher candidates. Chapter
2 describes the sampling strategy and response rates for the survey component and the
sampling strategy and coding methods of the interview component. It also describes the
labor market context for teachers in this study and the metropolitan area in which it was

located.

34

Chapter 3 examines the nature of the teaching ﬁeld and its relationship to the ﬁeld
of schools. It explores teacher candidates’ perceptions of what it takes to get a teaching
job and how principals make hiring decisions. By comparing teacher candidate
perceptions to the literature on principal hiring decisions, the chapter examines which
teacher candidates have the most accurate perceptions of the teaching ﬁeld.

Chapter 4 focuses on teacher candidates’ habitus and the practical logic they have
about the job search. This practical logic informs how they approach the job search,
including what types of distinctions they make between schools and how they think about
districts and schools in the process. It explores the characteristics of schools that teachers
prefer and how they gather and use information about schools when making decisions
about where to apply and where to work.

Chapter 5 links the individual experiences and decisions of teacher candidates in
Chapters 3 and 4 to a broader sample of prospective elementary teachers. It traces the
progression of teacher candidates’ job search as they submit applications, receive offers,
and accept jobs, and examines the distribution of teachers across schools at each of these
stages.

Chapter 6 considers the role of school and district hiring practices in the teacher
labor market and examines the types of teachers that schools and districts hired. It also
explores the future plans of teacher candidates who did not end up with a teaching job.

Chapter 7 summarizes the main ﬁndings from the dissertation and discusses the
importance of this study in the larger research base on teachers’ career decisions. It
outlines the implications this study has for policies aimed at improving the distribution of

high quality teachers as well as practices of teacher educators and school administrators.

35

CHAPTER 2: METHODOLOGY

Teacher hiring is a seasonal phenomenon. Schools and districts usually begin
posting available positions and accepting applications around April. This continues
through September, with many teacher candidates not receiving a job offer until late
August. Districts try to predict the number of teachers they will need before students
arrive for the school year, but their predictions may be inexact. Thus, some teachers may
be hired in the early months of the school year. While some teacher candidates are hired
mid—year as positions become available, most teacher hiring takes place within this April
to September window. This study occurred during the hiring season to follow teacher
candidates throughout this process.

This dissertation uses a longitudinal mixed-methods design and consists of two
components. The longitudinal design is necessary to capture the processes through which
teacher candidates make career decisions. A mixed-methods approach is used to make
connections between teacher candidates’ individual experiences on the job market and
broader labor market outcomes. First, there is a longitudinal survey of 289 prospective .
elementary teachers. I surveyed prospective teachers who were currently applying for
teaching positions in elementary schools before the hiring season began. I then sent a
follow-up survey to these teacher candidates after the start of the school year. Second, I
interviewed a sample of 27 prospective elementary teachers at two points in the hiring
season. The ﬁrst interview took place near the beginning of the hiring season and the

second interview took place near the end. All data collection occurred within the

36

Smithton2 metropolitan area, a large Midwestern city and its surrounding communities.
Sampling from one metropolitan area provided a geographical boundary to the teaching
ﬁeld, thus, making it possible to identify the relevant capitals and the logic that structured
teacher candidates’ choices.
Survey Component

A longitudinal survey of 289 prospective elementary teachers was conducted to
identify patterns among teacher candidates. The initial survey took place in January,
February and March of 2006 through the teacher candidates’ education courses. Items
were designed to measure criteria used for evaluating available jobs, perceptions of
particular schools, type and source of information about schools, job search strategies,
perception of self as a candidate, range of schools to which they will apply, demographic
characteristics, and teaching qualiﬁcations. It included measures relevant to economists,
organizational sociologists, and cultural reproduction theorists. Information on
prospective teacher candidates’ educational and social background was gathered to
explore possible relationships between their background and preferences for schools. The
initial survey was distributed as a paper and pencil survey—mostly during the
respondents’ university teacher education classes. A copy of the survey can be found in
Appendix A. Some items on the survey were based on the Schools and Stafﬁng Survey3
(SASS) and the Teacher Pathways Project,4 in order to make connections with existing

research.

 

2 All names included in this dissertation are pseudonyms, including the name of the large city and
metropolitan area, the colleges and universities involved, and individual teachers interviewed.

3 Copies of the SASS Public School Teacher survey can be found at http://www.nces.ed.gov/surveys/sass.
’ Copies of the Teacher Preparation Program Graduates survey can be found at
http://www.teacherpolicyresearch.org/teacherpathwaysproject/surveys.

37

The follow-up survey took place from mid-September to October of 2006 and
asked where the teacher candidates applied, how many job interviews they had, where
they had job offers, and where they ended up teaching. It also asked about the
characteristics of the school they chose compared to other job offers they received. All
respondents from the initial survey were asked to complete the follow-up survey. A copy
of the follow-up survey can be found in Appendix B. Some items on the survey were
based on the SASS, the Teacher Follow-up Survey5 (TF S) and the Baccalaureate and
Beyond study,6 in order to make connections with existing research.

Sampling

The population sampled consists of individuals looking for teaching jobs in
elementary schools in a single metropolitan area. The sample needed to be limited to
either elementary or secondary teachers as state certiﬁcation laws distinguish between
elementary and secondary teacher candidates, creating distinct labor pools of potential
teachers. Further, state certiﬁcation requirements are subject matter-dependent at the
secondary level, with the result being that secondary science teachers have different
qualiﬁcations than secondary mathematics teachers and thus would be not applying for
the same jobs. I chose to focus on prospective elementary teachers because one might
expect these individuals to be qualiﬁed for similar jobs and thus able to apply for the
same jobs. Still, prospective elementary teachers may seek speciﬁc grade level
endorsements (i.e., early childhood) or subject matter endorsements (i.e., mathematics,

reading, social studies) on their teaching certiﬁcates, creating variation in their

 

’ Copies of the SASS Public School Teacher survey and the Teacher Follow-up Survey can be found at
httpzl/www.nces.ed.gov/surveys/sass.
6 Copies of the Baccalaureate and Beyond surveys can be found at http://nces.ed.gov/surveys/b&b/.

38

qualiﬁcations. Further, prospective teachers with an elementary certiﬁcate may be
qualiﬁed to teach in middle schools if they seek a subject matter endorsement. Despite
these complexities in the qualiﬁcations of prospective elementary teachers and in the
types of jobs they may pursue, candidates for elementary teaching positions have broader
qualiﬁcations and face a more uniform set of available positions than candidates in
secondary schools. For this reason, this study was limited to the labor market for
elementary school teachers.

The sampling strategy consisted of identifying teacher preparation programs that
tended to supply elementary teachers to this particular metropolitan area. Individuals
completing elementary teacher preparation programs from these institutions were
contacted to complete the initial survey. Speciﬁcally, the sampling strategy targeted
individuals who were completing their student teaching requirement (typically the last
requirement of teacher education programs) in elementary schools. Universities were
chosen to get a diverse range of programs and students. Six colleges participated in this
study by providing opportunities for me to distribute the initial survey to their student
teachers. Table 1 provides descriptive information on the colleges and universities
included. The sample of colleges included both public and private schools, as well as

schools of various sizes.

39

Table 1 - Descriptive Information on the Colleges Used for Data Collection

Percent Elementary

 

Total non- education
University Sector enrollmentl White2 graduates3 Selectivity"
River State Public 35,000- 20 300-400 More selective
University 40,000
Valley Public 20,000- 35 100-150 Most selective
University 25,000
Fairtown Public 5,000- 30 100-150 More selective
University 1 0,000
Oak State Public 15,000- 30 500-600 Selective
University 20,000
Holy Immaculate Private, Less than 90 50-100 Less selective
College Catholic 1,000
Capital State Public 20,000- 50 400-500 Selective
University 25,000

 

I Undergraduate enrollment information obtained from http://nces.ed.gov/globallocator/.
2 Race/ethnicity information was obtained from http://nces.ed.gov/globallocator/ and is
an approximation to mask the identity of participating colleges. This represents the
composition of the entire college, not necessarily that of the teacher preparation
program.

Number of students who passed the certiﬁcation test for elementary education.
Obtained from www.title2.org.
4 Obtained from http://www.usnews.com.

Data collection proceeded in one of three ways, depending on the level of
participation the college allowed. In the ﬁrst method, the respondents completed the
survey in my presence, during a workshop or seminar in which student teachers were
gathered and time was allotted for the survey administration. In the second, I distributed
surveys to student teachers during an established seminar or workshop, along with
stamped envelopes. The prospective teachers completed the survey on their own time and
put it in the mailbox. Finally, a third method occurred when the college provided me with
a list of names and addresses of elementary student teachers and I mailed the survey to

respondents. In this case, I also sent a reminder postcard and a second copy of the survey

to respondents that did not respond in a timely fashion. In each case, I followed the

40

principles of survey design and administration set forth in Dillman (2000) to make
administration across these three settings as similar as possible. Also, all individuals
asked to complete the survey were given $2 as an incentive for participation. This token
incentive was designed to create goodwill between the respondents and I (Dillman,
2000). Respondents kept the money whether or not they completed the survey.
Response Rates

Table 2 shows the response rates for the initial survey. The overall response rate
was 60.7%. The total number of responses was 327, although 36 responses were
considered unusable as respondents indicated in the screening question that they would
not apply for any elementary teaching positions that year. Most respondents removed
from the sample in this way indicated that they would not be completing their
undergraduate degree this year, decided to take a job outside of teaching, or would be
leaving the country (i.e., doing missionary work). Two responses were considered
unusable because the respondent answered too few questions. Therefore the ﬁnal sample

size was 289 prospective elementary teachers.

Table 2 - Number of Responses and Response Rates by College

 

 

Number Number Response Number Final
College disbursed returned rate not usable sample size
R‘Ye’ SW" 120 90 75.0% 2 88
Unrversrty
Valley University 55 54 98.2 1 53
F airtown University 52 36 69.2 0 36
Oak State University 252 96 38.1 29 67
Holy Immaculate 23 20 87.0 3 17
College
Capital State
University 37 31 83 .8 3 28
Total 539 327 60.7 38 289

 

41

Response rates varied by institution, most likely due to the differential access to
prospective elementary teachers provided to me by each institution. The response rates
ranged from 38.1% in Oak State University to 98.2% in Valley University. Oak State
University would only allow me to approach elementary student teachers through the
mail, resulting in a low response rate. Oak State University also had a large number of
respondents indicating that they would not complete their student teaching in Fall 2007 or
would otherwise not graduate by the start of the next school year. The large number of
unusable responses from Oak State suggests that the mailing list provided was not up to
date. As such, it may be that the low response rate is the result of survey recipients
assuming that a survey about prospective elementary teachers did not apply to them. Still,
the low response rate is cause for concern. To avoid possible non-response bias due to the
low response rate from Oak State, I ran two sets of analyses—one including Oak State
respondents and one excluding Oak State respondents. Unless otherwise stated, I report
the results of all the institutions including Oak State; for those instances in which
including Oak State signiﬁcantly and substantively changes the ﬁndings, I include results
of all institutions including Oak State and then all institutions excluding Oak State.

Demographic characteristics of the ﬁnal sample were compared to characteristics
of the state’s elementary teachers from the 2004-05 Registry of Education Personnel
(REP). The REP is an administrative database that tracks all public school and district
employees in the state. Table 3 compares the demographic characteristics of teacher
candidates in this study to all of the state’s elementary teachers in their ﬁrst three years of
teaching from the REP. Demographic characteristics for only those new elementary

teachers who earned their certiﬁcation from one of the six colleges included in this study

42

were also analyzed. The sample of teacher candidates included in this study has a similar
percentage of Aﬁican American teacher candidates as all new elementary teachers in the
state, although there are fewer White, non-Hispanic teachers and more Asian teachers.
There are a similar number of male teacher candidates compared to the general
population of new teachers in the state. The teacher candidates in this study are less likely
to have a master’s degree. This difference is likely due to the inclusion of teachers with

up to three years of experience in the REP.

Table 3 - Demographic Characteristics of Survey Respondents and All Elementary

Teachers in the State

 

 

 

 

 

Sampled

This study State database colleges only]
Characteristic Percent SD Percent SD Percent SD
Male 13.1 0.32 12.9 0.33 12.3 0.33
Master's degree 6.8 0.25 11.8*** 0.32 14.2*** 0.35
American Indian 0.7 0.08 0.3 0.05 0.3 0.06
Asian 4.3 0.23 0.6*** 0.07 0.9** 0.10
Afma‘.’ Amfr’can’ 10.0 0.30 8.0 0.27 11.6 0.32
non-Hrspanrc
White, non-Hispanic 80.4 0.40 90.1*** 0.30 85.9* 0.35
Hispanic 2.5 0.16 1.0 0.10 1.3 0.11
Native Hawaiian/
Paciﬁc Islander 0'7 0'08 n/a n/a
White and Asian 1.1 0.10 n/a n/a
mute. and ”man 0.4 0.06 n/a n/a
Amencan
N 289 4863 1647

 

Note: Data on the state’s teachers come from the 2004-05 REP. The data reported
here only include elementary teachers in their ﬁrst three years teaching.

1 Only includes new elementary teachers in the state database who earned their
certiﬁcation from one of the six colleges sampled in this study.

* p<.05, ** p<.01, *"‘* p<.001. These p-values come from two-tailed t-tests on the
difference between the REP data and this study.

43

The demographic variables in this study allowed teacher candidates to report
more than one racial or ethnic identity, although few did so. Due to the small numbers of
respondents who reported more than one racial/ethnic identity or who identiﬁed
themselves as Asian, American Indian, Hispanic, or Native Hawaiian/Paciﬁc Islander,
some racial groups were combined in the analyses. For analytic purposes, the individuals
who identiﬁed themselves as both White and Aﬁic’an American are included in the
African American subsample. Individuals who identiﬁed themselves as Asian, American
Indian, Hispanic, or Native Hawaiian/Paciﬁc Islander were combined into one category,
noted as “Other racial minority.”

Table 4 presents the item nonresponse rates, focusing on the items with a
nonresponse rate over 10 percent. Of the 180 items on the survey, 45 had a nonresponse
rate above 10 percent. The average rate of item-level nonresponse is 6.7%, although this
varied from 0.35% to 52.9%. The two items with the highest nonresponse rates are the
items in which respondents named their top choice and least desirable schools. It is likely
that this nonresponse rate is an indication that teachers had difﬁculty naming speciﬁc
schools, as most of these teachers did name a district. Some respondents who did not
answer this item wrote on the margins that they could not think of a school within the
district, supporting this explanation of nonresponse on these items. Likewise, the item in
which teacher candidates were asked to name a district in which they do not want to work
also had a high nonresponse rate. Some of this nonresponse is probably due to
respondents being unwilling to eliminate any district from their job search. Many teacher

candidates who did not answer this item wrote that they would work anywhere. For these

44

reasons, the item-level response rates to these three items are considered pieces of data,

rather than indications of missing data.

Table 4 — Items With Highest Nonresponse Rate

 

 

Range of
Item Item description Missing % missing
Write the name of a speciﬁc school in which you 0
11a would NOT want to teach. School name 5294 A) n/a
Write the name of a speciﬁc school in which you
5a would most like to teach. School name 39°79 n/a
21c Features you feel will most HURT your own chances 31 83 n/a
of obtaining a teaching job, Third feature '
Features you feel will most HURT your own chances
21b of obtaining a teaching job, Second feature 2630 n/a
15c What features make this school LEAST desirable to 24 22 n/a
you, Third feature '
38b ACT score 23.53 n/a
15b What features make thrs school LEAST desrrable to 2111 n/a
you, Second feature
1 5a What features make thrs school LEAST desrrable to 2007 n/a
you, Frrst feature
Features you feel will most HURT your own chances
21a of obtaining a teaching job, First feature 19°03 n/a
11b Write the name of a speciﬁc school in which you 1 6 61 n/a
would NOT want to teach. District name '
. Average for sources of information on least desirable 15.22 to
16a-1 15.38
school 15.92
12 Academic performance of least desirable school 15.22 n/a
13 Community type of least desirable school 14.53 n/a
14a-w Average for characteristics of least desirable school 12.97 K5429“)
36 Name of high school you graduated from 12,46 n/a

 

There are four overall trends in the item-level nonresponse rates. First, the items

about respondents’ individual characteristics that would most hurt their ability to ﬁnd a

job had high levels of nonresponse. It may be that these items confused respondents.

Some individuals who did respond wrote in how they interpreted the question, indicating

that respondents did have trouble understanding the question. For this reason, these data

45

are not used. Second, the items in which respondents were asked questions about their
least desirable school had high levels of nonresponse. As noted above, some respondents
could not identify any school in which they did not want to teach, and thus they left this
section of the survey blank. Third, the items with the largest nonresponse were the items
at the end of the survey, suggesting that some individuals did not ﬁnish the survey.
Finally, the ACT score had a high level of nonresponse, probably because individuals
who took the ACT for college entrance did not remember their score.
F ollow- Up Survey Response Rates

The follow-up survey was sent to all respondents of the initial survey who
provided contact information, either a mailing address or email address. The follow-up
survey was initially distributed through an online survey website, Survey Monkey.7
Respondents that provided both an email address and mailing address in the initial survey
were sent a letter in mid-September through the mail alerting them to an email message
they would receive about the follow-up survey. This pre-notice letter was sent to increase
respondents’ awareness of the survey and thus improve response rates (Dillman, 2000).
Teacher candidates received an email a week later with a link to complete the follow-up
survey online. Teacher candidates were sent two email reminders to complete the survey
online. In mid-October, teacher candidates who had not yet completed the follow-up
survey were sent a paper copy of the survey through the mail. Nineteen teacher
candidates only provided mailing addresses or had email addresses that were not
deliverable. These respondents were sent a similar number of contacts, but all occurred

through the mail.

 

7 www.5urveymonkey.com

46

The follow-up survey was administered primarily through the Internet for two
main reasons. First, it included questions that were only applicable to a segment of the
respondents depending upon their progress through the job search. For example, all
teacher candidates were asked a similar set of questions about where they applied, but
only those respondents who received at least one job offer were asked additional
questions about their job offers. Teacher candidates who did not receive any job offers
thus completed a shorter follow-up survey. Also, teacher candidates who received more
than one job offer answered an additional set of questions comparing their various offers.
Using an Intemet-based survey program facilitated the navigation of teacher candidates
through these skip patterns so that respondents were only presented with questions that
pertained to their situation. Second, the majority of respondents were college students
about to graduate and would likely change their mailing address. As the colleges in this
study allowed their students to use university email accounts for several months
(sometimes years) after they graduated, it was more likely that the respondents could be
reached by email rather than postal mail.

Table 5 provides the sample sizes and response rates to the follow-up survey by
college. The overall response rate for the follow-up survey was 59.2%, although this
varied from a low of 38.9% at Fairtown University to a high of 75.0% at River State

University. The follow-up survey had a ﬁnal sample size of 171 teacher candidates.

47

Table 5 - Number of Responses to F ollow-Up Survey and Response Rates by College

 

Initial Follow-up Returned Returned Final
sample surveys through through sample Response

 

College size disbursed Internet mail size rateI
RN.“ St.“ 88 87 54 12 66 75.0%
Unrversrty

Valley University 53 53 19 6 25 47.2
Fairtown University 36 35 13 l 14 38.9
Oak State University 67 67 37 6 43 64.2
Holy Immaculate 17 15 7 2 9 529
College

C’P’m’ .S’a’e 28 28 11 3 14 50.0
Unrversrty

Total 289 285 141 30 171 59.2

 

1 The response rate was calculated by dividing the ﬁnal follow-up sample size by the
initial survey sample size.

Table 6 presents response rates to the follow-up survey by teacher candidate
demographic characteristics. To test whether the response rates differed by teacher
characteristic, chi-square statistics were computed. Male teacher candidates and African
American teacher candidates were less likely to respond to the follow-up survey than
their female, White, or other racial minority peers. Teacher candidates with and without a
master’s degree were equally likely to respond to the follow-up survey, as were new
college graduates who are entering teaching as their ﬁrst career and individuals who are
switching into teaching from another career. The low response rate for male and African
American teacher candidates indicates caution should be used in making inferences about

these groups using the follow-up survey data.

48

Table 6 - Number of Responses to Follow-Up Survey and Response Rates by

Demographic Characteristics

 

Initial Follow-up Response Chi-

 

Characteristic survey survey rate square Prob.
Gender

Male 37 14 37.8% 8.47 .004
Female 246 155 63.0

Highest degree

Bachelor’s degree 261 155 59.4 .10 .746
Master's degree or above 19 12 63.2

Career status

New college graduate 183 110 60.1 .01 .933
Mid-career changer 99 59 59.6

Race/ethnicity

African American, non 29 11 37.9 7.95 .019
Hispanic

White, non-Hispanic 226 139 61.5

Other racial minority 26 19 73.1

 

Table 7 presents item-level nonresponse rates for items on the follow-up survey,
focusing on the items with a nonresponse rate over 10 percent. Of the 81 items on the
survey, 7 had a nonresponse rate above 10 percent. The average rate of item-level
nonresponse is 3.3%, although this varied from zero nonresponse to 33.3%. The two
items with the highest nonresponse rates are the items that only were relevant for teacher
candidates who had turned down all of their job offers. Few teacher candidates did not
accept any of the jobs they were offered and the high nonresponse to these items may be
due to the small numbers of teacher candidates involved. Items about the industries of the
respondent’s parents also had higher levels of nonresponse. Teacher candidates were

asked about both the occupation and industry of each parent and respondents were more

49

likely to provide the occupation of each parent, even if they did not provide the industry.

Respondents were also more likely to omit their salary information.

Table 7 - Overall Item-Level Nonresponse Rate and Items With Highest Nonresponse for

the Follow-Up Survey

 

 

Item Item description Missing
F 18 What will you most likely do during the 2006-07 school year? 33 3%

(Asked of those who did not accept any teaching job offers.) '
F17 Why did you turn down your job offer(s)? (Asked of those 33 3

who did not accept any teaching job offers.) '
F3industry What is your mother’s industry? 29.2
F22 What is your base teaching salary? (Asked of those who 21 1

accepted a job offer.) '
F2industry What is your father’s industry? 17.5
F28c What are the three features that most inﬂuenced your decision

to accept the teaching job at your chosen school rather than
other schools in which you had job offers? Third most
important feature. (Asked of those with more than one offer.)

10.0

 

Analytic Methods

Details on speciﬁc analytic procedures used will be discussed in the relevant
chapters. In general, this dissertation uses descriptive frequencies, factor analysis, and
multivariate regression techniques to analyze the structure of the teaching ﬁeld, describe
teacher candidate behavior, and analyze patterns in their preferences and ultimate labor
market outcomes. The survey included measures of different types of capital, including
social capital, embodied cultural capital, institutionalized cultural capital,8 and human
capital. These measures were used to understand the types of capital teacher candidates
think is most valued on the teaching ﬁeld and as independent variables associated with

labor market outcomes. Data on teacher candidates’ preferences for schools and

 

8 For the distinction between embodied and institutionalized cultural capital, see Bourdieu (1986). I explain
this distinction more fully in Chapter 3.

50

construction of job alternatives was analyzed by teacher characteristics, such as gender,
race, social class background, and academic background. The follow-up survey was used
to make connections between teacher candidates’ initial preferences and the school in
which they ultimately ended up working.

Both the initial survey and the follow-up survey asked teacher candidates to name
speciﬁc districts or schools. The initial survey asked teacher candidates to name their
most desired and least desired schools and the follow-up survey asked them to name
where they applied and received offers. Data on these districts and schools were obtained
from the Common Core of Data (CCD) or Private School Universe Survey (PSS) through
the National Center for Education Statistics. The percentage of students eligible for free
or reduced-price lunch was obtained from the CCD and used as a measure of the
socioeconomic status of students in the school or district. The racial composition of the
student population, overall school size, pupil—teacher ratio, zip code of physical location,
and urbanicity were also obtained from the CCD. Data on per-pupil expenditures was
obtained from the CCD ﬁnance survey. The zip code of the teacher candidate’s home
address at the time of the initial survey, as provided by the teacher candidate, was used to
calculate geographic distances for applications and job offers.

There is no national source of data on beginning teacher salaries and so salary
information for districts was obtained from several statewide databases. The primary
source for salary data came from the actual salaries of new teachers in the Registry of
Education Personnel (REP). The average salary of new teachers in each district was used
to calculate the expected beginning teacher salary in that district. This data source

provided expected beginning salaries for most districts named by teacher candidates as it

51

came from the state that is the focus of this study. Information on expected beginning
teacher salaries in districts in other states came from the following websites:
http://www.ed-data.kl2.ca.us/, http://www.nctq.org/cb/, and
http://www.teachinﬂorida.com/Jobcenter/ﬂteachersalaries.asp. In all cases, information
on the annual salary for the lowest step on the salary schedule or the annual salary ﬁrst-
year teachers with a bachelor’s degree was used as the measure of the beginning teacher
salary.

Another strategy to examine the relationship between teacher candidate
characteristics and characteristics of where they applied consists of looking at the
characteristics of teacher candidates that applied to different districts. For this reason, I
restructured the dataset to look at patterns in the districts to which teacher candidates
applied. Rather than have one observation per teacher candidate, this dataset has one
observation per teacher candidate-application match. Variables in the dataset include
characteristics of the school or district where the teacher candidate applied and
characteristics of the teacher candidate. The 161 teacher candidates who remained in the
labor market and applied to at least one district named a total of 805 districts and/or
private schools or charter schools to which they applied.

One limitation of this approach is that it under-represents the applications of
teacher candidates who applied to many districts. The survey provided space for teacher
candidates to name seven districts.9 Nine teacher candidates wrote in multiple districts

per space. Eighty-four teacher candidates (52%) named fewer districts than spaces

 

9 Providing seven spaces for names of districts was a compromise between naming all districts and limiting
response burden. In the ﬁrst round of interviews, teacher candidates were asked to name speciﬁc districts
and could easily name 5-6 districts, while it took more time to remember a larger number of districts.

52

provided, suggesting that providing more space would not have induced half of the
teacher candidates to name more districts. This restructured dataset was used for analyses
in Chapters 5 and 6. In Chapter 5, it is used to explore correlations between teacher
candidate and district characteristics. In Chapter 6 it is used to analyze which applications
resulted in job offers.

Interview Component

The qualitative component of this study consists of interviews with 27 prospective
elementary teachers at two points during the teacher hiring season. The ﬁrst interview
took take place in March, April, or early May of 2006 and covered topics such as the
teacher’s social and educational background, their sense of themselves as a candidate,
criteria for applying to schools, job search strategies, and methods for gathering
information about potential schools. The second interview took place in July or August
2006, and covered topics such as evaluations of particular schools, experiences
interacting with school and district staff through interviews and reasons behind the
teacher’s ﬁnal job choice. Half of the teacher candidates did not consider their job search
over by the second interview and were interviewed again in early September.

The interviews occurred in a time and place convenient for the participants. Most
interviews were conducted in a university library, local public library, or coffee house,
although a few were in the teacher candidates’ elementary school, home, or other
workplace. Several teacher candidates indicated during the interview that they associated
me with this university; this is important because participants’ responses may have been
inﬂuenced by their perception of my social position: a young, White researcher from a

major university. Interviews lasted approximately 45 minutes to one hour, and were taped

53

and transcribed for analysis. A semi-structured interview protocol provided initial
questions and prompts for the interview, but I adapted the interview questions to allow
the participants to respond in ways that made sense for their experiences. The second and
third interviews also included questions about particular themes or experiences
mentioned by participants in the ﬁrst interview. The interview protocols can be found in
Appendix C.
Sampling

All survey respondents were invited to participate in the interview component of
this study. Slightly more than one-fourth of survey respondents (26.3%) volunteered for
an interview. There were no signiﬁcant differences in the racial or ethnic background,
gender, career status, or high school location of survey respondents who did or did not
volunteer. Respondents who volunteered to be interviewed were somewhat more likely to
come from Holy Immaculate College and River State University and less likely to come
from F airtown University. Among those who agreed, I compared demographic
characteristics from the initial survey to ensure a diversity of backgrounds and
experiences in the interview sample. I considered the college or university the
prospective teacher attended, his or her racial and ethnic background, gender, and career
status when choosing the interview sample. Although I tried to balance characteristics of
teachers across universities, this was not always possible due to the different composition
of prospective student teachers in each university. For example, most student teachers in
River State University are young adults just graduating from college, while those in Holy

Immaculate’s teacher preparation program tend to be switching into teaching from other

54

careers. Table 8 shows the demographic characteristics of prospective elementary

teachers in the interview sample.

Table 8 - Descriptive Information on the Interview Participants

 

Social class

 

Teacher College Race Gender Career status background
Stephanie River State White Female College graduate Mid-high
Claire River State White Female College graduate Middle
Heather River State White Female College graduate Middle
Daniel River State White Male College graduate Mid-low
Rachel River State Hispanic Female College graduate Middle
Melissa Valley White Female College graduate Middle
Courtney Valley White Female College graduate Mid-high
Ryan Valley White Male College graduate Mid-high
May Valley Asian Female College graduate Mid-low
Samantha Fairtown White Female College graduate Mid-low
Daija F airtown Aifnfgr'cizgn Female College graduate Mid-low
Valerie Holy Immaculate White Female Mid-career Mid-low
Denise Holy Immaculate White Female Mid-career Mid-low
Richard Holy Immaculate White Male Mid-career Middle
Nicole Holy Immaculate Afrrcan Female Mid-career Mid-low
Amerrcan
Vanessa Holy Immaculate Afrrcan Female Mid-career Low
Amerrcan
Kelly Oak State White Female College graduate Middle
Debra Oak State White Female Mid-career Middle
Monica Oak State White Female Mid-career Middle
Amanda Oak State White Female College graduate Middle
Jennifer Oak State Hispanic Female Mid-career Mid-low
Zoe Oak State 31:13:; Female College graduate Middle
Wayne Capital State White Male Mid—career Mid-low
Chris Capital State White Male Mid-career Low
Joshua Capital State White Male Mid-career Middle
Kia Capital State African Female Mid-career Mid-low
Amencan
Jasmine Capital State Asian Female College graduate Low

 

Teacher candidates’ race or ethnicity, gender, and career status come from their

responses on the initial survey. I categorized teacher candidates by social class based

55

upon their parents’ occupation and educational background, obtained during a discussion
of their background during the ﬁrst interview. Occupation, educational level, and income
are common measures of one’s social class (Gilbert, 1998; Nakao & Treas, 1992).
Teacher candidates in the lower class category have parents who did not graduate from
high school and held a variety of temporary, part-time work or periods of unemployment.
The mid-low category consists of teacher candidates whose parents completed high
school and may have some college education. Their parents also had low-level factory or
clerical jobs. Teacher candidates were categorized as middle class if their parents had
bachelors’ degrees and semi-professional occupations or skilled trades. Participants
whose parents were teachers generally fell in this category. The mid-high category
consists of teacher candidates whose parents had advanced degrees and professional
occupations.
Coding and Analysis Methods

The interviews attempted to capture teachers’ preferences for schools in general
and speciﬁc ways. First, teacher candidates were asked to list features of schools that
were important in their job search. Second, teacher candidates were asked to name and
talk about speciﬁc schools in which they wanted to work and did not want to work and
give their reasons. This dual questioning strategy was designed to help teacher candidates
talk about the details of their job search decisions. A holistic reading of the interview
transcripts suggested that teacher candidates offered different reasons for choosing
schools in response to these questions. That is, one teacher candidate, in the course of one
interview, gave different information about her preferences when answering different

questions.

56

For this reason, each criteria discussed by teacher candidates was coded as either
an espoused preference or a preference in use, depending on the context of the teacher
candidates’ response. Preferences in use are responses to speciﬁc questions in which
teacher candidates were asked to name schools in which they are or are not applying and
to explain those decisions. They are the characteristics of schools that teacher candidates
mentioned when they talked about why they preferred or avoided particular schools and
districts. Espoused preferences, on the other hand, are teacher candidates’ responses in a
general discussion about the features of schools that are important in their job search.
Espoused preferences are the features teacher candidates provided when they were asked
about what characteristics were important and how they compared schools to one another
in general. This method of distinguishing between espoused preferences and preferences
in use is consistent with the research by Agyris and Schon (1975) when they developed
this framework. In their work with organizational leaders, they asked participants to both
write down their goals and assumptions in an abstract situation (to capture their espoused
theories) and to describe the reasons and assumptions behind their actions in a speciﬁc
situation (to capture their theories in use).

The interviews were transcribed and analyzed at two levels. The ﬁrst stage of data
analysis involved creating cases for each teacher (Miles & Huberman, 1994). After the
ﬁrst interview, an initial analytic memo was prepared for each teacher that highlighted
the processes and preferences of each teacher candidate. Themes from these analytic
memos were discussed with the participants to check the validity of interpretations used

by the researcher (Maxwell, 1992).

57

For the second level of data analysis, the interviews were coded using a constant
comparative method (Glaser, 1965) to identify central themes in teacher’s decision-
making and experiences. The initial coding scheme focused on creating codes that
reﬂected themes from the various theoretical frameworks, including traditional labor
market studies, organizational sociology, and Bourdieu’s cultural reproduction
framework. Labor market codes focus on teachers’ statements about salary, beneﬁts, and
various working conditions of the schools they considered. Organizational sociology
codes focus on statements about teachers’ interactions with principals and other school
staff to determine the extent to which teachers get a sense of the work environment
within the schools and supports offered to new teachers. To capture cultural reproduction
processes, I used codes that captured how teachers perceive their position on the teaching
ﬁeld, how they deﬁned the school features that are important in their decision-making,
how they perceived the schools in which they applied and interviewed in relation to other
schools, and their perception of the structure of the school ﬁeld. Teachers’ perceived
position in the teaching ﬁeld are based on several types of statements, including explicit
comparisons teachers make between themselves and other teacher candidates; statements
teachers make in relation to their ability or inability to get a job in a particular school;
teachers’ beliefs about what schools look for when hiring teachers, and factors in their
preparation that help them get a job.

Initial coding and analysis required revisions to the list of codes to extend existing
themes, create distinctions within heavily used codes, and ﬁll in emergent themes not
present in the initial coding scheme (Miles & Huberman, 1994). In particular, teacher

candidates offered reasons for preferring or avoiding particular schools that were not

58

suggested by existing research and additional codes were created to reﬂect these other
criteria. The interviews were coded using Atlas.ti software and allowed me to compare
the criteria teacher candidates used in their job search with whether or not the criteria
were provided as an espoused preference or preference in use. After the codes were
created, another coder and I independently coded 12 interviews. We agreed 76% of the
time. In cases where the two coders disagreed, we talked about the discrepancy and came
to a consensus about how the interview should be coded. To check internal consistency
and possible drifting of codes over time, I coded 20% of the remaining interviews twice.
This internal consistency agreement was 93%. These reliability measures are in line with
standards for qualitative analysis (Miles & Huberman, 1994).

Data analysis proceeded with particular attention paid to making connections
between the qualitative and quantitative data. Relationships or themes that emerged from
the interviews were used to analyze the survey data. Likewise, when signiﬁcant patterns
become evident through the survey analysis, I searched for underlying causes behind
those patterns in the qualitative data.

Advantages and Disadvantages of Research Design

One advantage of this research design is that it relies on data collected from
prospective teachers currently looking for a teaching job. Most research on teachers’
career decisions uses retrospective data about teachers’ previous decisions, or it focuses
on reasons for teacher turnover. It is possible that reasons teachers leave a school are
different from the factors that inﬂuence their initial match to their ﬁrst teaching job.
Some researchers may consider this sampling strategy a disadvantage as I do not know in

advance which prospective teachers will ultimately be offered jobs, and of those offered

59

jobs, what types of schools the teacher candidates will choose. I view this as an
advantage, however, as interviewing teachers about their decisions about where to apply
and understanding their perceptions of schools throughout the application and interview
process still provides valuable information about their preferences for schools and the
processes involved in ﬁnding a job, even if they are not ultimately offered a job.

Further, exploring how prospective teachers conceive of their job alternatives in
the face of no full-time job offer also provides insight into how teachers make career
decisions. My pilot study suggested that, when teachers did not have a job offer by late
August, they either made plans to substitute teach in a desirable district, expanded their
search to include different types of schools, or began looking for non-teaching jobs.
Understanding why teachers make different decisions when faced with poor job prospects
may help to understand how teachers conceive of their job alternatives and their
preferences for particular types of schools.

One disadvantage of this sampling strategy is that it emphasizes a particular type
of newly hired teacher for schools and districts. While districts frequently hire new
teachers, these new hires may fall into four categories: newly prepared graduates of
certiﬁcation programs, other ﬁrst-time teachers who have delayed entry into teaching
after receiving their initial certiﬁcation, teachers transferring from other districts, and
teachers returning to schools after taking a break from teaching (US. Department of
Education, 2000). This sampling strategy relies on newly prepared teachers. Over time,
however, districts appear to be relying more on newly prepared teachers and delayed

entrants, rather than transfers or returning teachers. For example, ﬁrst-time teachers made

60

up only 30% of new hires in 1987-88, but 45% of new hires in 1993-94 (US. Department
of Education, 2000).

This sampling strategy also relies on prospective teachers who complete their
teacher training in university-based programs. As such, the sample is limited to more
traditional types of teachers, rather than teachers who enter the profession without
completing a university preparation program. Understanding the preferences of
traditionally-prepared teachers is not a bad thing, as they continue to be the main source
of new teachers. In particular, as this study is based in a state that is a net supplier of
teachers, emphasizing university-based programs is less problematic as most new
teachers go through this route. Still, it represents one limitation of the generalizability of
this study. Although this study does not include teachers from alternative certiﬁcation
programs, the colleges involved did have a combination of undergraduate, post-
baccalaureate, and master’s level certiﬁcation programs. Thus, the respondents did come
from a variety of training backgrounds, even though all were university-based.

Teacher Supply and Demand in the Smithton Metropolitan Area

As teachers’ career decisions may be inﬂuenced by the conditions of teacher
supply and demand, the general labor market, and the structure of the education system, it
is necessary to understand the speciﬁc context of this study. The research design focuses
on teacher candidates in one metropolitan area surrounding Smithton, a large,
Midwestern city. The Smithton economy has relied on traditional manufacturing jobs for
many years and the rising instability of the manufacturing sector has taken its toll on the
city. In particular, the economic woes of several large manufacturers based in the

Smithton area led to record job losses as the companies cutback on production and

61

eliminated jobs. Due to the size of these companies and the state economy’s reliance on
this industry, the impact was felt throughout the state. In 2006, Smithton and its
surrounding area continued to suffer from a sluggish economy.

Smithton’s population of about 900,000 has been declining as residents move out
of the city into the surrounding suburbs. This population shift has continued for several
decades and as Smithton’s residents move into the immediate suburban communities,
residents of those communities move into more distant suburbs. These demographic
trends are exacerbated by the rise in public charter schools and inter-district choice
programs in which even more students exit the Smithton Public School District. Charter
schools have operated in the state for over 10 years. There are now over 200 charter
schools in the state, many of which are clustered around Smithton. This expanding sprawl
has had implications for schools in and around Smithton. Districts are funded on a per-
pupil basis from the state, so districts losing students through school choice options or
residential relocation also lose ﬁmding. The result is that the Smithton Public School
District has lost signiﬁcant amounts of revenue in the past decade, forcing the district to
close schools and layoff teachers. The district has closed schools each year for the past
several years.

The ﬁnancial troubles of public school districts affects more than the Smithton
Public School District itself. Surrounding districts also face ﬁnancial difﬁculties as the
state per-pupil funding grant has stagnated in the past couple years while health care costs
have risen, leaving districts struggling to ﬁnd ways to keep costs down. Further, the
population shift as families moved from the communities closest to Smithton to those

farther away has created substantial demographic changes in some communities in recent

62

years. Manufacturing declines have forced layoffs by major employers in the area and
many residents are following jobs out of state. In the past couple years, it has become
common for districts to give pink slips to all of their ﬁrst year teachers near the end of the
year. While many of those teachers are hired again during the summer, it creates
instability in the teacher labor market.

Overall, the Smithton area has an over-supply of teachers and the job prospects
for new teachers looked dim, as there are many more new teachers than there are open
positions. The labor market conditions are clearly inﬂuenced by enrollment patterns
induced by the trends outlined above. Further, the state is historically an exporter of new
teachers, with many newly prepared teachers leaving the state for other regions in the
nation with an overall shortage of teachers. Still, the state bureau of labor rated
elementary teaching among the occupations with the largest number of annual job
openings. One recent trend that eases the job market for new teachers is the presence of
retirement incentives for experienced teachers. Like the rest of the country, the state
teaching force is graying, with many teachers hanging on to jobs until they retire. As
district budgets struggle to pay teachers at the top of the salary schedule, many districts in
the area have offered incentives for experienced teachers to retire, creating openings for
new teachers.

The teacher hiring season in the Smithton area begins in early April when all
colleges in the area host teacher job fairs. Districts from around the area—and some from
around the country—are brought to campus to recruit potential teachers. While a few
districts post vacancies and accept applications before the job fairs, these fairs usually

represent teacher candidates’ ﬁrst opportunity to submit applications and talk to hiring

63

ofﬁcials. Representatives from schools or districts often hold mini—interviews with
teacher candidates at the job fair and collect resumes. There is a combination of districts
in the area with centralized or decentralized hiring practices. While most districts have at
least a centralized ﬁltering process to assess applicants’ qualiﬁcations, many districts
give principals and individual schools signiﬁcant hiring authority.

Nearly all districts in the area use online application programs in which
candidates submit application materials over the Internet. These online applications are
located on both district human resource websites and on county job banks. The Smithton
metropolitan area consists of three main counties, which each list vacancies in districts
throughout the county. However, not all districts participate in their county job bank, thus
the three county job banks do not have all openings in the area. Further, charter schools
are not part of these county job banks. There is also a statewide job bank in which teacher
candidates can upload application materials to districts throughout the state, but district
participation in this state job resource is also sparse. For example, about at the start of the
job search season, about 35 districts and/or charter schools participated in this online job
bank.

Smithton Public School District

As this study is particularly concerned with the distribution of teachers across
schools and how teachers think about working in urban schools that serve high
proportions of poor and minority students, some description of the Smithton Public

School District is in order. The Smithton School District enrolls around 115,000 students,

64

although overall enrollment has dropped 40% in the past ten years. ’0 The district receives
about 70% of its revenue from state sources and faces state restrictions in raising
additional local funds. Due to concerns about ﬁnancial mismanagement, the state took
control of district operations and the district returned to a traditionally elected school
board in 2005. About 90% of students enrolled in the district and 85% of the city’s
population under age 18 are African American. About 70% of Smithton students are
eligible for free or reduced-price lunch. About 80% of the adult population has completed
high school and 25% have a college degree. The city has an unemployment rate of 7.8
and the median household income is about $30,000. This compares to a national
unemployment rate of 5.8 and a national median income of $42,000.

The Smithton District has been particularly hampered by the enrollment patterns
throughout the region. The district has consistently lost students—due to school choice
and residential moves—as well as the funding that accompanies each student. The
Smithton teacher union went on strike in the fall of 2006 as the teachers in this study
were preparing to start their ﬁrst year of teaching. Smithton schools canceled the ﬁrst few
days of the new school year. As a result, many parents enrolled their children in other
districts or charter schools, causing Smithton to lose more students than expected and
freeze new teacher hiring.

As noted above, the district has been forced to close many schools and lay off
teachers. Yet, Smithton does hire new teachers each year and some indicators suggest
that Smithton has difﬁculty hiring fully certiﬁed teachers. For example, the state Highly

Qualiﬁed Teacher Report indicated that Smithton employed over 200 teachers who were

 

’0 District information in this section comes from the National Center for Education Statistics online district
proﬁle. Community information in this section comes from the US. Census Bureau website.

65

not highly qualiﬁed. Further, many charter schools in Smithton and the immediate
suburbs reported over half of their teachers were not highly qualiﬁed, suggesting that
some schools encounter stafﬁng difﬁculties despite an overall large supply of new
teachers. Smithton’s problems in recruiting teachers may be supported in part by
portrayals of the city in the local news. For example, the Smithton Public School District
was in the news for several attacks on teachers in public schools, including an attack on
an elementary teacher. The resulting focus on security in Smithton schools played itself
out in the major area newspapers and media outlets.

Although the entire metropolitan area is quite diverse when considered as a
whole, the suburbs surrounding Smithton are highly segregated. The main counties
around Smithton are predominantly White. Most of the African Americans who live in
the area live within Smithton’s city boundaries, though several suburban towns close to
Smithton also have large proportions of African American residents. There is also a large
population of Arab Americans. The suburbs are generally divided into a group of older,
more diverse suburbs immediately surrounding Smithton and younger suburbs farther out
from Smithton. The older suburbs tend to be less affluent and the younger suburbs are
more affluent and predominantly White. These younger suburbs formed during
successive population shifts as residents move farther away from Smithton.

Colleges and Universities in this Study

Six colleges and universities that prepare elementary teachers are included in this
study. These institutions were chosen because they are located close to the Smithton area
and traditionally supply new teachers to schools in and around Smithton. While the

colleges must design their teacher certiﬁcation programs to meet state requirements, the

66

structure of the programs and the populations served by each institution do vary. As such,
recognizing the signiﬁcant aspects of each institution is important to understanding how
teachers graduating from these institutions think about their teaching careers.

Table 9 provides descriptive information on the teachers in this study by the
institution from which they were contacted. As is evident from this table, the
demographic characteristics of teachers from each college vary dramatically. Except for
Oak State University, response rates from each college were reasonably high. This
suggests that the differences in teacher characteristics between the colleges stem from
real differences in the population of student teachers in each college, rather than from
differential nonresponse. In particular, Holy Immaculate College and Capital State
University have many more African American student teachers than the other institutions.
Teachers graduating from these two institutions also tend to be older, male, and switching
into teaching from another career.

Capital State University is located near downtown Smithton and is a large urban
research university that prides itself on being accessible to all students. The average age
of students in the program is 35, reﬂecting the fact that many students are entering
teaching as a second career. The program serves a nontraditional population of teachers, a
large African American population, and almost a third of the students are male. Capital
State teachers also have the lowest incoming ACT scores. Capital State offers both
undergraduate and graduate teacher certiﬁcation programs and emphasizes the needs of
urban educators in its programs. Capital State also offers a provisional certiﬁcate
program aimed at mid-career changers in which individuals with full-time positions in

Smithton may complete certiﬁcation requirements while teaching.

67

Table 9 - Demographic Characteristics of Teacher Candidates From Each Institution

 

Teacher River Valley F airtown Oak Holy Capital
characteristic State University University State Immaculate State
Age 23.9 23.3 28.3 27.7 36.3 35.0
ACT score 24.2 27.6 22.5 23.1 21.8 20.8
Race/ethnicity Percent Percent Percent Percent Percent Percent
“lime: .m’n' 86.4 82.7 91.2 87.5 35.3 53.9
Hrspanrc

Aﬁican

American, non 3.4 3.9 5.9 3.1 58.8 38.5
Hispanic

the‘.’a°‘a’ 10.2 13.5 2.9 9.4 5.9 7.7
mrnorrty

Graduated

“mom“ 11 137 29 141 250 148
state high ' ' ' ' ' '
school

M"”"”""” 8.0 5.8 48.6 53.1 88.2 88.5
changer

Male 11.4 5.7 8.6 14.1 23.5 30.8
Master's degree 0.0 0.0 0.0 6.4 68.8 15.4

 

Valley University is the ﬂagship campus of the state’s public university system.

Over 82% of Valley’s teachers are White, with the next largest racial group coming from

Asian backgrounds. Few Valley teacher graduates are male and Valley teachers have the

highest incoming ACT scores. Valley also stands out among the colleges in this study as

having a large percentage of teachers from other states but mostly young new college

graduates. While Oak State, Holy Immaculate, and Capital State have large percentages

of teachers who grew up in other states, they also have older teachers who are switching

into teaching from another career. Valley also attracts traditional age college students

from other states. Compared to the rest of the university, Valley’s undergraduate

68

education program is small. Valley also offers an intensive full-time graduate
certiﬁcation program designed for mid-career changers.

Oak State University is a large comprehensive public university and has one of
the largest teacher preparation programs in the country. It is located in a large city close
to Valley University. Oak State teachers have an average age of 27.7 and over 87% of
Oak State teachers are White. Slightly more than half of Oak State teachers are switching
into teaching from another career. Oak State has undergraduate, postbaccalaureate, and
graduate programs to earn certiﬁcation and education is among the largest programs at
Oak State. Oak State has partnerships with schools in the smaller urban area in which it is
located as well as schools in suburban Smithton. It also has an urban education program
that prepares individuals for teaching in Smithton.

F airtown University is part of the same public university system as Valley, but the
campus operates independently. F airtown is located closer to Smithton and is just outside
the city borders. Nearly all of F airtown teachers graduated from high schools within the
state. Fairtown has the highest percentage of White teachers, with over 91% of teachers
being White. Fairtown offers undergraduate and graduate certiﬁcation programs,
including online master’s programs. The teacher preparation programs are among the
largest programs in the university.

River State University has a nationally ranked teacher preparation program.
Although it is located about 80 miles from Smithton, the teacher preparation program
places a substantial proportion of its student teachers in schools in and around Smithton,
and offers classes for these student teachers in a local school. Most students are White

and are new college graduates. Very few River State students attended high school

69

outside of the state. River State is unique among the colleges in this study for providing
prospective teachers with a full year of student teaching experience in the same
classroom.

Holy Immaculate College is a small, Catholic college in Smithton with a strong
commitment to social justice. Holy Immaculate students are the oldest and most likely to
be African American among the colleges in this study. They are also most likely to have
graduated from an out of state high school and over 88% of Holy Immaculate students
are switching into teaching from another career. Teacher education is the largest area of
study at Holy Immaculate. Holy Immaculate offers undergraduate, postbaccalaureate,
and master’s degree programs for certiﬁcation, but all cater to individuals who work full-
time. Most courses are offered in the evenings or weekends. Many students in Holy
Immaculate’s teacher education programs work as long-term substitutes or uncertiﬁed
teachers in local schools while completing their coursework. The capstone of Holy
Immaculate’s teacher education program is a teacher as researcher project in which
students participate in a service learning project, reﬂect on their learning, and present

ﬁndings in an annual symposium.

70

CHAPTER 3: TEACHER CANDIDATES’ PERCEPTIONS OF THE

STRUCTURE OF THE TEACHING FIELD

Bourdieu’s cultural reproduction framework can help us understand teachers’
career decisions by offering insight into how the structure of society and the labor market
in particular inﬂuences individual teachers’ tastes and perceptions. The relational
dynamics—or the relationships between positions—that occur through interaction on a
speciﬁc ﬁeld are particularly important to developing a framework for understanding
teachers’ career decisions. As such, this chapter explores how teacher candidates perceive
the structure of the teaching ﬁeld, their place on the ﬁeld, and the relationship between
the teaching ﬁeld and ﬁeld of schools.

A study of teachers’ career decisions should be aware that teachers occupy
different positions on the teaching ﬁeld. Teachers may hold different positions depending
on their possession of the various types of capital valued in the ﬁeld (cultural, economic,
and social). A teacher’s position on the teaching ﬁeld inﬂuences the teacher’s habitus, or
set of dispositions and perceptions that guide his or her decision-making. For example,
accepting a job in an urban school could mean many different things to a teacher—it
could represent the teacher’s inability to secure a position in a more afﬂuent public
school, a chance to work with children from low-income backgrounds and help disrupt
patterns of social reproduction, or a chance to work with students from backgrounds
similar to their own. How the teacher perceives an urban teaching placement is
inﬂuenced by how the teacher perceives his or her position on the teaching ﬁeld as well

as the position of urban schools.

71

Teachers are individuals who make decisions about where to work based upon
their individual circumstances. Therefore, it does not make sense to assume that all
teacher candidates who choose an urban school ascribe the same meaning to this
decision. Indeed, many teachers may accept a job in an urban school without attributing
any larger meaning to the decision because they simply think the job is a good ﬁt for
them. Yet, patterns in teacher candidates’ decisions about where to work suggests that
teacher candidates’ social or cultural background inﬂuences these decision processes.
Therefore, a study of teachers’ career decisions should recognize the relative positions of
teachers within the labor market and how they perceive differences among schools.

Bourdieu argues that homologies exist between ﬁelds such that an individual’s
position on one ﬁeld is associated with his or her position on another ﬁeld. For a study of
the teacher labor market, the relevant ﬁelds are the teaching ﬁeld and the ﬁeld of schools.
Teachers are positioned on the ﬁeld according the amount and type of capital they
possess. Likewise, schools are arrayed on a ﬁeld according to the capital they have. The
types of capital valued on the teaching ﬁeld or on the school ﬁeld are empirical questions.
Bourdieu’s cultural reproduction framework suggests that teachers who occupy a
dominant position on the teaching ﬁeld will work in schools that occupy a dominant
position on the school ﬁeld.

This chapter examines how teacher candidates perceive the structure of the
teaching ﬁeld and the ﬁeld of schools; it explores the homologies, or structural
similarities, between the teaching and school ﬁelds seeking to understand how teacher
candidates’ sense of their position may inﬂuence their preferences and behavior. I ﬁrst

describe Bourdieu’s concept of ﬁeld and outline the types of capital that may be valued

72

on a ﬁeld. I then review the literature on teacher hiring to identify the characteristics that
existing research links to the structure of the teaching ﬁeld.

I then present the ﬁndings in three sections. The ﬁrst section explores how
prospective elementary teachers viewed the structure of the teaching ﬁeld and their place
within it. Teacher candidates recognized the independent inﬂuence of cultural, human,
and social capital in hiring decisions. While most teacher candidates’ perceptions agreed
with the literature on principal hiring decisions, a smaller group of teacher candidates
thought other factors structured the teaching ﬁeld. The second section explores how
teacher candidates viewed the ﬁeld of schools. Some teacher candidates noticed that
particular schools were selective in their hiring, placing those schools in a more dominant
position. Teacher candidates considered other schools, such as charter schools and
schools in other states, as not selective and therefore not in a dominant position. Smithton
Public Schools and schools in low-income suburbs were described as being in a weak
position on the ﬁeld as they faced ﬁnancial difﬁculties. The third section examines the
characteristics of teacher candidates who preferred to teach in schools in weak positions
on the ﬁeld in order to explore homologies between the teaching and school ﬁelds.

This chapter uses both the survey and interview data. The survey data are used to
identify factors teacher candidates think are important in the job search. The interview
data are used to explore what these factors mean to teacher candidates. Both the survey
data and interview data provide information on the relative importance of these factors,
and I search for consistencies or inconsistencies between the two types of data. The
interview data are also used to explore how and why teacher candidates describe some

schools as being in a dominant or weak position on the ﬁeld of schools. The survey data

73

are used to investigate structural similarities between the teaching ﬁeld and ﬁeld of
schools by examining which teacher candidates preferred to teach in schools in a weak
position.
Field and Capital

There are several types of capital that structure ﬁelds. Bourdieu’s theory focuses
mostly on cultural and economic capital (Bourdieu, 1986) while economic theory
includes human capital (Cohn & Geske, 1990). Cultural capital exists in knowing or
possessing the cultural norms, habits, and interaction styles that provide access to power
and high status positions and exists in embodied, institutionalized, and objectiﬁed forms
(Bourdieu, 1986). Embodied cultural capital consists of interaction patterns and social
norms recognized as appropriate by the dominant culture and resides inside the
individual. Embodied cultural capital cannot be directly transferred, although individuals
can obtain it through a socialization process. Institutionalized cultural capital represents
academic qualiﬁcations and credentials that may be used as tools to exclude individuals
or make objective distinctions between individuals. Due to its relationship to academic
degrees, institutionalized cultural capital is similar to the idea of human capital in the
economic literature, which consists of the necessary skills, expertise and knowledge for
particular occupations (Cohn & Geske, 1990; Schultz, 1977). Obiectiﬁed cultural capital
consists of physical objects, such as clothing, books, or cars.

In addition to these three types of cultural capital, social and economic capital
may also inﬂuence the structure of the teaching ﬁeld or the ﬁeld of schools. Teachers
with social capital have networks of friends, family, or other acquaintances from which

they can obtain resources that help them with their job search——either by giving them

74

access to particular jobs (Bourdieu, 1986) or providing information (Coleman, 1988).
Social capital on the ﬁeld of schools is also important as parents care about their
students’ peers, particularly the racial characteristics of their potential peers (Lankford &
Wyckoff, 2005).

Economic capital may be more important on the ﬁeld of schools than on the
teaching ﬁeld. As school funding is usually linked to property values, districts vary
greatly in their ability to raise revenue through property taxes (Berry Cullen & Loeb,
2004). In the Smithton area, districts receive most of their funding through per-pupil
foundation grants from the state. While this school ﬁnance program intended to create
more equity in school ﬁnance, afﬂuent districts are allowed to raise additional revenue
through local taxes. Further, the inability of some districts to adequately fund capital
expenditures results in wide disparities in the condition of physical buildings and
facilities (Arsen et al., 2005).

Previous Research on Teacher Hiring

Despite the importance of teaching quality and, therefore, teacher hiring, little
research exists on the criteria schools and districts use to hire teachers (Harris, Rutledge,
Ingle, & Thompson, 2006). Although some research exists on the timing and structure of
the hiring process and ﬂow of candidates through various steps in hiring, less research
exists on the criteria by which hiring personnel make decisions about which teachers pass
through the various steps (Liu & Johnson, 2003; Wise, Darling-Hammond, & Berry,
1987). This section reviews previous research on criteria in teacher hiring, focusing on
the personal characteristics and professional qualiﬁcations of teachers that are important

on the teaching ﬁeld.

75

Principals predominantly rely on personality characteristics when selecting
teachers and making decisions about whom to hire (Harris et al., 2006). In particular,
principals want teachers who are caring, enthusiastic, and motivated (Harris et al., 2006).
Principals also rate honesty, emotional stability, and confidentiality as among the most
important criteria used in hiring teachers, even though there are few objective ways to
assess these characteristics (Place & Kowalski, 1993). Districts rate teacher candidates’
human relations skills more important than their teaching skills (Ralph, Kesten, Lang, &
Smith, 1998). Indeed, even when districts focus on teaching skills, they consider the
ability to facilitate relationships (such as communication ability, rapport with students,
and creativity) more important than instructional planning and strategies (Ralph et al.,
1998). The interview is one of the most important tools principals use (Delli & Vera,
2003; Harris et al., 2006) and job candidates that interviewers consider more “likable”
receive more job offers (Delli & Vera, 2003). Principals may look for clues to a teacher
candidate’s personality, human relations skills, and work ethic in how they talk and the
types of questions they ask during an interview (Trimble, 2001).

There is mixed evidence on the importance of demographic factors on teachers’
success in the job market. Principals rated age as one of the least important characteristics
of teachers in hiring (Place & Kowalski, 1993), although some evidence suggests that
principals prefer the enthusiasm and adaptability of younger teachers (Harris et al., 2006).
Experimental evidence, however, found that principals prefer older teaching candidates.
For example, when principals examined resumes for candidates who were identical
except for their age, 49 year old candidates were more likely to be offered an interview

than were 29 year old candidates (Young & Fox, 2002). Young and Fox (2002) also

76

found that age and ethnicity interact in teacher hiring. While young Hispanic candidates
were less likely to be called back for an interview than all other candidates, older
Hispanic candidates were just as likely as candidates from other ethnic backgrounds to be
offered an interview.

Principals’ desire to hire teachers who ﬁt into the school may explain the
conﬂicting evidence for the importance of age in teacher hiring. Principals want to
maintain a balance on their staff or in a particular grade level or subject matter team such
that whether they prefer to hire a younger or older teacher depends on the composition of
the existing staff (Harris et al., 2006). Thus, principals are sensitive to the organizational
and social context of their schools and prefer to hire teachers who ﬁt in with the rest of
the staff. Likewise, principals’ hiring decisions suggest that they prefer to hire individuals
they consider to be socially proximal to them (Young & Delli, 2002). That is, principals
want to hire teachers who are similar to them.

Teacher qualiﬁcations appear less important in teacher hiring than personal
characteristics (Harris et al., 2006; Ralph et al., 1998). For example, teacher candidates
who graduated from the most selective colleges and universities are hired at lower rates
than teachers from other colleges (Ballou, 1996), although principals who themselves
went to selective colleges are more likely to hire teachers from selective colleges (Baker
& Cooper, 2005). Likewise, principals rate intelligence and coming from a speciﬁc
university as some of the least important teacher characteristics in hiring because they
assume that all college graduates meet the minimum threshold of intelligence necessary

for teaching (Harris et al., 2006). Principals do consider subject matter knowledge when

77

evaluating teacher candidates, although it is not as important as personality characteristics
(Harris et al., 2006).

The limited importance of general intelligence and speciﬁc teaching skills may be
due to the structure of the teacher hiring process. While hiring is decentralized, schools
and districts still rely mainly on paper credentials and usually do not observe the
individual teaching or review a portfolio of the candidate’s teaching experiences (Liu &
Johnson, 2003). As such, districts usually only have opportunities to assess a candidate’s
teaching ability only after he or she has been hired and started working (Wise et al.,
1987). Still, the interview is an important tool in teacher hiring and a study connecting
principal ratings of teachers during interviews with teacher performance data found that
most—but not all—principals can accurately distinguish between strong and weak
teachers during an interview (Emley & Ebmeier, 1997).

The economic literature refers to qualiﬁcations and ability as human capital. In
cultural reproduction terms, qualiﬁcations are forms of institutionalized cultural capital
and personal characteristics are forms of embodied cultural capital. That is, personality
traits are characteristics that cannot be learned or transferred to another person while
qualiﬁcations and human capital are obtained through academic organizations (Bourdieu,
1986). Thus, the literature on teacher hiring suggests that principals and districts
emphasize personal characteristics and personality—embodied cultural capital—when
hiring teachers. Qualiﬁcations and teaching skills—institutionalized cultural capital or
human capital—are of secondary importance. The teacher hiring literature provides little
evidence on the importance of the third major form of capital—-—social capital. Principals

look for caring, honest, and enthusiastic teachers and they assume that certiﬁed

78

candidates have the requisite instructional skills and intelligence. Principals also want
teachers who are similar to them and who complement the existing teaching staff.

In this way, the extant literature suggests that the hiring process may work to
reproduce social distinctions among teachers and schools. As principals appear to rely
predominantly on personality traits and perceptions of ﬁt with the current staff, their
decisions reﬂect a sense of the ﬁeld and of which teacher candidate is a ﬁt for the job in
their particular school. Thus, principals’ hiring decisions are driven by their own sense of
the teaching ﬁeld and its relation to the ﬁeld of schools. This chapter examines whether
teacher candidates make similar distinctions between professional and personal
characteristics in the teaching ﬁeld and what characteristics they think determine which
candidates are hired.

Teacher Candidates ’ Perceptions of the Teaching Field

This section examines the extent to which teacher candidates distinguish between
different types of characteristics as structuring the teacher labor market. The teacher
hiring literature suggests that principals make distinctions between personal and
professional characteristics when hiring teachers (Harris et al., 2006) and the general
employment literature suggests that social networks are important in securing a job
(Granovetter, 2005; Hanson & Pratt, 1992). Little empirical evidence exists on whether
teacher candidates also recognize these three types of characteristics as structuring
opportunities for teaching jobs. The survey component of this study asked teacher
candidates to rate a variety of teacher characteristics on the extent to which they make a

candidate more likely to get a job. The characteristics were chosen to represent the

79

theoretical constructs of interest in this study, including social capital, human capital or
institutionalized cultural capital and embodied cultural capital.

To analyze teacher candidates’ perceptions of the structure of the teaching ﬁeld, I
perforrrrcd an exploratory factor analysis on these 23 items to look for patterns in how
teacher candidates think about the characteristics that helps one get a job. Figure 1 shows
the scree plot of the obtained factors. As shown in this ﬁgure, three factors had an
eigenvalue greater than the Kaiser criterion of l and the slope of the scree plot starts to
ﬂatten out after the third factor. This suggests that three factors should be retained and

interpreted. Together, the three factors explain 82.3 percent of the variance in the data.

 

 

4
3.5 J

3 u
2.5 .

Eigenvalue 2 ~

 

 

 

 

1.5 «
1 -
0-5 3 \ﬁ—
0 i 1 T . T T i T 1 1
1 2 3 4 5 6 7 8 9 10 11
Factor

 

 

 

Figure l - Scree Plot of Factors in Teaching Field

I obtained factor loadings using a varimax orthogonal rotation method. I
considered items to load on a factor if their factor loading was higher than .3. Table 10
shows the items loading on each factor and their factor loadings. As shown in this table,

the three retained factors reﬂect the different types of capital—social capital, embodied

80

cultural capital, and institutionalized cultural capital or human capital—used to construct
the items. Teacher candidates did not distinguish between institutionalized cultural
capital and human capital, which was not surprising as these two types of capital are
often operationalized using similar measures. For example, having a high GPA may
represent human capital (i.e., having the knowledge and skills that teachers and college
instructors use to determine grades) as well as cultural capital (i.e., knowing the social
norms for student behavior that results in recognition by the teacher and, thus, good
grades).

Four characteristics loaded onto the ﬁrst factor, representing social capital. Of
these characteristics, three loaded very strongly on this factor. These included direct
measures of whether an individual teacher candidate may have family or friends who are
teachers, principals, or school district ofﬁcials. Teacher candidates who have other
education professionals in their social network may have greater access to information
about available jobs or how to get a job. A ﬁnal characteristic loaded weakly on this
factor: being in the right place at the right time. While this may be an indirect measure of
social capital, it may still represent the idea that meeting the right people—in this case
meeting them at an opportune time—makes it more likely that a teacher candidate will
get a job offer. The Cronbach’s alpha for this factor was .76, indicating that this measure
had good internal reliability.

Nine variables loaded onto the second factor, which measured human capital and
institutionalized cultural capital. The variables that loaded on this factor all included
some measure of work experience, academic credentials, or teacher certiﬁcation. This

measure had a reliability of .70.

81

Table 10 - Factor Analysis Results and Item Loadings From Teacher Candidates’

Perceptions of Impact of Various Characteristics on the Likelihood of Getting a Job Offer

 

Item
number Item text Loading

 

 

Factor l—Social capital

 

 

 

 

Q19K Have a relative or friend who is a teacher 0.64
Have a relative or friend who works in a school

ngM district ofﬁce 0'85

Q19N Have a relative or friend who is a school principal 0.80

Q19W Are in the right place at the right time 0.35

 

 

Factor 2—Human capital

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q 1 9A Have experience tutoring or mentoring children 0.53
Q19B Have several years of teaching experience 0.32
Have ex erience workin with children in non-

Q19C school Sgttings g 0'52
Q19D Have a high GPA 0.46
Q19F Have a full year of student teaching experience 0.46
QI9H Have high scores on teacher certiﬁcation exams 0.51

Have ex erience teachin in urban schools
Q19] anludirfg student teachiﬁg) 0'37
Q19D Are certiﬁed to teach multiple subjects or grades 0.47
Q19T Come from a highly ranked university 0.46
Factor 3—Embodied cultural capital

Come from the same racial/ethnic back round as
Q19G other teachers in the school g 0'59
Q19I Fit in well socially with the school staff 0.32
Ql9P Look like other teachers in the school 0.54
Q19U Are male 0.39
Q19V Are White 0.65

 

 

 

Note: Data come from teacher candidate survey. N=289

 

Five teacher characteristics loaded on the third factor. This factor included
measures of teacher candidates’ race, gender, physical appearance and ability to ﬁt in
socially with the school staff. The Cronbach’s alpha measure of internal reliability for
this measure was .64, suggesting that the items in this third factor were less closely

related to one another than the ﬁrst two factors.

82

Overall, teacher candidates identiﬁed three distinct forms of capital as structuring
the teaching ﬁeld and inﬂuencing who gets a job offer. Table 11 shows the means and
standard deviations for the three factors. The variables ranged on a scale from 1 (much
less likely to get a job) to 5 (much more likely to get a job). The 95% conﬁdence
intervals for social and human capital overlapped, indicating that teachers thought these
two factors are of about equal importance in getting a teaching job. Teacher candidates
rated embodied cultural capital rated slightly lower than both social capital and human

capital.

Table 11 - Mean and Standard Deviation for Each Measured Factor

 

 

Standard
Factor Mean deviation
Social capital 4.02 0.63
Human capital 3.96 0.38
Cultural capital 3 .45 0.49

 

N=278

The interview data also provided evidence on the relative importance teacher
candidates gave to various characteristics in the labor market. During the interviews,
teacher candidates were asked open-ended questions about what they thought schools
looked for when hiring teachers. Table 12 reports the most frequently mentioned
characteristics. Teacher candidates named certiﬁcation or basic qualiﬁcations for the job
most frequently and over half of the teacher candidates said having the qualiﬁcations to
teach in particular subject areas—such as math, science, or special education—help one
get a job. Yet most teacher candidates also thought that individuals with particular

demographic characteristics—such as being a member of a racial minority group or

83

male—are more likely to get a job offer. Teacher candidates also thought that schools

look for individuals who ﬁt in or have particular personality characteristics.

Table 12 - Percentage of Teacher Candidates Who Named Various Characteristics as

What Schools Look for When Hiring Teachers

 

 

Characteristic Percent
Certiﬁcation, or a generally qualiﬁed, competent teacher 66.7
Come from a racial minority background 63.0
Has a credential in a particular shortage area 55.6
Is a good ﬁt with the school or principal 55.6
Personality characteristics 51.9
In the principal’s social network 51.9
Are male 48.1
Have experience teaching, are not ﬁrst-year teachers 40.7
Commitment, passion for children 33.3
Have worked as a substitute in that district 29.6
I don't know what schools look for 18.5
Come from a speciﬁc university 18.5
Have a full year of student teaching experience 18.5
Does not matter if White or from a racial minority group 18.5
Are a team player 18.5
Have a certain look 14.8
Demonstrate an interest in that school, court the school 14.8
Experience working with diverse students 14.8
Experience working with children in non-classroom setting 14.8
Positive letters of reference 14.8
Candidates with a master’s degree are not hired 11.1
Are female 11.1
Are a good communicator 11.1
Individuals who are similar to the principal 11.1

 

Note: Data come from teacher candidate interviews. Total number of teacher
candidates is 27.
Overall, the interview data suggested a different relative importance of the types
of capital that structure the teaching ﬁeld. The survey data indicated that teacher
candidates thought social and human capital are equally important, with embodied

cultural capital playing less of a role. The interview data indicated that embodied cultural

84

capital (which combines personality characteristics, communication skills, and
demographics) was most important, followed by human capital and then social capital.
While embodied cultural capital and human capital were named by equal numbers of
teacher candidates, as discussed below, the ways in which they talked about human
capital suggested it was a threshold factor. The next three sections examine each of these
types of capital in more detail, including what they meant to teacher candidates, why
teacher candidates thought they are important, and whether teacher candidates varied in
the importance they placed on each type of capital.
Embodied Cultural Capital
Personality and ﬁt
Similar to existing research on principals’ decisions, teacher candidates did not

necessarily think that schools hire based only on professional characteristics or
qualiﬁcations. Although two-thirds of teacher candidates interviewed mentioned being
competent or qualiﬁed as one factor that schools look for in new hires, they also
mentioned personality characteristics and being a good ﬁt for the school. In many cases
teacher candidates mentioned certiﬁcation as a prerequisite before describing other
features that play a larger role. For example, Joshua said,

They’re looking for people who are qualiﬁed and have their

certiﬁcation and looking for teachers who are passionate

about what they do. They’re looking for teachers who are

enthusiastic, looking for chemistry, teachers who will be a
good ﬁt with the rest of the teachers.

Joshua ﬁrst mentioned certiﬁcation and qualiﬁcations as necessary characteristics for a
teacher to be hired, but quickly turned to personality traits, such as passion and
enthusiasm, and ﬁtting in with a school staff. Based on the interview data, teacher

candidates thought that schools hire individuals who are conﬁdent, dependable, ﬂexible,

85

and passionate. Although these characteristics may be highly individualized traits, being
identiﬁed as conﬁdent or dependable requires individuals to exhibit the behaviors and
mannerisms that principals can recognize as signs of conﬁdence or dependability. That is,
seemingly individual behavior characteristics are identiﬁed as positive personality traits
through culturally constructed deﬁnitions. Further, while enthusiasm and ﬂexibility may
be traits necessary for effective teaching, they are best considered forms of embodied
cultural capital rather than human capital as they are acquired through a process of
socialization rather than training. Thus, it seems that teacher candidates deﬁned elements
of embodied cultural capital as important characteristics that schools look for when they
hire teachers. As these traits are those that principals also prefer, it appears that many
teacher candidates did have a “feel for game” (Bourdieu, 1984).

In addition to personality characteristics, teacher candidates thought schools hire
teachers who ﬁt in with the rest of the staff. Over half the teacher candidates said schools
look for a “good ﬁt” when hiring teachers; a good ﬁt was the third most frequently cited
factor in teacher hiring. Indeed, some teacher candidates placed ﬁtting in as more
important than professional qualiﬁcations. For example, Daniel said,

First and foremost I think you gotta have a ﬁt... it’s all
about the ﬁt. And then I think they look at, you know, your

abilities. And then basically it’d be, you know, what
type of person that you are.

Daniel described ﬁt as “just kind of how you connect with one another” and said it is
assessed during the interview. Other teacher candidates characterized ﬁt as getting along,
feeling at ease, feeling you were having a conversation rather than an interview, and
making eye contact. Fit then appears to be a form of embodied cultural capital, as it

cannot be separated from an individual and represents interaction skills and habits.

86

Denise implied that a “good ﬁt” might, at least in part, be based on social class.
Denise felt out of place in a school in a wealthy neighborhood where she was initially
placed for student teaching. She ended up not student teaching there because the assigned
cooperating teacher did not want to work with her. She described how she felt he was
brushing her off. She said,

I’m not that upper class so I thought maybe it was a class
issue. Could he tell that about me? I don’t know.
...Sometimes I get self conscious like I know I’m not a
petite [wealthy town] woman, you know. ...I have no idea.
Seeing the women in the building, I wasn’t dressed like

them. I didn’t have the expensive Saks 5th Avenue clothes.
I had Kohl’s clothes.

Denise felt uncomfortable because she did not look like other teachers in the school; she
did not think she had the appropriate embodied and objectiﬁed cultural capital because
she had a different body type and wore different clothes. As a result, she thought she was
reassigned to a new student teaching placement because she did not ﬁt in with the other
teachers in the school.

Most teacher candidates who talked about ﬁtting in with a school talked about
working well with others. For example, Melissa deﬁned a good ﬁt as “being able to
interact well with the other staff.” Fitting in, then, is about working cooperatively with
other teachers and forming relationships. That Melissa saw ﬁt as a type of capital that
helps her on the job market was evident when she expressed concern about competing
with teachers who have a couple years of experience. Melissa said,

Maybe there are other candidates that have taught for three
years or something and they, you know, kind of already

know how the show runs. But I think on the positive side,
like I, I do think I ﬁt very well with, I can, I can work with
pretty much all people. . .I think that part could give me, or

would be like the part that would give me a good edge in
that aspect.

87

Despite being a brand new teacher, Melissa thought she could be hired over more
experienced candidates because she works well with other people. Interestingly, Melissa
deﬁned ﬁt as something that is applicable to all schools—she could ﬁt in to all schools—
rather than being dependent on matching some characteristic of the school. For Melissa
and other teacher candidates who deﬁned a good ﬁt as being able to get along with other
teachers, ﬁtting in is a unidimensional construct. Some teachers can ﬁt in with all schools
because they are able to work well with others, while other teachers do not ﬁt in with any
school because they cannot get along with their colleagues. The ability to ﬁt in then
becomes a form of cultural capital and is closely linked to personality traits, such as
ﬂexibility and the ability to work in a team.

Other teacher candidates deﬁned ﬁt as a multidimensional construct. When
teacher candidates deﬁned ﬁtting in as having things in common with other teachers in a
school or sharing the school’s instructional philosophy, then they recognized that one
teacher could ﬁt in with one school but not with another. For example, Stephanie said,

I think they’re also looking for somebody who’s a good ﬁt
for their school. Like for example, if the person is very,
very well dressed, very, like a high maintenance type
person that person might go well in one school where
everyone’s like that so then they’re gonna get along with
the teachers, they’re gonna get along with the principals.
But they might not get along with the high maintenance
person might not ﬁt in well there where it’s very relaxed.
So I think the principals look for people who are gonna ﬁt

in with the personalities of the school, as well as the
philosophies of the school.

Stephanie talked about ﬁtting in with a school by matching the personality characteristics
and teaching philosophies of other teachers in a school, as well as having a certain look.
She described a “good ﬁt” as complementing the existing personalities and dispositions

in the school. A teacher may ﬁt into one school but not another. Stephanie would only ﬁt

88

into certain schools—those with a progressive pedagogical approach. This contrasts with
Melissa, who described a good ﬁt as more universal.

Teacher candidates did vary in the emphasis they placed on personality and ﬁt in
the teacher labor market. In particular, teacher candidates differed over the relative
importance of personality traits or professional characteristics in getting a teaching job.
For example, younger teacher candidates were more likely to focus on personal traits. All
fourteen of the teacher candidates under the age of 25 talked about the importance of
personal traits in getting a job, but only three of the seven teacher candidates over age 35
talked about these types of characteristics. Likewise, new college graduates—who tend to
be younger—emphasized personality characteristics more than mid-career changers. All
15 new college graduates said that principals and schools care about personal traits, while
only 7 of the 12 mid-career changers thought principals considered personality
characteristics. As enthusiasm, ﬂexibility, and motivation are traits that evoke youthful
images, it is not surprising that younger teacher candidates were more likely to emphasize
these personality characteristics.

There was also a relationship between the socioeconomic status of teacher
candidates and whether they talked about personal characteristics and ﬁtting in with a
school. For example, 75% of teacher candidates who came from upper middle class
backgrounds emphasized a good ﬁt when describing what schools look for when they
hired teachers, compared to 33% of those from working class backgrounds. Overall,
White teacher candidates (16 of 18) were more likely than African American teacher
candidates (0 of 5) to focus on personal characteristics when describing what schools

look for when hiring teachers. They emphasized traits like good communication skills,

89

enthusiasm, and motivation. As principals themselves did consider personality and ﬁt
with the existing staff most important in choosing teachers, it seems that young, White
teacher candidates and those from middle or upper middle class backgrounds had the best
feel for the game. That is, they were most likely to emphasize the traits on which
principals do in fact base their hiring decisions.

That teacher candidates varied in the relative importance they placed on embodied
cultural capital is important for two reasons. First, the characteristics that teacher
candidates thought were most important in the job search are likely to be those they
emphasized on applications and in interviews. That is, teacher candidates may present
themselves in different ways to principals or other hiring personnel depending on what
they thought is most important to the hiring ofﬁcial. Second, the distinctions teacher
candidates saw as structuring the teaching ﬁeld may be related to how they thought about
their decisions on the job search. Teacher candidates may have approached the job search
quite differently if they thought professional qualiﬁcations were more important on the
teaching ﬁeld than perceptions of ﬁt.

Demographic Characteristics

To examine how teacher candidates think demographic characteristics inﬂuenced
the hiring process, I analyzed the survey data on teacher candidates’ ratings of the effect
of being White or coming from a racial minority background on one’s ability to get a
teaching job. The variables ranged on a scale from 1 (much less likely to get a job) to 5
(much more likely to get a job). Each variable was standardized (mean = 0 and standard
deviation = 1) and teacher candidates’ rating of the impact of teacher race (either being

White or coming from a racial minority background) was regressed on various teacher

90

candidate characteristics. Table 13 presents the means and standard deviations for these
variables and results from the individual regression analyses. Comparing the means
shows that, overall, teacher candidates thought that coming from a racial minority
background made a teacher more likely to get a job offer (t-statistic = 3.01, p—value =

0.003).

Table 13 - Regression Analyses of the Inﬂuence of Being White and Coming From a

Racial Minority Background in Teacher Hiring

 

Are ﬁom a racial minority

 

Are White background
Variable Mean SD Mean SD
Mean 3.14 0.74 3.34 0.76
Respondent characteristics Coeﬂicient SE Coeﬂicient SE
Intercept -0.09 0.08 0.18* 0.08
Master’s degree 0.32 0.27 0.15 0.27
Mid-career changer 0.05 0.14 -0.53*** 0.14
Male -0.23 0.18 0.16 0.18
Aﬁican American 0.62M 0.23 -0.19 0.23
Other racial minority 0.10 0.21 0.07 0.21
R-squared 0.060 0.065
N 268 271

 

Note: The dependent variables for these regression equations are teacher candidates’
perception of the impact of being White and of coming from a racial minority
background on their likelihood of getting a job offer. The results in this table are
different if Oak State respondents are excluded from the sample. See Table A1 in
Appendix D.

* p<.05, ** p<.01, *** p<.001.

The only statistically signiﬁcant effect on teacher candidates’ perception of the
importance of being White on the teaching ﬁeld is whether the teacher candidate was
African American. Aﬁican American teacher candidates were more likely than White
teacher candidates to say that being White made someone more likely to get a job offer.
Aﬁican American teacher candidates seemed to have recognized that they face some

91

difﬁculty in getting a teaching job because they were not White. This may reﬂect the
historical racial discrimination that Aﬁican Americans have faced. The low R—squared,
however, suggests this model does not explain much of the variance in teacher
candidates’ perception of the importance of being White in getting a teaching job.

Due to a low response rate from Oak State University, I also ran the analysis
excluding Oak State respondents to test for nonresponse bias. Including Oak State
respondents changes the regression results for teacher candidates’ rating of the impact of
being White on the likelihood of getting a job. Aﬁican American teacher candidates still
gave higher ratings than White teacher candidates (coefﬁcient-:50, SE=.25), but there
were differences for the coefﬁcient on teacher candidates with a master’s degree and
male teacher candidates. Those with a master’s degree gave higher ratings to the impact
of being White on the likelihood of getting a job while male teacher candidates gave
lower ratings. Full results can be found in Table A1 in Appendix D.

Although there were few differences in teacher candidates’ perceptions of the
value of being White on the teaching ﬁeld, there were differences in their perceptions of
the effect of coming from a racial minority background on getting a teaching job.
Although African American teacher candidates thought that being White gave one an
advantage, they rated the importance of coming from a racial minority background the
same as other teacher candidates. Thus, it seems that African American teacher
candidates thought that White teacher candidates and teacher candidates from racial
minority backgrounds had about equal chances of getting a teaching job while other

teacher candidates thought that racial minorities were more likely to get a job.

92

This ﬁnding is supported by the qualitative data. Even though equal numbers of
African American teacher candidates in the qualitative component said that race did or
did not matter, three times as many White teacher candidates said that minority
candidates had an advantage than said that race did not matter. Most White teacher
candidates said schools wanted to hire minority candidates. For example, Heather said,

I deﬁnitely think I’m at a disadvantage because I’m a
White female. I had one principal that actually told me that
he basically had to hire minorities. . .they have to be in

compliance with some sort of minority quota . . . if they’re
not then they have to hire minority teachers.

Many teacher candidates, like Heather, implied that racial minority candidates beneﬁted
from a form of afﬁrmative action in teacher hiring because most existing teachers are
White women. Some teacher candidates also thought schools wanted to hire teachers of
color because they recognized the diversity in the student population and the need to have
teachers reﬂect that diversity. While these teacher candidates still felt they faced a
disadvantage in obtaining a teaching position, they recognized the reasons behind it.
While some African American teacher candidates did recognize that they might

have an advantage due to schools’ desire to diversify their teaching staff, at least one
teacher candidate expressed reservation about teaching in a district that would hire her
because of her race. Zoe is avoiding one district because she heard they are facing a
lawsuit due to the lack of diversity in the teaching force. She said,

I think with the achievement gap and things, I think they

are deﬁnitely interested in minorities. But are you

interested in my qualiﬁcations or just the color of my skin?

So, I mean, like I didn’t apply for districts like [district

omitted] because like I said that’s what I think they’re
interested in.

93

Many teacher candidates described this district as being one of their top choices and were
working hard to get a job in this district. Indeed, several White teacher candidates
expressed frustration with their diﬂiculties in getting a job due to racial preferences. Yet
Zoe was avoiding this district because she thinks they value her skin color over her
qualiﬁcations.

Many teacher candidates thought that males were also more likely to get a job
offer, especially since they were looking for elementary teaching jobs. About half of the
teacher candidates thought males held a privileged position in the teaching ﬁeld. For
example, Valerie said,

I think men normally get hired quicker they’ll hire them
because there aren’t that many male candidates, so I think
they would pick a man teacher over a women, most likely,

even though, you know, most people would say ‘we’re
going to pick the best candidate.’

Valerie, and other teacher candidates like her, recognized that there were few male
elementary teachers. Both male and female teacher candidates noted that men held an
advantage. It is of interest that three teacher candidates—in particular those who were
older and came from lower socioeconomic backgrounds—talked about females having an
advantage on the teaching ﬁeld. For example, Wayne said, “Education is aimed towards
females. It’s always had a female basis for the most part because most teachers are
females.” These teacher candidates thought that since most elementary teachers were
female, female teacher candidates ﬁt into elementary schools. While race and gender are

considered secondary characteristics in cultural reproduction theory and are not

94

foregrounded in Bourdieu’s work (Bourdieu, 1984),’1 they take on a more primary
importance for the teacher candidates in this study.

In summary, most teacher candidates’ perceptions of the structure of the teaching
ﬁeld corresponded to the characteristics principals consider most important. Most teacher
candidates thought that personality was important in the teacher labor market and many
teacher candidates thought personal characteristics were more important than
qualiﬁcations, assuming that all candidates were certiﬁed. White teacher candidates, new
college graduates, and teacher candidates from middle or upper middle class backgrounds
were most likely to say that personal characteristics were important in teacher hiring,
suggesting these teacher candidates had the best sense of the game in terms of what
principals consider in hiring.

Most teacher candidates also thought that race mattered on the teaching ﬁeld. In
particular, White teacher candidates thought that racial minority candidates had an
advantage, while African American teacher candidates tended to think that race did not
matter or that they did not want to work in a school that hired them because of their race.
Finally, teacher candidates thought that schools looked for teachers who would ﬁt in with
the existing staff. They described a “good ﬁt” as matching the personalities, philosophies,
dispositions, and appearances with other teachers in the school. Thus, teacher candidates
appeared to understand conclusions from the teacher hiring literature—that schools hire
teachers who cohere with the school staff. Yet the teacher candidates in this study used a
different deﬁnition of “ﬁt” than is evident in the teacher hiring literature. While

principals seem to want teachers who supplement others on the staff to create a balanced

 

” The relative lack of attention to race in Bourdieu’s work is most likely due to the European focus in
much of his writing.

95

teaching force, teacher candidates thought that schools hired teachers who complemented
them.
Human Capital/Institutionalized Cultural Capital
The teacher hiring literature suggested that professional qualiﬁcations were of

secondary importance to principals. While the teacher candidates in this study thought
personality, demographics, and a good ﬁt were important, qualiﬁcations were mentioned
more often in the interview data. Still, as noted above, teacher candidates described the
importance of professional qualiﬁcations as a threshold factor that did not distinguish
candidates above a minimum level. The survey data provided stronger evidence that
teacher candidates thought human capital was more important than embodied cultural
capital (see Table 11). The qualitative data provided evidence on what teacher candidates
meant by professional qualiﬁcations. Professional qualiﬁcations were a form of human
capital or institutionalized cultural capital as they were represented by certiﬁcations,
subject matter knowledge, and teaching skills. Many teacher candidates did recognize the
importance of qualiﬁcations and competence for teaching. For example, Vanessa said,

I think schools look for certiﬁcation, experience, and what

you can offer their students. And they look at your

student teaching evaluation. Did you fulﬁll everything to

the Xth mark? Did you go above and beyond, not just
mediocre? So they look for someone that can add to the
teaching ﬁeld, but also adapt. They looked for those that
can teach but are aware of the laws that are involved.
They’ll ask you, do you know about the IDEA law? . . . And
they really want you to be current on everything that is
occurring in the teaching ﬁeld and the pedagogy.

Vanessa referred to qualiﬁcations speciﬁc to teaching, including certiﬁcation,
pedagogical knowledge, and awareness of speciﬁc education laws. Teacher candidates

talked about being qualiﬁed in general, as well as having speciﬁc qualiﬁcations that were

96

particularly in demand. Valerie talked about the need for special education teachers when
she said, “I know now that it probably helps to be specialized in an area. Special
education is big right now, a lot of people are looking for that.” In addition to special
education, teacher candidates noted that schools were particularly interested in candidates
with a math or science background, English as a Second Language, and technological
proﬁciency. Thus, teacher candidates’ perceptions of shortage ﬁelds in teaching appeared
to match conventional accounts of those vacancies schools have the most difﬁculty
ﬁlling.

Teacher candidates varied somewhat in the importance they placed on human
capital and professional qualiﬁcations. Older teacher candidates talked more about
credentials and qualiﬁcations—in general and in speciﬁc teaching areas—while younger
teacher candidates emphasized other characteristics. While nearly all teacher candidates
said qualiﬁcations were a factor in hiring decisions, teacher candidates over age 35 talked
about qualiﬁcations twice as much, with an average of 3.4 statements about the
importance of qualiﬁcations, compared to an average of 1.7 statements by candidates
under the age of 25. Likewise, Aﬁican American teacher candidates focused more on
professional qualiﬁcations than did White and Asian teacher candidates (3.2 and 2.1
statements per teacher, respectively. As the teacher hiring literature suggests that
principals place primary emphasis on personality and a good ﬁt, it seems that older
teacher candidates and Aﬁican American teacher candidates had less of a feel for the
teacher labor market.

Like the relative importance of embodied cultural capital, the degree to which

teacher candidates placed different emphases on human capital in the teacher labor

97

market could have implications for how they presented themselves to hiring ofﬁcials they
approached the job search. If teacher candidates thought the ﬁeld was structured by
professional qualiﬁcations, then they were likely to continue making distinctions in their
job search based upon professional criteria, rather than a sense of ﬁt.
Social Capital
In addition to personality and ﬁt, demographic characteristics, and professional

qualiﬁcations, teacher candidates recognized the importance of social capital in getting a
teaching job. About half of the teacher candidates indicated that having social ties with
other teachers, principals, and district personnel gave someone an edge in the hiring
process. Stephanie described the importance of social networks when she said,

Having a connection gets your resume pulled. . . . It gets a

good recommendation. Like if my connection in [one

district], he’s kinda higher up so if he tells the principal,

you know, I think she’s really good he might not give

me the job but he’s going to at least interview me and look
at me in a different light than he would some stranger.

Stephanie thought her chances of getting a job offer were higher than other teacher
candidates because she had a contact with a district official that could get principals to
look at her resume. Teacher candidates also talked about the importance of making a
personal connection and getting one’s face known throughout a district. They thought
principals were more likely to call them in for an interview if they could put a face to the
name on the resume. While social networks may not be enough to guarantee a job, it
would get teacher candidates through to an interview, where they would compete in a
much smaller pool.

Some teacher candidates actively sought to expand their social network to

improve their chances of getting a teaching job. Their strategies included making an

98

effort to meet principals in person when dropping off resumes so that principals would
recognize their face later; getting recommendations from colleagues in their student
teaching school who would drop names when meeting new principals; and substitute
teaching. By substituting in various schools, teacher candidates thought they could meet
new teachers and principals and let them know they were applying for teaching positions.
Richard was perhaps the most active in accumulating social capital. He contacted
multiple principals and district ofﬁcials and asked them to do a mock interview with him.
He said,

I asked her to do a mock interview and then I asked her if

she’d be willing to set up other mock interviews for me in

other districts. Because I think giving a face with

administrators who haven’t seen you before is signiﬁcant.

they may say, you know what, if I have any opportunity,

or if somebody calls me and says I need a teacher do you

have any candidates in mind, I just had an interview with
this person that, you know seems great.

Richard set up these mock interviews not to improve his interviewing skills but to meet
principals and expose himself to multiple schools. By expanding his educational network
in this way, Richard hoped to improve his chances of getting a teaching job.

Finally, teacher candidates whose parents’ held occupations with similar salaries
and social status as teachers placed the most emphasis on social networks and substitute
teaching in schools as giving someone an advantage on the teaching ﬁeld. Compared to
teacher candidates whose parents held occupations with both more and less social status
and income than teaching, teacher candidates whose parents had occupations similar to
teaching were both more likely to mention social networks as a factor in teaching hiring
(8 out of 11 compared to 9 out of 16) and made more statements about its importance (1.7

statements compared to .6 statements). Many of these teacher candidates had parents who

99

were in fact teachers themselves, so they may have had more extensive networks within
schools and districts. White teacher candidates also placed more emphasis on social
networks (average of 1.3 statements), and Aﬁ'ican American teacher candidates (.6
statements) placed less emphasis on social networks.

The principal hiring literature does not provide much evidence on the relative
importance of social capital in hiring decisions. If teacher candidates’ perceptions were
accurate and social capital had an independent effect on who was hired, then this has
implications for prospective teachers’ equitable access to teaching jobs for teacher
candidates who do not have families with educators. If research on principals’ hiring
practices conﬁrms the importance of social networks, then teacher education programs
should explore possible ways to develop teacher candidates’ social networks.

The qualitative data suggested that teacher candidates were quite conﬁdent in
their ability to get a job. Although most teacher candidates were aware of the slack in the
teacher labor market and that many districts were laying off teachers, they thought their
own chances of getting a job were rather good. Only two teacher candidates were really
concerned they might not get a teaching job. One reason teacher candidates seemed
secure in the idea that they would ﬁnd a job is because they described themselves as
having the characteristics they thought schools wanted to hire. That is, teacher candidates
deﬁned the criteria they thought schools looked for in ways that made them highly
desirable candidates. As a result, teacher candidates recognized the poor job prospects for
elementary teachers in general but were conﬁdent of their own individual prospects

because they thought they matched what schools wanted.

100

While nearly all teacher candidates deﬁned the ﬁeld in ways that put them in the
most dominant position, not all teacher candidates were accurate in their perceptions of
the ﬁeld, given what we know about the characteristics important to principals. Young
White teacher candidates from middle class backgrounds had the best sense of the game.
That is, these teacher candidates’ perceptions of the structure of the ﬁeld most closely
matched the ﬁndings from the literature on teacher hiring.

Teacher Candidates ’ Perceptions of the Field of Schools

A cultural reproduction ﬁamework highlights the presence of homologies
between positions on different ﬁelds. This suggests that teacher candidates who occupy a
dominant position on the teaching ﬁeld will teach in schools that occupy a dominant
position on the ﬁeld of schools. Likewise, teacher candidates in a weak position on the
teaching ﬁeld will teach in schools that occupy a weak position. While the structure of
the ﬁeld of schools is an empirical question, existing research on the characteristics of
schools that face persistent staffing difﬁculties provides some evidence on this question.
Schools with high concentrations of poor, minority, and low-achieving students are more
likely to have teachers who have failed the teacher certiﬁcation exam or lack full
certiﬁcation (Lankford et al., 2002).

Schools can vary dramatically in their ability to be selective in teacher hiring. A
study of the conditions of schools with chronic teacher turnover (Guin, 2004) found that
even within one district, schools in poor neighborhoods received about 3-5 applications
for every open position while schools in afﬂuent areas received about 150 applications.
My data suggested that teacher candidates did recognize differences between schools’

selectivity in hiring. Indeed, some teacher candidates in this study wanted to work in

101

schools they thought were selective in hiring and did not want to work in schools they
perceived as unselective. That is, for those teacher candidates who noticed the relative
selectivity of schools, selectivity was a desirable characteristic and made them want to
work in that school. Some teacher candidates did want to work in the schools considered
unselective by their peers, but the selectivity or lack of selectivity was not a salient
feature for them. This suggests that some teacher candidates recognized the structure of
the ﬁeld of schools and wanted to work in schools in a dominant position on the ﬁeld.
While teacher candidates were aware of the tight labor market in the area in

general, they did note that it was harder to get jobs in particular schools. Teacher
candidates linked the difﬁculty of getting a job to the quality of the district. For example,
Daija said she was very happy to have secured a job in a suburban district. She called the
district “the best district in the state” and explained,

There are a billion screenings for teachers. The district

is well funded. It is made up of areas that are very rich, that

are middle class, that are poor... not poor, but lower. They

have a lot of money for supplies. I have $250 to spend on

classroom supplies. . .. They have all kinds of programs for
their students.

Daija made a connection between better districts and many screenings of teachers. She
called the district “hard to get into” and that her acquaintances were impressed that she
was hired. Daija also linked the district’s status to the socioeconomic status of the
population and their ability to buy materials and support programs for students. Other
teacher candidates made a similar connection between a school’s status and funding and
their ability to be selective in hiring teachers. Chris described his reaction to a job
opening at a secular private school in an afﬂuent suburban area that said they preferred

someone with experience.

102

Chris: They’ll be picking who they want. They’ll pick and
choose. They’ve got some prestige, they’ve got the
money and they’ll pick who they want. . .Experience
preferred. Catholic schools put that and that’s, they
would prefer it but they’re saying we’ll also consider
others. Some people put that, they mean just that.

Interviewer: And how do you know the difference?

Chris: Cause they’ve got the money, they’ve got the
prestige. And money talks. It’s just the way it is.

Chris noted a distinction between this secular private school and Catholic schools in their
ability to be selective. He thought both types of schools may prefer to hire similar
teachers, but the money and prestige associated with the secular private school allowed
them to “pick and choose” among candidates. The implication is that other schools, such
as Catholic schools and schools with less money and prestige, are not able to “pick and
choose.” One-third of teacher candidates recognized that prestige, socioeconomic status,
and money place schools in different positions on the ﬁeld. While there may be other
characteristics of schools that placed them in a dominant position for students or
parents—or more experienced teachers—for teacher candidates in this study, schools that
served more aﬂluent populations had more money and prestige and thus were in a
dominant position. Schools that served less afﬂuent populations and had less money and
prestige were in a weaker position.

While teacher candidates thought getting a job offer in most districts in the area
would be hard, they made distinctions between districts in the source of the job difﬁculty.
As noted above, some teacher candidates indicated that afﬂuent suburban schools were
selective in hiring because they received many applications. Several teacher candidates
said they heard stories about 2000 candidates for one job in suburban Smithton districts.
These were the districts where teacher candidates thought they really had to be the best to

get a job.
103

Teacher candidates also thought it was difﬁcult to get a job in Smithton or low-
income suburbs, but they attributed the difﬁculty to the lack of funding for jobs rather
than selectivity by the district. For example, Rachel said, “they’re shutting down schools
so it’s not like I’m gonna go and apply for schools that are laying off teachers.” Teacher
candidates frequently mentioned teacher layoffs in Smithton as causing difﬁculty in
getting a job there, not that Smithton was selective because it had many candidates. That
it was hard to get a job in a particular school or district was not the same thing as the
school or district’s ability to be selective. Teacher candidates’ explanations of why it is
difﬁcult to get a job in high-income suburbs or urban districts demonstrated the opposing
positions of these districts. Affluent suburban districts were in a dominant position
because they could select whom they wanted. Urban districts, on the other hand, were in
a less dominant position because of ﬁnancial instability. Through their explanations of
the sources of difﬁculty in getting a job in different districts, teacher candidates classiﬁed
districts into opposing positions on the ﬁeld.

Just as teacher candidates recognized it was particularly hard to get a job in
particular schools, they also thought it was easy to get a job in some schools. Teacher
candidates thought that it was much easier to get a job out of state. Some teacher
candidates moved out of state to get a job, others just lamented the fact that the state did
not have as many jobs as other states. For example, Valerie said, “some people are
getting hired out of the state, some people don’t even care if you have your teaching
certiﬁcate, they’ll just hire you right on the spot. The standards in [this state] are so high
that makes it difﬁcult.” Like the teacher candidates above who equated good schools

with selectivity in teacher hiring, Valerie said it is the high standards for elementary

104

teachers in the state that made it difﬁcult to get a teaching job. Other states have lower
standards as they “don’t even care” if the teacher candidate is certiﬁed. Samantha
described Florida as “desperate” for teachers and she moved to Florida because she did
not think she would get hired in this state. She said,

The way that it is here in [state], with nobody getting jobs, I

mean, I went down there thinking I’ll be lucky to just get a

job. That’s what I thought but oh, no. ...they’re desperate

for teachers down there. The rumor that you heard, it’s true.
It’s true. They need them.

Samantha and several other teacher candidates moved out of state because they did not
think they could get a job in this state. Teacher candidates thought it was easier to get a
job in another state because they thought other states had lower standards and were less
selective. These perceptions came from recruiting fairs in which districts from other
states offered contracts after short meetings with potential teachers.

Teacher candidates also thought it was easier to get a teaching job in a charter
school. While one teacher candidate preferred to teach in a charter school, only 5 of the
27 teacher candidates considered charter schools on an equal basis with regular public
schools in their job search. Most teacher candidates either ruled out charter schools
completely or considered them only if they did not get a job in a traditional public school.
Richard said he would not apply to charter schools initially because “they’re not my ﬁrst
tier job search.” Likewise, Melissa said,

My plan is now is that I’m gonna avoid them as long as I
can and then possibly if it gets to like July/August and I

really feel like I want a teaching position, I might look into
it more.

Most teacher candidates were unfamiliar with charter schools and were wary about the

low salaries and lack of union protections in charter schools. Teacher candidates did not

105

see charter schools as equal alternatives to traditional public schools as a place to work.
One exception to this ﬁnding was that Aﬁican American teacher candidates were more
likely to consider charter schools as equal to regular public schools. That most teacher
candidates thought charter schools occupied a weak position on the ﬁeld of schools is
evident in the fact that most teacher candidates thought they could ignore charter schools
for the ﬁrst part of the hiring season but apply at the end of the summer and still get a job.
J ust as teacher candidates viewed schools in a dominant position as having high

standards for teachers and being selective in teacher hiring, they thought it was easy to
get jobs in charter and out of state schools because these schools were not selective. The
lack of selectivity in schools made them less attractive to teacher candidates. For
example, Stephanie said,

They ended up asking us one question. They basically gave

a contract to every single person that came. So I don’t

know if I really wanta go there because they’re, I mean,

they’re so desperate... There was a man there in sweat pants
and a hat and a plaid shirt, and he got a contract.

Stephanie did not want to work in a district that was “desperate” and would give a
contract to any candidate. Note that it was the selectivity of the district, rather than other
characteristics, that made the school undesirable.

Teacher Preference for Schools Occupying Weak Positions on the Field

According to Bourdieu’s cultural reproduction theory, teacher candidates in
dominant positions on the teaching ﬁeld would teach in schools that occupied a dominant
position on the ﬁeld of schools and teacher candidates in a weak position on the teaching
ﬁeld would teach in schools that occupied a weak position on the ﬁeld of schools. Thus,
examining the characteristics of teacher candidates that preferred to teach in schools in

dominant or weak positions on the ﬁeld may provide an assessment of their position on

106

the teaching ﬁeld. This section examines the characteristics of teacher candidates who
preferred to teach in schools in a weak position on the ﬁeld. The previous section
identiﬁed low-income schools, urban schools, charter schools, and out of state schools as
occupying a weak position on the ﬁeld of schools. That some teacher candidates
preferred these schools may indicate that they hold a weak position on the teaching ﬁeld.

Table 14 shows the percentage of teacher candidates who preferred schools in
Smithton Public Schools, urban or low-income suburban schools, charter schools, and
schools out of state. The percentage of teacher candidates who preferred to teach in
Smithton represents the percentage that named a public school in Smithton as their top
choice school. The other percentages were calculated similarly.

These frequencies suggest that the patterns of teacher candidates who wanted to
teach in Smithton, urban or low-income suburban, and charter schools were similar,
while teacher candidates who wanted to teach out of state were somewhat different.
African American teacher candidates were more likely to prefer to teach in Smithton.
Aﬁican American teacher candidates were also more likely to prefer charter schools and
urban or low-income suburban schools. Teacher candidates who graduated from high
schools with a high concentration of students eligible for free or reduced-price lunch
were more likely to prefer to teach in Smithton or other urban or low-income suburban
schools. Teacher candidates who were moving into teaching ﬁom another career were
more likely to prefer to teach in Smithton Public Schools or charter schools. There were
no statistically signiﬁcant differences between male and female teacher candidates except

that males appeared more likely to prefer to teach in charter schools.

107

Table 14 - Percentage of Teacher Candidates Who Preferred to Teach in Various Types

 

 

 

of Schools

Smithton Public Urban or low- Out of
Variable Schools income suburban Charter state
Total 6.9% 20.4% 2.2% 9.0%
White 2.1 15.2 0.0 8.7
African American 462*“ 55.6*** 14.3*** 10.3
Other minority 4.6 19.2 0.0 12.0
New college graduate 3.3”” 18.3 0.0" 12.9"
Mid-career changer 13.0* * 23.9 5.3“ 2.1* *
Female 6.6 19.2 0.8" 9.2
Male 9.1 25.0 8.3" 8.3
Non-high poverty high 48*" 19.1~ 2.0 9.2
school
High poverty high school 333*" 363'” 5-3 5-3
Note: N=289

~ p<.1, ** p<.01, *** p<.001.

The patterns were different for out of state schools. There were no differences
between racial or gender groups in choosing out of state schools as their top choice
school. Likewise, teacher candidates who graduated from low poverty high schools were
about as likely to name an out of state school as their top choice as those who graduated
from more affluent high schools. New college graduates were more likely than mid-
career changers to prefer to teach out of state, possibly because these teacher candidates
were also younger and less settled in a particular area.

The pattern for out of state schools may be different due to teacher candidates’
recognition that schools in other states occupied different positions on the ﬁeld,
depending on their relative status within their own state. For example, Stephanie is a

White, new college graduate whose mother teaches in an affluent district in suburban

108

Smithton. Stephanie wanted to teach in a neighboring district, but realized that she would
face stiff competition in that district and therefore sought out a job in another state that
she called “the best district in the country.” Stephanie recognized that some districts in
other states might have the same or more prestige as her preferred district in this state.

To disentangle the inﬂuence of the various teacher candidate characteristics on
wanting to teach in particular types of schools, I ran a binomial logistic regression. The
dependent variable is the log odds of the probability that the teacher candidate named an
urban or low-income suburban school as their top choice school. Due to the low numbers
of teacher candidates who preferred to teach in Smithton, charter schools, or out of state
schools, logistic models predicting who named these schools as their top choice did not
converge. Therefore, I only present the results for preferring an urban or low-income
suburban school.

Table 15 presents the results from a logistic regression on the likelihood of
teacher candidates with various characteristics preferring to teach in urban or low-income
suburban schools. African American teacher candidates were more likely to prefer to
teach in urban or low-income suburban schools; they had over 10 times the likelihood of
preferring to teach in an urban or low-income suburban schools than their White
colleagues. This may reﬂect their preference to teach African American students or to
teach close to where they lived as most Aﬁican American teacher candidates lived in
urban areas. Table 15 supports the results from the frequencies in Table 14 and
underscores the importance of race, as being Aﬁican American remained statistically

signiﬁcant even when controlling for other teacher candidate characteristics.

109

Table 15 — Logistic Regression Results Estimating the Likelihood That Teacher

Candidates Preferred to Teach in an Urban or Low-Income Suburban School

 

 

 

Variable Estimate Error Odds ratio
Intercept -1.68*** 0.23
Mid-career changer —O.23 0.38 0.80
Male 002 0.51 0.98
African American 2.32*** 0.57 10.19
Other minority 0.38 0.54 1.47
High poverty high school -0.52 0.67 0.60
Pseudo R2 0.082
N 249
*** p<.001.

Tables 14 and 15 suggest that the teacher candidates who had the best sense of the
ﬁeld—that is, the teacher candidates whose perception of the ﬁeld best matched the
research on principals’ preferences in teacher hiring—were least likely to apply to
schools in weak positions on the ﬁeld of schools. White teacher candidates and young,
new college graduates had the most accurate perception of the teaching ﬁeld. This
suggests they may be in the most dominant position on the ﬁeld. Yet African Americans
and mid-career changers were most likely to prefer urban or low-income suburban
schools Smithton Public Schools, or charter schools—schools that most teacher
candidates described as being in a weak position on the ﬁeld. This suggests there are
homologies or structural similarities between the teaching ﬁeld and the ﬁeld of schools.
Teacher candidates who had the best sense of the ﬁeld wanted to teach in schools that
were selective and avoided schools that were not selective. Yet teacher candidates who
did not seem to have an accurate perception of the teaching ﬁeld were most likely to want

to teach in schools that were viewed as not selective.

110

Discussion and Conclusion

Teacher candidates can identify the various forms of capital that structure the
teaching ﬁeld and recognize the independent effects of embodied cultural capital, human
capital, and social capital. Teacher candidates’ perceptions of the structure of the ﬁeld
correspond to principals’ hiring criteria as most teacher candidates emphasize the
importance of personality, ﬁt, and demographic characteristics in teacher hiring. That is,
teacher candidates realize that even though a successful candidate must have basic
qualiﬁcations, they think an individual’s personality and the ability to ﬁt into a school
determine who gets a teaching job and who does not. These characteristics highlight the
importance of embodied cultural capital as they represent traits that are acquired through
a socialization process (one that begins well before entering college) rather than explicit
instruction in a teacher education program. These forms of cultural capital are embodied
and present in how an individual carries oneself, interacts with others, and behaves in
particular social situations.

The notion of ﬁt is also a form of embodied cultural capital. While traditional
ideas of ﬁt deﬁne it as a multidimensional construct in which it depends on both the
candidate and the potential employer, many teacher candidates in this study deﬁne ﬁt as a
unidimensional construct. That is, some teacher candidates think they can ﬁt in all
schools and others cannot ﬁt in any school. Teacher candidates ﬁt if they have good
people skills, work well on a team, and are easy to get along with. From this perspective,
ﬁt is another form of embodied cultural capital as it represents interaction patterns and
personality traits. This chapter demonstrated how teacher candidates thought that those

who can ﬁt into a school are more likely to get a job. The next chapter will show that

111

teacher candidates also prefer schools where they think they ﬁt. Thus, perceptions of ﬁt
become important constructs in the teacher job search process.

While most teacher candidates recognize the criteria important to principals in
teacher hiring, some teacher candidates emphasize different characteristics. African
American teacher candidates and older teacher candidates, for example, are more likely
to emphasize qualiﬁcations and human capital rather than personality and embodied
cultural capital. White teacher candidates from middle class backgrounds just graduating
with their ﬁrst college degree are the most likely to recognize the structure of the
teaching ﬁeld represented in principals’ hiring criteria. That is, White middle class
teacher candidates have the best sense of the game and the easiest time negotiating the
teaching ﬁeld. As these teacher candidates also come from backgrounds mest similar to
principals and the majority of current teachers, it is likely that their cultural background
inﬂuences the characteristics they View as most important. Their habitus—shaped by
their cultural and social upbringing—allows them to see distinctions between teachers
based upon cultural traits. The habitus of other teacher candidates—such as African
Americans—focuses their attention on qualiﬁcations and instructional skills.

The variation in teacher candidates’ perceptions of the teaching ﬁeld is important
for two reasons. First, the approach teacher candidates take to the job search is likely
informed by how they view the structure of the teaching ﬁeld. Most teacher candidates
have a sense through their interactions with administrators, school personnel, and teacher
educators that ﬁt is important in the teacher labor market. Yet some teacher candidates
think professional qualiﬁcations are most important. As habitus mediates between the

structure of the ﬁeld and one’s perceptions, how teacher candidates view the teaching

112

ﬁeld may inﬂuence how they view schools. For example, teacher candidates who think
that personality and ﬁt are most important in hiring decisions may be more attuned to
personal characteristics of the schools in which they apply, while those whose attention
focuses on professional qualiﬁcations may likewise focus on professional attributes of
schools.

Second, the characteristics that teacher candidates emphasize on applications and
in interviews are likely to be those they think are most important in the job search. That
is, teacher candidates may present themselves in different ways to principals or other
hiring ofﬁcials depending on what they think is most important to the hiring ofﬁcial. If
teacher candidates emphasize the wrong type of capital in hiring situations, they may be
at a disadvantage in getting a job offer.

It is possible that teacher candidates who emphasize professional qualiﬁcations
over personality and ﬁt have accurate perceptions of a segment of the teaching ﬁeld. In
other words, it is possible that while most principals emphasize personal characteristics
when making hiring decisions, principals from urban or low-income suburban schools
care most about professional qualiﬁcations and teaching ability. Most existing research
on principal hiring decisions uses smaller samples that do not allow for making
comparisons between principals in different contexts. Future research on hiring decisions
should examine this further.

The importance of ﬁt and personality may reﬂect the particular context of the
study—elementary teachers in an area with an over-supply of teacher candidates. Caring
is a central component in descriptions of elementary teaching (N ias, 1999); it may be that

professional qualiﬁcations are emphasized more in secondary schools with their

113

disciplinary orientation. Further, the labor market context was such that basic
qualiﬁcations for teaching could be assumed. Principals did not need to make trade-offs
between certiﬁed teachers and other personal traits. The tight labor market conditions
may inﬂuence teacher candidates’ perceptions of the ﬁeld.

Many teacher candidates also make distinctions between schools based upon
cultural traits—namely prestige and selectivity. Teacher candidates recognize that it is
hard to get a job in many types of schools, but they attribute the difﬁculty to opposing
reasons. Afﬂuent suburban schools are considered highly selective and therefore
desirable, compared to charter and out of state schools that are perceived to have low
standards and therefore are undesirable. Yet teacher candidates attribute the difﬁculty in
getting a job in urban or low-income schools to the ﬁnancial instability and lack of jobs
rather than to a sense of selectivity. Urban, low-income, and charter schools are then
perceived by most teacher candidates as undesirable places to teach, in part due to their
low prestige. In viewing the selectivity of a school as a desirable characteristic, teacher
candidates appear to be engaging in herd behavior and assuming others have received
signals that these schools are desirable places to work (Banerjee, 1992). If individual
teacher candidates observe large numbers of people choosing a particular school, they
tend to ignore any information they do have and follow the behavior of others (Banerjee,
1992; Bikhchandani, Hirshleifer, & Welch, 1998).

That teacher candidates are in different positions on the ﬁeld is important because
teacher candidates’ positions structure their habitus. As habitus mediates between the

structure of the ﬁeld and one’s tastes and behavior, how teacher candidates view the

114

structure of the ﬁeld has implications for how they approach the job search. The next

chapter explores the practical sense that teacher candidates bring to the job search.

115

CHAPTER 4: UNDERSTANDING THE PRACTICAL LOGIC OF THE
TEACHER JOB SEARCH: ESPOUSED PREFERENCES AND PREFERENCES

IN USE

This chapter focuses on the processes by which teacher candidates ﬁnd teaching
positions to explore how their preferences shape their job search. The purpose of this
chapter is to develop and apply a theoretical understanding of how teacher candidates
make decisions about where they want to teach and how they use information while
making these decisions. In doing so, I respond to the limitations that existing frameworks
on teachers’ career decisions present by using cultural reproduction theory; this
framework highlights the practical sense that teacher candidates bring to the job search
and how these processes are situated within social and cultural contexts.

My analysis in this chapter indicates that teacher candidates’ behavior during the
job search leads them to reproduce their own schooling experiences by selecting a job in
a school similar to the ones that they attended as children. While teacher candidates
espouse preferences for schools with particular organizational characteristics that on the
surface reﬂect rational calculations of costs and beneﬁts, they prefer particular schools
partially based upon feelings of comfort, ease, and familiarity that are inﬂuenced by their
race and class. Further, teacher candidates’ methods of identifying open positions ﬁlter
their job options to schools that are socially proximal to them. In this way, teacher
candidates sort themselves into schools in ways that reproduce existing social divisions.

This chapter is divided into six sections. The ﬁrst section further develops the

theoretical framework outlined in Chapter 1 and focuses on the practical sense that

116

teacher candidates bring to the job search. In doing so, it highlights the social proximity
or distance that develops between teachers and schools and the difference between
teacher candidates’ espoused preferences and preferences in use.

The second section describes the espoused preferences teacher candidates
explicitly state as governing their job choice. The third section highlights how teacher
candidates are using their practical sense of what differences between teaching jobs are
important by focusing on districts rather than schools in their job search. In doing so,
many teacher candidates act in ways inconsistent with their espoused preferences. The
fourth section describes how the ways in which teacher candidates identify available
teaching positions restricts their job choices. Teacher candidates’ practical sense leads
them to ignore unfamiliar schools that may potentially match their espoused preferences.
The ﬁfth section highlights how teacher candidates avoid schools in districts where they
think they would feel uncomfortable. These are typically districts with high percentages
of minority and low-income students. Rather, teacher candidates seek out districts where
they think they would ﬁt in because the school community matches their own
background.

Finally, the last section suggests that teacher candidates have few sources of
information available to them to gather information about schools, particularly about
characteristics represented in their espoused preferences. This results in teacher
candidates making assumptions about the work environments in schools based upon
information they do have, student demographic and achievement characteristics. I
conclude by suggesting that ignoring the cultural and social contexts of teacher

candidates’ job search decisions may result in ineffective or inefﬁcient policies.

117

Theoretical Framework

This chapter synthesizes ideas from Bourdieu’s cultural reproduction theory
(Bourdieu, 1984, 1998) and theories of action (Argyris, 1985; Argyris & Schon, 1975) to
illustrate how teacher candidates’ job choice decisions are guided by their sense of where
they belong and their culturally informed perceptions of schools. As discussed in the
introduction, Bourdieu’s cultural reproduction theory emphasizes how individuals have
an embodied, practical logic, or a sense of appropriate actions given their position on a
ﬁeld. This practical understanding—rather than an explicit set of rules or preferences—
inﬂuences behavior. Likewise, a theory of action framework suggests there are
differences between an individual’s espoused theory—a theory someone uses to explain
his or her behavior—and that person’s theory in use—the logic that guides actual
decisions and actions. Together, these frameworks focus attention on how teacher
candidates may engage in actions during the job search that do not have their stated
preferences as intended outcomes.
Habitus and Social Proximity

Bourdieu’s cultural reproduction theory emphasizes how individual behavior is
guided by a practical sense of what is appropriate in a given situation (Bourdieu, 1980,
1998). This practical sense is embodied in individuals and guides speciﬁc actions even if
individuals are not adhering to explicit rules or strategies. Rather, individuals have an
embodied understanding of what is a natural course of action. The concept of practical
sense is different from a rational actor framework because it assumes decisions are made

more holistically, instead of a calculative weighing of the costs and beneﬁts of different

118

courses of action. It also assumes that individuals have a sense of what is desirable
without understanding the speciﬁc features that make it desirable.

Understanding one’s habitus is central to understanding his or her practical sense,
as habitus is the generating principle behind individual behavior. Habitus generates our
everyday cultural practices and perceptions and acts as a mediator between the structure
of the ﬁeld and our practices. As our habitus generates our tastes and practices, a social
proximity or distance develops between us and other actors. As discussed in the previous
chapter, activity occurs on a social ﬁeld in which individuals occupy a social space
deﬁned by the amount and type of resources held. Our position is closer to some social
spaces and more distant to others (Bourdieu, 1998).

There are two implications of social distance. First, we are less likely to come into
contact with some actors because our tastes and practices put us is different spaces
(Bourdieu, 1998). Teacher candidates are more likely to come in contact with schools—
and the principals and teachers in those schools—that are socially proximal to their own
position. Teacher candidates may become aware of a school because it is close to their
home or near a restaurant they frequent, or because themselves, a relative, or a ﬁiend
attended the school. Thus, what schools are familiar to a particular teacher candidate is
related to the teacher candidate’s social and cultural context. This explains how habitus,
through social distance, may essentially ﬁlter the schools to which a teacher candidate
chooses to apply.

Second, when we do come in contact with a socially distant position, it does not
appeal to us because we have opposing tastes (Bourdieu, 1998). Teacher candidates may

still come in contact with socially distant schools during the job search, but they are

119

likely to feel uncomfortable in these schools or view them as undesirable places to work.
Things may seem out of place or their interactions with personnel at the school may feel
forced or inappropriate. On the other hand, when teacher candidates encounter a socially
proximal school, their expectations for teacher-teacher or student-teacher interactions
may be met and their own interactions with school personnel would feel more natural.
When individuals with similar habitus are brought together, they feel a kindred
relationship with each other (Lamaison & Bourdieu, 1986). This afﬁnity is the result of
similar social experiences and conditionings (Lamaison & Bourdieu, 1986).
Espoused Theory and Theory in Use

A theory of action framework provides a useful analytic distinction to help
understand the notion of practice. This framework highlights the norms, values, and
assumptions behind individual behavior. Individuals hold a set of explicitly stated beliefs
and goals that constitute our espoused theories of action. These espoused theories outline
how we describe our behavior; it is our own explanation of what guides our actions
(Argyris, 1985; Argyris & Schon, 1975). Yet individuals may act in ways that are
inconsistent with what they espouse as their intentions (Argyris, 1985; Argyris & Schon,
1975). In this way, the distinction between espoused theories and theories in use reﬂects
the distinction between stated and revealed preferences in the economic literature. While
individuals may have espoused theories to explain their own behavior, another set of
norms, values and assumptions may better predict actions (Argyris, 1985). These theories
in use are powerful guides for one’s actions, but they are tacit and individuals are usually
unaware of how they may be inconsistent with their espoused theory (Argyris, 1985;

Argyris & Schon, 1975). Theories in use are tacit because they are undiscussable or

120

inaccessible to individuals (Argyris & Schon, 1978). In the teacher labor market, teacher
candidates may have espoused preferences that they provide as explicit criteria or
characteristics of schools that guide their decisions. Yet they may act in ways that are
inconsistent with these espoused preferences. Instead, the decisions and assumptions
made by teacher candidates during the job search may be better explained by
understanding their preferences in use.

Theory in use is similar to practical logic. They both represent explanations of
how everyday behaviors are guided by sets of assumptions about how things are related
without the individual actor recognizing those assumptions or explicitly following a set of
rules. Bourdieu (1977b) makes a distinction similar to that of espoused theory and theory
in use, arguing that the hypothesis that best predicts someone’s behavior may not be the
set of rules or goals that the individual provides for his or her actions. Indeed, the
presence of regularities does not mean that rules are being followed. Rather, regularities
may occur because individuals from similar social and cultural backgrounds respond to
situations using similar practical understandings of the structure of the social ﬁeld
(Bourdieu, 1977b). A theory of action framework suggests that individual actions have
“microtheories” that guide behavior in speciﬁc situations (Argyris & Schon, 1975). But
one person’s various microtheories are related to each other as they are all guided by a
similar logic, based upon a model of the behavioral world (Argyris & Schon, 1975). This
is similar to the practical logic of cultural reproduction theory that speciﬁes the habitus as
the unifying principle that generates everyday actions.

Theory in use and practical logic are not identical ideas, however. Indeed, Agyris

and Schon’s framework continues to assume individuals engage in goal-setting and

121

rational decision-making that Bourdieu’s cultural reproduction theory rejects, even if the
goals are tacit or unconscious to the actor. My use of the term preferences in use does not
assume that teacher candidates engage in this type of goal-setting. I use the term
preferences in use to highlight the perceptions teacher candidates have about what makes
a school desirable and the processes that teacher candidates undergo to ﬁnd a teaching
job. Still, the difference between espoused theories and theories in use remains a useful
analytic distinction as it emphasizes how individuals may act in ways that are
inconsistent with their stated beliefs. Rather, individual behavior may be congruent with a
separate, practical logic that is implicit and based upon experience with the social world.
Examining teacher career decisions, then, should involve trying to understand the
practical logic teacher candidates possess as well as how their espoused theories and
theories in use compare. A cultural reproduction theory predicts that teacher candidates
end up working in particular schools not entirely as the result of a rational calculation or
adherence to explicit rules, but because they act in accordance with their sense of their
social position and where they ﬁt. Existing research on teachers’ career decisions
suggests that teacher candidates have espoused preferences for particular features of
schools, such as supportive principals or small classes. This conception of teacher
candidates’ job choices assumes that teacher candidates are consciously establishing their
preferences and researching schools to ﬁnd the best match given their list of preferences.
A cultural reproduction view of teachers’ job choices, on the other hand, suggests that
teacher candidates are not acting out these preferences in a linear fashion. Rather, teacher
candidates are navigating their way through the hiring process using their practical

understanding of how they could make the best match.

122

Espoused Preferences

Table 16 shows survey data on the percentage of teacher candidates from the
initial survey that listed a given characteristic as one of the three most important features
of schools in their job search. Teacher candidates’ espoused preferences placed the most
importance on organizational characteristics of schools, location, and issues of
curriculum and curriculum materials. Teacher candidates had strong preferences to teach
in schools that support them in their ﬁrst year of teaching. Thirty-eight percent of teacher
candidates said working in a school that provided them with mentors or other forms of
beginning teacher support was one of the most important features in their job search.
Likewise, 32.5% of teacher candidates said working in a school with a supportive
principal was one of the most important characteristics. Teacher candidates also cared
about the relationship between teachers in a school. Almost one-quarter of the teacher
candidates (22.2%) said they wanted to teach in a school where the teachers get along and
17.0% wanted a school where teachers talk about their instruction.

The qualitative data mirrored this double meaning of staff collegiality. Teacher
candidates wanted to work in schools where teachers collaborated and instruction was a
collective practice, but they also wanted to be ﬁiends with their colleagues and feel part
of a family. For example, Heather described visiting a school where her peers had
completed their internships. She liked the staff because they were, “part of this whole
family at the school and I really am looking for a school like that, where I feel like I
belong and I have friends on the staff.” For Heather and other teacher candidates, the

social relationships among teachers were just as important as the professional ones.

123

:mOVQ * a—.VQ 2
.ome Z $3.5m REE 2: 80¢ 088 8% 3.2:. ”802

 

124

 

 

 

 

3 3 ed 3 mo 3 859% Sizes?
em 3 mm 2 M: Z 868% 83628-33
. . . . . . as a 2:863
m m o v \1 n m m _ m m m 38288208 088 05 80¢ mﬁoeam
ms E 82 ”.2 ...3 2 853m €658 233368 as“:
2 S E 2m: 3 2 nice 85.6 8228 88
m: as we we we 2: 8326 :25
a: as E SN 3: 2; has gen
3: _.2 n: 2: 3: of :88 s 82 _ as gauge
m: N: zwdm ”.2 4.2 q: 8:855 .65 36% was as mam <
4.2 we «.2 2: ea 3: nausea 8288 be 8388 e80
roam 1.2 3m .2 4.8 52 use 55, 63:88:66 sons .8 gaseous:
3: NE mi N: N. _ N SN 6:8 2:386 aoeam 236% 5.
m. a New 4.: e: 3N NAN 6&2. be 9 £585 686
EN 2N 2: an ...wsm Nam :63 was saw as ecu. <
New SN 2m 3m SN mam masses 856 as 38868 3:58:63.
com Em 6.8 EA mam 2a @2682 36:8 ozeoaasm
swam $4.6m so? .2858 $3M ea? 8&3 same assets 2385 page
Suﬁsm $385 biota: sectssw 83x: $83328 onwmumcoexese 393m.
cavemen imZ $38.3: ESQ :82th $538 as
.4533. .333 am one; Am
Schwom

£2. 5 833cm acetone: $02 moi... 2: mo 25 we UoEmZ 352038an ”moocouomoi womnoamm ,moamEvSwU 853,—. - 2 233.

Location was another school feature that was very important to teacher
candidates. Slightly less than one-quarter of teacher candidates said that being in a school
that was close to their home was one of the most important characteristics. Location
appeared to be even more important to teacher candidates in the qualitative data as
location was the most frequently mentioned characteristic they preferred. Teachers
wanted to work close to home. This was true both of older teacher candidates who own
homes and have children in school and of young college graduates, many of whom want
to live close to their parents. The interview data also suggested that wanting a school in
close proximity to home means that teacher candidates did not want to move in order to
ﬁnd a teaching job.

When teacher candidates shared their espoused preferences, characteristics of
students appeared less important to them and many teacher candidates explicitly stated
that these features did not matter in their job search. Teacher candidates were least likely
to say that student characteristics were one of the most important factors in their job
search. For example, only 1% of teachers said teaching high-achieving students was one
of the most important features. Yet, several teacher candidates in the interview sample
did want to teach in Title I schools or schools that served low-income students. Many of
these teacher candidates saw teaching as a form of service. For teacher candidates who
wanted to teach students from low socioeconomic backgrounds, they oﬁen spoke about
wanting to give back to their community or make an impact.

Further, the survey data suggested that the characteristics most important to
teachers are likely to vary among schools within a district. Of the ten most important

characteristics, seven are school-level characteristics, such as beginning teacher support,

125

supportive leadership, relationships between teachers, teaching practices, and school
discipline policies. Another factor, curriculum, could be determined at either the school
or district level. Two characteristics, availability of materials and proximity, are more
strongly associated with district characteristics and may vary less within a district.

Teacher candidates’ espoused preferences did vary somewhat. White teacher
candidates were more likely to say that having a staff that gets along was one of the most
important features and less likely to say that working in a school with many racial
minority students was most important. The pattern for teacher candidates from other
racial minority backgrounds was reversed; they were more likely to say that working in a
school with many minority students was one of the most important features and less
likely to say that having a staff that gets along was important. African American teacher
candidates were less likely to say that having beginning teacher support or a school
mission that matches their own were among the most important features in their job
search, but more likely to say that good condition of the facilities was. Compared to new
college graduates, mid-career changers cared more about having a mission that was
compatible with their own and less about having students from the same socioeconomic
background as themselves.

Activating Practical Logic and Preferences in Use

The espoused preferences of teacher candidates were consistent with previous
research on teachers’ career decisions. Teacher candidates cared about how they would
be supported in their teaching, including the availability of beginning teacher support,
support from their principal, and collaboration with their colleagues (Johnson &

Birkland, 2003; Kardos et al., 2001; Smith & Ingersoll, 2004). But the way teacher

126

candidates activated their practical logic during the job search was often inconsistent with
their espoused preferences. These preferences in use appeared to give more weight to
familiarity, feelings of comfort, and perceptions of schools based on student
characteristics. In particular, the way that teacher candidates approached the job search at
the district level reduced the importance of school-level organizational characteristics and
draws attention to how teachers decided to apply or not apply in particular places due to
feelings of where they would be comfortable. As described below, the strategies that
teacher candidates used to identify available teaching positions required that they be
familiar with and have social connections to schools. Further, the lack of information on
schools, especially on characteristics in teacher candidates’ espoused preferences,
narrowed their attention to student demographics and achievement.
Approaching the Job Search at the District Level

Teacher candidates in this study tended to focus on districts, rather than schools
within districts, in their job search. The level that teacher candidates chose reﬂected their
practical sense of what created distinctions between teaching jobs and highlights
contradictions between their espoused preferences and preferences in use. Survey
respondents were asked to name speciﬁc schools that they considered their most desired
and their least desired place to teach, although many teacher candidates only named
districts. Table 17 shows the percentage of teacher candidates who named a speciﬁc
school and the percentage that only named a district. When thinking about their most
desired place to teach, 39.8% of teacher candidates only named a district. The difference
was more apparent when teacher candidates thought about places in which they did not

want to work, suggesting that teacher candidates were especially likely to avoid entire

127

districts as places they did not want to work. ‘2 To some extent, it made sense to focus on
districts rather than schools when looking for a job; districts did organize much of their
recruitment and screening through central human resource departments. Thus, teacher

candidates appeared to be responding to the structure of the teacher labor market.

Table 17 - Percentage of Teacher Candidates Who Named a Speciﬁc School, Named a
District but Not a School, or Did Not Name Any District as Their Most Desired and Least

Desired Places to Teach

 

 

Type ofregoonse Most desired school Least desired school
Named a speciﬁc school 55.7% 30.1%
Named a district but not a school 39.8 53.3

Did not name either school or district 4.5 16.6

 

Note: These data come from the initial survey. N=289.

Yet the reasons teacher candidates gave for focusing on districts in their job
search indicated other motivations. One reason teacher candidates gave was that they did
care about district-level features such as geographic location, availability of instructional
materials, and, to a lesser extent, salary. In this way, teacher candidates may have acted in
accordance with some of their espoused preferences when they focused on districts in
their job search. However, the focus on districts also suggests that their preferences in use
were inconsistent with their espoused preferences because it placed more weight on
district, rather than school, characteristics. Although teacher candidates talked about

selecting schools, they actually focused on geographic areas or districts.

 

'2 The 5% of teacher candidates who did not name either a school or a district represents item nonresponse.
Likewise, the 17% of teacher candidates who did not name either a school or a district represents a
combination of item nonresponse and, most likely, an indication that some teacher candidates will teach
anywhere. Some teachers who did not name either a school or district as least desired wrote in that they
would work anywhere.

128

The second reason that teacher candidates focused on the district in their job
search is that they saw fewer differences between schools in a district than they did
between districts. Seventy percent of the teacher candidates interviewed saw few
differences between schools within their desired districts. For example, Kelly said, “I
primarily focused on the district... When I looked at schools within the districts, it
seemed from the information I could get that they were the same.” Like Kelly, many
teacher candidates saw fewer differences between schools than between districts.

This takes on increased importance as teacher candidates anticipated staying in
one district once they were hired. Claire, for example, recognized that district personnel
policies reward teachers who remained in one district and she felt she would be “stuck”
in a district because, “Well, with retirement and tenure and the job market that it is, it’s
really hard to switch districts and once, like if you’re in a district for ten years, you lose
all those years if you switch districts.” Claire thus focused her job search at the district
level because, if you must spend your whole career in one district, “you want it to be one
that you like.” This raised the stakes of where they took their ﬁrst job because teacher
candidates did not want to be “stuck” in the wrong district. Teacher candidates like Claire
and others who used the district as the focus of their job search were acting based on their
practical sense of what differences between teaching jobs were important. Many teacher
candidates who applied to homogeneous districts thought distinctions between districts—-
such as the community served and location—were the most important in deciding where
they wanted to spend their career.

In contrast, teacher candidates who wanted to teach in large urban districts were

more likely to talk about differences between schools within a district, which was not

129

surprising given the greater variation within large districts. Teacher candidates who
wanted to teach in Smithton, for example, were more likely to recognize that not all
Smithton schools were the same. In particular, African American teacher candidates were
more likely to recognize differences between schools in a district. On the survey, 30% of
African American teacher candidates named two different schools in the Smithton district
as their most desired and least desired schools, compared to only 3% of teacher
candidates from other racial backgrounds who chose two schools in the same district as
their most and least desired schools. The differences between teacher candidates due to
race is noteworthy and suggests that African American teacher candidates were activating
a different practical logic or sense of how the teacher labor market is structured than
teacher candidates from other racial groups. African American teacher candidates—
although they largely conﬁned their job search to Smithton Public Schools and charter
schools—saw a lot of distinctions between schools. Other teacher candidates, who were
mostly White, avoided Smithton schools and viewed them as all the same. The
composition of teacher candidates’ social networks, described below, may contribute to
how they came to know different districts.
Identifying Open Positions

Teacher candidates could not apply to jobs if they did not know they were
available. By far, the number one way teacher candidates identiﬁed open positions is
through the Internet. Teacher candidates looked at both individual district websites and at
consortiums of education agencies, although they were most likely to go directly to the
personnel website of speciﬁc school districts. According to popular lore, the Internet is

supposed to be a great equalizer, providing access to a wide range information in a non-

130

discriminatory way.l3 Thus, teacher candidates in one metropolitan area should all have
equal chances to learn about job opportunities in various districts; districts’ use of the
Internet to publicize their available teaching positions on the Internet should not have any
effect on teachers’ job searches, so long as they post jobs in a public space in some
manner. Teacher candidates’ use of the Internet to identify available teaching positions,
however, did impact which open positions they identiﬁed and pursued.

More than three-quarters of teacher candidates went directly to the websites of
speciﬁc districts in which they wanted to work to check for available positions. This was
not surprising given that job postings and application screening are usually organized by
district human resource personnel. Yet the districts which teacher candidates decided to
check for open positions were limited to those with which they were already familiar. For
example, when asked if she applied to every open teaching position she saw, Claire
replied, “Well, the only districts that I’m checking are the ones I’d potentially be
interested in. So that kinda narrows it down.” Thus, Claire’s espoused preferences for a
supportive principal and collaboration among teachers were not directly supported by the
processes she uses to identify open positions, which emphasized two preferences in use:
location and familiarity.

Claire elaborated by saying she was “just going on ones that I have heard about or
ones in this area because I think that’s the scariest thing about teaching is you don’t know
what you’re getting yourself into unless you know about that school district.” In other

words, Claire used a self-imposed ﬁlter of familiarity. Positions in districts that were

 

'3 All teacher candidates interviewed for this study had Internet access and, despite the fact that they were
not asked this question directly on the survey or in the interview, the frequency with which they discussed
using the Internet suggested they were experienced lntemet users.

131

unfamiliar to her were not even considered. Claire would not know those positions were
open because she relied on checking speciﬁc district websites. Claire’s narrowing of
districts she would “potentially be interested in” occurred early in the job search process.
She relied on her practical sense of what it would be like to teach in particular districts
before she investigated whether there were open teaching positions. This practical sense
was structured by informal experiences in or stories about these various communities and
it structured whether she perceived these districts as desirable places to work.

In contrast to Claire’s strategy of focusing on speciﬁc district websites, teacher
candidates who identiﬁed open positions through county-wide consortiums had more
opportunities to learn about positions in districts with which they were not already
familiar as they could search a wide geographic area for teaching jobs. Still, even the way
teacher candidates used consortiums restricted the types of schools to which they were
exposed. For example, about 40% of teacher candidates who used county-level education
job banks did not pay equal attention to all teaching jobs posted on those consortiums.
For example, Richard said,

There’s a couple of general websites that, they have all the
districts, say in [the state], but I don’t just go out to any
district, some of them I don’t even know. I think there’s
this one, [Cross Creek], I have never heard of that one, and
it may be here near [Redville] . .. I’ve never heard of it
before, where is that one? But in general, I’d like to be
closer, [River Rock] [Orangelake is] starting to get a
little bit further north, but I would be interested in going up
there, again I have some contacts up there, it draws me up
in that direction. And then I have some contacts that are up
in [Hampton]. A little farther, probably 40 minute drive,

but they have a position that’s a middle school math teacher
posted right now.

As in Claire’s experience, we see the importance of familiarity in Richard’s strategies for

identifying available jobs. While a short commute was one of Richard’s most important

132

espoused preferences, he did not want apply to districts he had never heard of before.
Richard was hesitant to respond to the Cross Creek job, although he deﬁnitely wanted to
apply to Orangelake and Hampton—two districts that were farther away but familiar to
him through his social network.

In addition to a sense of familiarity and comfort, Richard’s experience also
emphasizes how teacher candidates relied on their social networks to help them locate
teaching jobs. Richard was particularly active in using his social network on the job
search. He said,

Then you continue to follow through with word of mouth
with friends and friends of friends who are working in
districts. Just try to get information, you know, making sure
people are aware and knowing that I need help. An
administrator that I worked with at student teaching, she
did a mock interview for me. And then, a friend’s dad

used to be a superintendent in a district and he, he’s at least
reviewed some information and given me feedback.

Richard’s social network included the principal he worked with during his student
teaching and his friend’s dad. Like Richard, many teacher candidates relied both on
professional colleagues—such as contacts from their student teaching, tutoring, or
substitute teaching experiences—and personal friends to help them identify open
positions.

Like Richard, Vanessa also had a wide social network because she worked as a
teacher aide in Smithton Public Schools and had been tutoring adults in a community
college. Thus, she knew many people who worked in education. Yet Vanessa and
Richard talked about very different schools when describing their job search, despite
graduating from the same teacher preparation program. Richard’s network was focused in

wealthy suburbs and allowed him to learn about available positions in these more afﬂuent

133

areas. Vanessa’s network was focused in Smithton Public Schools and in low-income
suburbs immediately surrounding Smithton. The segregation of the Smithton
metropolitan area by race and socioeconomic status resulted in Vanessa and Richard
having different social networks. Thus, their similar strategy of using social contacts to
help secure a teaching job contributed to teacher sorting as they identiﬁed jobs in schools
in a familiar social position.
Deciding Where to Apply

Teacher candidates’ methods for identifying available positions served as an
initial ﬁltering mechanism in their job search as they did not learn about all open
positions in the area. Yet many teacher candidates did not apply to all open positions they
did identify. Some teacher candidates said they would work anywhere and did not claim
to use any ﬁltering criteria once they identiﬁed an open position. Yet these teacher
candidates were in the minority. While 40% of teacher candidates made it clear in their
interviews they would not apply to some districts, 20% of teacher candidates said they
would apply to any open position. The remaining 40% of teacher candidates gave
conﬂicting evidence, indicating at one point in the interview that they were open to any
job and indicating at other times that there were some teaching jobs they would not
accept. Teacher candidates were hesitant to avoid particular schools altogether because of
the tight labor market. Still, 80% of teacher candidates in this study who recognized the
lack of available jobs in the area indicated they would not apply to every open teaching
position they identiﬁed. This section explores the extent to which teacher candidates
applied to a narrow set of schools and the characteristics associated with where teacher

candidates did not apply. The schools teacher candidates avoided were those they told me

134

they felt uncomfortable in or assumed they would be uncomfortable. This uncomfortable
feeling was based on the lack of afﬁnity between their habitus and the school community.

Table 18 reports survey data on the percentage of teacher candidates who planned
to apply to various types of schools.14 It was clear that teacher candidates did not plan to
apply to all schools, even within one geographic area. Most teacher candidates—almost
three-quarters—planned to apply to moderate-income suburbs around Smithton. Yet
under half of teacher candidates planned to apply to low-income suburbs around

Smithton and just one-quarter planned to apply to Smithton Public Schools.

Table 18 - Percentage of Teacher Candidates Who Planned to Apply to Various Schools

 

 

Type of school Percent
Regular public schools in Smithton 26.1
Regular public schools in high income suburbs near Smithton 67.5
Regular public schools in moderate income suburbs near Smithton 74.5
Regular public schools in low income suburbs near Smithton 48.0
Charter schools 46.7
Private schools with a religious afﬁliation 37.9
Private schools with no religious afﬁliation 46.4
Schools in rural areas 60.0

 

Note: These data come from the initial survey. N=289.

Denise, like several teacher candidates in the study, would not apply to Smithton
Public Schools because it was “an inner city school” where she would be “totally out of
[her] element.” Denise explained,

As much as idealistically you’d like to be somewhere to
make a difference, I have to recognize my limitations and
know if I’m not entirely comfortable I wouldn’t be doing a
good job. So I’d like to stick at least somewhere within my
realm of experience socially. You know, I mean, like
somewhere in the suburbs.

 

" Chapter 5 examines in more detail where teacher candidates applied and the variations between teacher
candidates in where they applied.

135

Denise recognized that inner city schools needed good teachers. But she also believed
that she would be uncomfortable in Smithton Public Schools and preferred to teach in a
school that was “within [her] realm of experience socially.” Because Denise was not
familiar with the experiences of inner city children, she would be “nervous.”

Viewed through a cultural reproduction lens, Denise preferred to teach in a school
where her habitus was structured by similar experiences of her students. As our habitus is
structured by past experiences, when we interact in situations similar to those past
experiences, actions feel natural. It is this sense of feeling that actions are natural or
ﬁtting in that teacher candidates described as feeling comfortable. On the other hand,
when we are placed in a social situation that is foreign to our past experiences, our
habitus provides us with few guides for appropriate behavior and actions feel forced or
we notice thinking through what would be an appropriate action. When teacher
candidates were unsure of the natural course of action, they described a sense of feeling
out of place and being uncomfortable. This was what Denise expected to feel if she were
to teach in Smithton.

Denise’s experience highlights a preference in use many teacher candidates had
when thinking about where they wanted to work. When teacher candidates explained why
they wanted to work in particular schools, about half of them indicated that these schools
were desirable because they were places where they felt comfortable. Almost all
instances of teacher candidates talking about wanting to work in schools where they felt
comfortable occurred when describing why they focused their job search in speciﬁc areas
or avoided particular districts. Only one teacher candidate explicitly said that comfort

was a criterion she used to evaluate potential jobs. This suggests that while feeling

136

comfortable was an important preference in use, it was not an espoused preference.
Teacher candidates were more active in seeking out job opportunities in places they felt
comfortable. These included schools where they student taught and schools they attended
as a student, as well as schools they thought were “like” those they with which have
previous experience. On the other hand, teacher candidates avoided districts where they
felt uncomfortable, or, rather, assumed they would feel uncomfortable.

The role of feeling uncomfortable in teacher candidates’ job decisions illustrates
how their practical logic led teacher candidates to sort themselves into schools. Not all
teacher candidates felt uncomfortable in Smithton schools, as Denise did. Indeed, African
American teacher candidates felt most comfortable in Smithton as they had larger
numbers of African American teachers on staff. For example, Zoe said, “I prefer not to
work at a school where I am the only minority on the staff. . .It’s my comfort level. I don’t
want everybody to come to me and look at me as the minority spokesperson.” Further, it
was not the case that teacher candidates’ comfort level increased monotonically with the
socioeconomic status of the community surrounding the school. Actually, several teacher
candidates—including Denise—felt uncomfortable in both very afﬂuent and very poor
communities. A pre-student teaching visit to a school in Hampton, an afﬂuent community
near her hometown, made Denise think she did not want to work there. She said,

Well, like in [Hampton], it was a wonderful building. I’m
sure the people were nice but they made me nervous. I
didn’t feel like I’d ﬁt in, but I can’t put a ﬁnger on why. I
just felt it. I don’t drive a fancy car. I don’t wear the
same clothes. ...I didn’t spend a lot of money on those
extemals. And when I’m around people that do, I guess I
feel intimidated. Sounds awful materialistic but I know
that’s one thing that makes me nervous. But in my

community, in [Lakeview], I’m comfortable. We’re all
working hard to raise our kids. None of us have money.

137

Denise exempliﬁed the embodied aspect of teacher candidates’ practical sense when she
described her feelings of discomfort in Hampton. Denise recognized that the tastes and
dispositions of teachers in Hampton—their habitus—did not match her own. Denise had
a bodily reaction to this mismatch—she was nervous and intimidated. Thus, Denise did
not feel comfortable in Smithton or in Hampton, but was most comfortable when the
social position of the school matched her own. She could understand the experiences and
attitudes of both students and teachers in her home community—that is, there was an
afﬁnity between their habitus—and so preferred to work there. Denise’s experience on
the job search illustrates how teacher candidates avoided schools where their habitus
provided them with little knowledge of appropriate behavior and where the social
experiences of students and teachers in the school did not match their own. As teacher
candidates came from varying social and cultural backgrounds, they sorted themselves
into schools based upon their own background.

Denise’s experience underscores the role of social class in teacher candidates’
perceptions of where they “ﬁt.” Denise picked up on the cultural practices of other
teachers that may indicate their social class and implied that these practices affected the
collegiality in the school. This suggests that beneath teacher candidates’ espoused
preferences for collegiality is a desire for a sense of familiarity, social ease, and class
similarity. Indeed, these factors were interrelated as teacher candidates were most likely
to have positive perceptions of staff collegiality in schools where they felt comfortable.
Gathering and Using Information in a Job Search

A rational decision-making approach to teachers’ career choices assumes that

teacher candidates have espoused preferences for schools and compare schools to each

138

other based upon those espoused preferences. This assumes that teacher candidates have
access to and use information about the school characteristics they most prefer. While the
teacher candidates in this study did research schools, the processes they used to gather
and interpret information about schools were sometimes inconsistent with their espoused
preferences. In particular, teacher candidates’ most important source of information
focused their attention on student demographic and achievement information. Further,
teacher candidates did not seek out information on unfamiliar schools. The practical logic
that teacher candidates brought to the job search led them to interpret student
characteristics as indicators of school organizational characteristics they preferred and
ignored schools where they did not already have information.

Sources of Information

 

Teacher candidates had access to several sources of information during their job
search, although these sources varied in the types of information available and the
distinctions they made between schools. Tables 19 and 20 present data on teacher
candidates’ sources of information. Table 19 presents survey data on their sources of
information on their most desired and least desired schools. Table 20 uses interview data
and lists the percentage of teacher candidates who named various sources of information
as how they know what schools are like. While the qualitative and quantitative data
showed differing emphases on the most important source of information—perhaps due to
questionnaire wording15 —they both indicated that teacher candidates had three main

sources of information: the Internet, word of mouth, and interviews.

 

‘5 There are two possible reasons for this disagreement in the data. One, because the survey data asks
teacher candidates to name speciﬁc schools as their most desired and least desired places to teach, they
listed schools that have high salience for them. Thus, they likely listed schools in which they knew teachers

139

Table 19 — Sources of Information on Teacher Candidates’ Most Desired and Least

Desired Schools, as Indicated on the Initial Survey

 

Most desired Least desired

 

Source of information school school
Conversations/interview with teachers in this school 71.5 54.7
Conversations with family/ friends who have direct

experience with this school 69'2 62'2
School or district website 66.3 36.1
Other direct personal experience with this school 59.9 49.4
Experience student teaching at this school 48.2 34.2
Conversatlons/rntervrew With the prrncrpal of this 46.8 250
school

My mentor/cooperating teacher 46.2 28.4
Newspapers/Media/TV 3 1.3 46.7
My university supervisor 23.3 28.2
N 214 195

 

Almost all teacher candidates said they used the Internet to gather information
about schools, including speciﬁc school and district websites and websites such as
www.greatschools.net. While the Internet did have information about a broad range of
schools, the type of information available was limited. Teacher candidates generally
obtained information on student and community demographics and test scores from the
Internet. For example, Stephanie described how she used the Internet to help guide her
decision-making:

I use GreatSchools.com as one. It gives you statistics on the
racial makeup of the school and the free and reduced lunch
percentage. And then it gives some parent reviews which
are a little skewed But they also give you test results, so

how the school’s doing But then I also use the website
epodunk.com which gives you information on community.

 

or other people with some experience at those schools. Two, the survey asked speciﬁcally about using the
school or district webpage as a source of information from the Internet. While many teacher candidates did
visit school websites directly, they also relied on general websites to gather information on schools.

140

...And then I also use, this is an odd one but I use Century
21 .com just to get an idea of the homes that are in the area
and also, it gives you an idea of the schools.

While Stephanie seemed to be a critical consumer of information—that is, she was aware
that Internet information was “skewed”—her use of the Internet narrowed the distinctions
she was able to make between schools. This illustrates how teacher candidates’
preferences in use did not support their espoused preferences. Websites rarely provided
information on teacher candidates’ espoused preferences, such as principal support and
teacher collegiality. Instead, the Internet focused their attention on student and
community characteristics. Stephanie said she strongly preferred to teach in a school with
a progressive instructional philosophy, but when gathering information to “help [her]

make that decision,” Stephanie focused on demographics, test scores, and house values.

Table 20 - Percentage of Teacher Candidates Who Said They Know About Schools

Through Various Sources of Information, as Indicated Through Interviews

 

 

 

Sources of information Percent
Internet 85.2%
Ask questions in interview/visit 59.3
Visit, observe, walk around school 59.3
I don't really know 55.6
Educational colleagues-Teachers 51 .9
Interactions during interview/visit 44.4
Ask friends who live in the area 40.7
Educational colleagues-Student teachers 29.6
I've lived here for a while 18.5
Educational colleagues-University 14.8
I've been working there 14.8
Look at test scores 14.8
Reading news 11.1
Pretend to be a parent 7.4
Note: N=27

141

Just as teacher candidates identiﬁed available positions using word of mouth, they
also gathered information about what it is like to teach in various schools through their
personal and professional social networks. When teacher candidates heard of openings in
various schools, they asked their professional network about those schools. Vanessa said,
“Well, the schools that I’m going to apply at, other teachers on the outside know those
teachers, schools, principals, so I can ask.” Yet teachers relied on personal contacts as
much as professional contacts to gather information and form opinions about the
desirability of teaching in particular districts. These personal contacts did not work in
education and therefore may not have had insight into a school from a teacher’s
perspective. For example, when Courtney was researching schools, she sought out
information from residents in the area. She said, “If I know somebody in the area, give
them a call and say what do you think about this community. Do you know anything
about the school district?” While teacher candidates did value information from other
teachers who had more ﬁrst-hand knowledge of working in schools, they also relied on
information from community members and parents. This strategy for gathering
information about schools illustrates how teacher candidates’ preferences in use may not
be consistent with their espoused preferences. While area residents may have information
about schools, they were less likely to have information about the characteristics teachers
said they most preferred.

Teacher candidates also obtained information on schools during job interviews.
Except for the few schools which teacher candidates knew about through direct
experience working there (perhaps as a substitute, student teacher, or tutor), asking

questions in a job interview was the best way to learn about a school. But teacher

142

candidates also formed perceptions of schools during the job interview just by interacting
with other people. This type of information was distinguished from asking questions in an
interview because teacher candidates focused less on the content of the discussion than
on the nature of the interaction. Teacher candidates talked about getting a “feel for the
principal and other teachers” from their clothes, their manner of speech, and their body
language. Amanda thought that talking with potential colleagues was the best way to
know what it would be like to work with them. She was looking for “Friendliness,
personable. You can just tell people are not being genuine when you see them one on
one.” Teacher candidates had a hard time expressing how they get a sense of schools
through these types of interactions, but that does not diminish their importance. It was
these interactions that guided teacher candidates’ perceptions of whether they felt
comfortable in a school.

Indeed, teacher candidates also felt they could get a sense of a school without
interacting with anyone. Claire described how she was eager to work in a school due to
the positive atmosphere she felt. She said,

I was really, really struck, and I’m being completely honest
when I say this, when I walked in the building, there was a
very positive atmosphere. It was kinda weird actually

because I didn’t see anyone initially but it just felt like a
really good place.

Even though Claire did not talk to anyone when she ﬁrst entered the school, she was
“struck” by a positive feeling about the school and said it “just felt like a really good
place.” Claire may have tacitly recognized an afﬁnity between her habitus and that of the
school community when she entered the building. A cultural reproduction framework
suggests that it is teacher candidates’ habitus that provides them with a feel for where

they ﬁt. When teacher candidates interacted in a situation that corresponded to their

143

habitus, they felt comfortable and did not notice cultural markers that structured the
interaction. On the other hand, when they interacted in a situation that did not correspond
to their habitus, they noticed distinctions and believed that things were “out of place,”
resulting in a feeling of discomfort.
Missing Information

Over half of teacher candidates indicated that they really did not know about
some features of schools. Table 21 shows the percentage of teacher candidates who did
not know whether or not their most desired and least desired schools had the top ten
features teacher candidates most strongly preferred. This table shows that teacher
candidates lacked a lot of information about schools they considered their top choice
school. For example, while 38% of teacher candidates strongly preferred to teach in a
school with beginning teacher support (see Table 16), 48.9% of teacher candidates did
not know whether their top choice school provided mentors or beginning teacher support.
This suggests that the sources of information teacher candidates had on schools did not
provide the information necessary to make informed decisions. Teacher candidates were
most likely to have information on the geographic location of the school, the availability
of textbooks and other instructional materials, and whether the schools had a good
reputation for teaching practices. Signiﬁcantly, teacher candidates were least likely to
know about their most important espoused preferences—the provision of mentors or
other forms of beginning teacher support and relationships between teachers in the
school. This again highlights how teacher candidates’ preferences in use may not be

consistent with their espoused preferences.

144

Table 21 - Percentage of Teacher Candidates Who Did Not Know Whether Their Most

Desired and Least Desired Schools Had the Characteristics They Named as Most

Important

Most desired Least desired

 

School characteristic school school
Mentors/beginning teacher support are provided to 48. 9% 66.7%
new teachers in this school.

This school has a supportive school leadership. 23.8 53.2
Textbooks and other instructional materials are 12 6 24 3
available when needed in tlus school. ' '
Teachers in this school get along well. 34.5 56.4
This school is close to my home. 3.57 3.97
This school has an effective student discipline policy. 24.4 33.7
13150 35:00] 5 missron or vrsron rs compatible With 220 53.0
This school has a good reputation for teaching 14.3 252
practices.

Teachers in this school talk about their instruction. 33.1 57.5
This school has a curriculum that I like to teach. 19.7 47.4
N 214 201

 

Note: These data come from the initial survey.

Further, Table 21 shows that teacher candidates were much more likely to lack

information about schools in which they did not want to work. For some characteristics,

teacher candidates were twice as likely to lack information on their least desired school

than on their most desired school. This included whether the school had supportive

leadership (53.2% compared to 23.8%), availability of textbooks (24.3% compared to

12.6%), whether the school mission or vision was compatible with their own (53.0%

compared to 22.0%), and whether the school has a curriculum that the teacher likes

(47.4% compared to 19.7%). This suggests that rather than gathering information about a

number of schools to assess which school would give them the most satisfaction, teacher

candidates eliminated some schools after very little consideration.

145

As noted above, teacher candidates focused their job search efforts in places that
were familiar to them. If teacher candidates were unfamiliar with a school prior to
entering the job market, they were less likely to do research on a school to see if it
matched their preferences. For example, Claire received three job offers simultaneously
during her job search. Two offers were in districts which she knew a great deal about—
one from student teaching and one from extensive research she had done about the
district. Claire “never gave much consideration” to the third offer, however. While she
did apply, she did not research schools in the district through the Internet or try to visit
the schools as she had in the others. Claire did engage in a rational decision-making
process as she “looked at the pros and cons” of the various schools, but Claire never
included the third school in her list of pros and cons. She just “didn’t know anything
about” the third school. Rather than try to gather information about the school to see how
it might match her preferences—and despite a positive interview where she liked the
principal and thought the “attitude was so welcoming”—Claire never considered the third
school alongside the other two.

Claire’s experience, like many of the others’ experiences, suggests that when
teacher candidates lacked information about potential schools or districts, they did not
always try to ﬁll in the gaps. Rather, teacher candidates used proxies for the
characteristics they said they really cared about, again belying their espoused preferences.
When teacher candidates did not have information about something, they were likely to
make assumptions based on the population of students. For example, Stephanie cared
deeply about teaching in a school with progressive pedagogies and approaches to literacy

instruction, but her perception of whether schools would have progressive practices came

146

from her student teaching experience in a suburban school with a high concentration of
African American students. Stephanie said,

My assumptions based on what I’ve experienced are that

it’s gonna actually be a more White area and a more middle

class to afﬂuent community. Now, I could be wrong and

I’m willing to be proved wrong but at this point, when you

have very little information to go on based on schools, I’m

using demographics to guide my decisions.
Stephanie felt she had to make “assumptions” in the face of “very little information to go
on.” Her practical sense about the relationship between progressive practices and student
demographics came from how her habitus led her to interpret her student teaching
experience. She was placed in a predominantly African American community where the
median income was lower than other parts of the county (although still higher than
Smithton and the state average) that is just outside the Smithton borders. She explained,

I just have had so many problems this year that I’ve been

really turned off by the demographics that I’ve seen. I get

very little parent support and I think that’s part of it. And I

think nobody really is aware what’s going on in the school

...they’re failing within the central ofﬁce so then you

would expect them to be failing outside. And I see that as

the demographics changed, the infrastructure has changed.

And I can’t say for sure if it is really related but at this

point in my career I’m letting it help me decide.
Stephanie linked the infrastructure of the district to the changing demographics and is
“turned off by the demographics.” Because she saw this relationship in one district, she
extrapolated to other districts. That is, she assumed that other districts with the same
demographic characteristics would also lack adequate infrastructure and parent support,
and will be “disorganized” and “failing.”

Heather also thought she knew what it would be like to teach in a school based

upon the students it served. Despite a desire to teach in an urban school due to her

147

religious beliefs, Heather avoided particular neighborhoods in an urban district because
of her assumptions about the possibility of success. She said she applied to schools in one
neighborhood but not another and explained,

It’s, you know, really dangerous and I’d probably be the

only White person per mile... There’ll probably be some

needy kids there but it’s not gonna be the lowest of the low,

I guess. I don’t wanta, in my ﬁrst year of teaching, go into
something that would be like impossible to succeed at.

Heather, like several other teacher candidates, felt that it would be impossible to succeed
in some schools, even though she only knew the neighborhood of the schools. Heather
drew drawing on cultural characteristics of the community in forming these perceptions
and linked the absence of White people from an area to the “lowest of the low” and the
ability to be successful as a teacher.

Another proxy that a few teacher candidates used in the face of missing
information about schools is test scores. Four teacher candidates said they made
assumptions about schools based upon their test scores. Daniel, for example, was
concerned about principal leadership. He wanted to work in a school where the principal
had a vision for the school. He looked at trends in test scores to give him information
about the leadership and teachers in a school. He said,

Just cause you’ve got the mean state average doesn’t mean
a lot. But if you’re consistently, you know, kind of right
there or a school getting better, I think you can kind of
gauge off of that. I think if you’re consistently bad then

something’s gotta be up. Whether it’s teachers, leadership,
curriculum, something.

Daniel assumed that schools with consistently low test scores did not have effective
leadership. Daniel, like a few other teacher candidates who said they learned about a

school by looking at test scores, was not saying that high test scores themselves are

148

important, but that they served as indicators for features they did consider important.

While these assumptions may reﬂect beliefs about effective schools present in the media,

they also reﬂected a teacher candidates’ practical sense of the ﬁeld of education.
Discussion and Conclusion

This chapter outlines the practical logic that guides teacher candidates’ behavior
during the job search. This practical logic is structured by teachers’ social and cultural
background and focuses their attention on particular districts, leads them to ignore
schools in other districts, and affects how they interpret the limited information they have
about various schools. In this way, teacher candidates’ practical sense of the labor market
emphasizes characteristics such as comfort and familiarity that guide job search
decisions. The strategies teacher candidates use to identify open positions results in them
not learning about opportunities in districts with which they are unfamiliar. Teacher
candidates often eliminate unfamiliar districts from their job search without trying to
become more familiar with them because they assume they will be uncomfortable in
those places. Teacher candidates’ strategies for gathering information on schools
emphasizes perceptions of comfort as they give equal weight to the nature of interactions
and “feeling” of a school as they do to the content of the interviews.

Previous studies of teachers’ career decisions emphasized their preferences for
organizational characteristics that support their instructional growth and particular needs
as new teachers (Johnson & Birkland, 2003; Smith & Ingersoll, 2004). Other research
suggested these decisions result in teachers leaving schools that serve high proportions of
racial minority, low-income, and low performing students. This study’s focus on how

teacher candidates deal with limited information suggests that both of these ﬁndings may

149

represent some truth. Teacher candidates in this study may prefer, for example, to work
with a supportive principal that attends to conditions of teaching and learning in the
school. But since they have no reliable way to assess this feature in most schools, they
make assumptions based upon student demographics and achievement information,
which is widely available through the Internet. Further, teacher candidates avoid schools
where they feel they do not “ﬁt” and prefer to teach in schools where they share social
and cultural experiences with the students and staff. As most teacher candidates come
from White, middle class backgrounds, they seek out schools in similar social positions.
African American teacher candidates, many of whom in this study come from low
socioeconomic backgrounds, also seek out schools in a similar social position as
themselves. Thus, teacher candidates from varying socioeconomic backgrounds sort
themselves into schools serving different student populations. If the supply of teacher
candidates from racial minority or low socioeconomic backgrounds does not meet the
demand, some schools will face continual stafﬁng problems.

Consistent with this previous research, teacher candidates’ espoused preferences
include a supportive principal, mentoring, ready availability of instructional materials,
and location. But the way that teacher candidates navigate through the job search process
suggests that other criteria play just as important a role in how teachers come to teach in
particular schools. In short, the attributes that teacher candidates say they want in a
school do not necessarily match the attributes of the school in which they want to work.
These preferences in use include location, familiarity with schools, level of comfort in
schools, and the sense of ﬁt. Comfort and familiarity go hand in hand because teacher

candidates are most comfortable in schools with familiar social and cultural

150

characteristics. The lack of explicit talk about comfort and familiarity as espoused criteria
for evaluating potential schools does not diminish their importance. Teacher candidates
use these features to talk about places where they really want to work and where they
focus their job search efforts. As such, being familiar with a district or feeling
comfortable in a school are powerful ﬁltering agents in teachers’ job searches. Indeed,
location may serve as a proxy for familiarity in existing research as teacher candidates
speak of familiar districts as those that are both geographically and socially proximal to
them. This study underscores the importance of location as an espoused preference that
also operates as a preference in use.

Teacher candidates rely primarily on the Internet and word of mouth for
information both on which schools and districts have available positions and what it
might be like to teach in those schools. Bourdieu (1984) describes how individuals,
through interacting on the ﬁeld, get a sense of their social trajectory. If teacher candidates
have homogeneous social networks, they will only be exposed to schools that are socially
proximal. Despite the fact that Richard and Vanessa attended the same teacher
preparation program, the structural holes (Granovetter, 2005) in their social networks
resulted in them learning about open positions in very different schools. Teacher
candidates identify open positions in districts that are already familiar to them because
they are located in communities where they live or visit. Again, these districts are socially
proximal to teacher candidates. Yet, as this study shows, even when teacher candidates
do learn about job opportunities in unfamiliar districts, the social distance still results in

teacher candidates feeling uncomfortable or out of place and they choose not to apply to

151

these districts based on their habitus. As their habitus inﬂuences how they think they “ﬁt”
with other people, it reduces the impact of these novel sources of information.

That familiarity in and of itself made teaching jobs more attractive to teacher
candidates is not surprising. Mere exposure to an object leads individuals to have positive
evaluations (Zajonc, 1968). Advertising and political campaigns spend millions of dollars
to familiarize us with their products and candidates. In other occupations, prior exposure
or familiarity with a company leads jobseekers to have a positive image of working for
the company and to pursue jobs with the company (Cable & Turban, 2003; Gatewood,
Gowan, & Lautenschlager, 1993). The ﬁndings presented here suggest a broader notion
of familiarity than mere exposure, one that incorporates the degree of social similarity.
Teacher candidates want to work in schools that are familiar to them—both literally and
ﬁguratively. They want to work in schools with which they have some prior experience,
as well as in schools they see as similar to those they attended. Teacher candidates do
learn about opportunities in districts that are otherwise unfamiliar to them—meaning they
are not previously thinking of those districts as possible places of employment—but as
those districts are socially distant they are likely to seem like foreign and unfamiliar
places.

Bourdieu’s cultural reproduction theory is distinct from rational choice theory
because practical logic rejects the idea that individuals act instrumentally. Rather,
individual behavior is spontaneously generated by activating one’s practical logic in a
speciﬁc situation. This framework may appear limited in the context of teachers’ career
decisions because teacher candidates do believe they are acting strategically to secure a

teaching job. But teacher candidates are not consciously aware of the ways their actions

152

narrow their job search. Teacher candidates use explicit strategies to improve their
chances of getting a job offer in districts they most desire, but the features that are related
to which district they most desire are not chosen strategies. Making decisions with a lack
of information or the importance of familiarity in decision—making may be similar to
notions of bounded rationality. Yet the contradiction between teachers’ espoused
preferences and their preferences in use suggest that teachers act in somewhat irrational
ways.

The crowded labor market faced by teacher candidates in this study created
difﬁculties in their job searches. The teacher candidates thought they were conducting
quite broad job searches as they realized available jobs are scarce. Yet as some of the
ways teacher candidates identiﬁed open positions narrowed their search, teacher
candidates say they apply to every position they saw, not realizing that some districts
went unnoticed. Teacher candidates do become more open to schools they initially avoid
as the summer stretches on. Yet while some teacher candidates apply to a broad range of
districts, they focus their efforts in places they most want to teach. This might mean a
difference between submitting an online application on a district website with no follow-
up phone call and personally dropping off resumes at every elementary school in a
district dressed in a suit and calling later to follow-up with the principal. It could mean
the difference between sending a form letter as a cover page or personalizing a cover
letter after researching the school’s mission statement and curricultun. When teacher
candidates focus their job search in particular areas, they also increase the chances they

will be offered jobs in those areas.

153

This chapter illustrates how teacher candidates’ embodied sense of where they ﬁt
guide their decisions during the job search. Teacher candidates seek out places that are
comfortable and familiar. Teachers feel most comfortable in schools that are socially
similar to their own backgrounds as they can relate to students and feel they ﬁt with the
staff. This chapter challenges existing research on teachers’ career decisions that focus on
teachers’ espoused preferences for schools by outlining how teachers’ actions during the
job search are sometimes inconsistent with their espoused preferences. The next chapter
connects the experiences of the teacher candidates in the interview sample to broader
patterns in the teacher labor market. It focuses on where teacher candidates apply and

how their preferences relate to where they ended up working.

154

CHAPTER 5: SORTING AND SELF-SORTING: HOW PREFERENCES AND

HIRING PRACTICES RELATE TO THE DISTRIBUTION OF TEACHERS

The previous chapter illustrated how the practical logic that teacher candidates
brought to the job search led them to reproduce their own schooling experiences by
focusing on preferences in use such as familiarity, comfort, and perceptions of ﬁt. This
chapter explores the overall degree of sorting of teachers among schools. In doing so, this
chapter attempts to connect the individual decisions made by teacher candidates as
portrayed in Chapter 4 with general patterns of teachers’ career decisions evident in
existing literature.

As discussed previously, existing research suggests that teachers are unequally
distributed across schools in ways that disadvantage racial and ethnic minority and low-
income students. This chapter explores the labor market processes that lead to the sorting
of teacher candidates among schools based upon the racial and class composition of the
students in the schools. In particular, this chapter examines the extent to which teacher
candidates sort themselves among schools. Some existing research suggests that teacher
preferences—such as preferences to work close to where they grew up or to teach
students from their own ethnic backgrounds—contribute to the sorting of teachers among
schools (Betts et al., 2000; Boyd et al., 2005a, 2005b; Hanushek et al., 2004). The
decisions teacher candidates make during the job search process can play an active role in
where they end up working. As shown in Chapter 4, the ways in which teacher candidates
identify open positions serve as an initial ﬁltering mechanism for the types of schools

they see as available. Likewise, teacher candidates may decide to apply for jobs in some

155

schools or districts but not in others. Each step in the job search—applications,
interviews, and offers—provides opportunities for teacher candidates to sort themselves
into or out of particular types of schools.

A competing explanation for the inequitable distribution of teachers among
schools focuses on school and district hiring practices in high and low-income areas. This
explanation minimizes the role teachers or teacher behavior may have in teacher sorting
and focuses attention instead on school hiring practices and demand for newly hired
teachers. For example, teachers in low-income schools are more likely to be hired late in
the summer or after the school year started than teachers in hi gh-income schools
(Johnson et al., 2004). Large urban districts—which tend to have poorer students and
higher concentrations of racial minority students—have longer or delayed hiring
processes due to teacher union transfer provisions, late vacancy notiﬁcations, and a
reliance on state budgets that are often not ﬁnalized until the end of the summer (Levin &
Quinn, 2003). As more afﬂuent districts make job offers sooner, they can hire the best
teachers, leaving lower quality teachers for districts that hire at the end of the summer,
such as urban districts. While exploring the extent to which teachers self-sort themselves
into schools, this chapter considers how the timing of job offers received from various
districts affects teacher candidates’ decisions.

It may be that factors other than student characteristics guide teachers’ decisions.
As districts with large concentrations of poor and minority students are often also under-
resourced districts, then it may be teachers’ preferences for high resource districts rather
than preferences for particular types of students that lead to an inequitable distribution

among schools. Teachers may be seeking out higher salaries or better working conditions,

156

such as smaller class sizes, availability of materials, or supportive work environments.
Chapter 4 provided some evidence against this by demonstrating that teachers’
preferences in use focus on feelings of comfort and ﬁt—feelings that are guided by
cultural understandings and perceptions—and that teachers rule out some schools or
districts even though they have little information about the working conditions in those
schools. In exploring how teachers become sorted among schools with varied student
characteristics, this chapter will also examine the resource and working condition
differences among schools.

To examine the degree of sorting in the teacher labor market, this chapter pays
special attention to the extent to which teacher candidates’ social and cultural background
inﬂuences the decisions they make on the job search and how teachers may self-sort into
schools. In doing so, it explores the extent to which the socioeconomic and demographic
characteristics of teacher candidates are related to the types of schools in which they end
up working. I do not assume that teachers of particular class or race backgrounds are
more or less effective teachers. Rather, examining whether teacher demographic
characteristics inﬂuence where teachers end up teaching is important given the difﬁculty
in recruiting individuals from particular backgrounds to enter teaching at all. For
example, Aﬁican American teachers are under represented in the teacher labor force
(Kirby et al., 1999; Kirby & Hudson, 1993). If White teachers and African American
teachers sort themselves into schools that predominantly serve students of their own
racial background, then this creates shortages in predominantly African American
schools. These shortages may force predominantly African American schools to hire

lower quality teachers to compensate for the labor shortage even if the teachers initially

157

attracted to these schools are equally as effective as those attracted to predominantly
White schools.

In addition to differences between teachers based upon their social and cultural
backgrounds, this chapter also explores whether teachers’ verbal ability inﬂuences how
they are sorted across schools. Previous research links teachers’ general ability with
higher student achievement and uses teacher test scores as an indicator of teacher quality
(Ferguson & Ladd, 1996; Greenwald, Hedges, & Laine, 1996). For this reason, I use
teacher candidates’ self-reported ACT score as an indicator of general ability. Teacher
candidates who score in the top quartile of the distribution of ACT-takers in the sample'6
are compared to those who scored in lower quartiles and those who did not take the ACT.
I also explore whether male and female teacher candidates apply to or accept jobs in
different schools. These results are located in Appendix B because, in general, there are
fewer differences by gender. Further, while there is concern about the overall lack of men
among the elementary teaching population, policymakers are less concerned about the
distribution of male and female teachers across elementary schools.

This chapter begins by describing the steps in the job search. The rest of the
chapter explores how teacher candidates’ social and cultural background inﬂuenced how
they navigated these steps and follows their progression through the job search. The
second section examines which teacher candidates removed themselves from the labor
pool and the number of applications submitted by those who stayed in the labor market.
The second section follows teacher candidates’ progression in the job search as they

submitted applications, received offers, and accepted jobs. In doing so, it focuses on the

 

'6 The top quartile of ACT-takers in the sample had a score of 26 or above.
158

student characteristics, employment conditions such as working conditions and salary,
and location of teacher candidates’ job options at each stage to examine how teacher
candidates were distributed among schools. The third section explores the timing of
hiring decisions to see if some schools made job offers later than other schools. The
fourth section examines the relationship between teacher candidate initial preferences and
their labor market outcomes. In doing so, this section explores characteristics of teacher
candidates and their preferences that end up teaching in schools with high concentrations
of poor or racial minority students and urban schools.
Steps in the Job Search

Figure 2 presents a model of the steps in the job search. It also shows the number
of individuals in this study at each stage. For example, 171 respondents completed the
follow-up survey and 90 of these respondents had a teaching job by the end of October
2006. Which teacher candidates ended the job search process with a teaching job and
how they were matched to schools depended on decisions by both teacher candidates and
hiring authorities in schools and districts (Boyd, Lankford, Loeb, & Wyckoff, 2006).
Individuals entered the teacher labor market by deciding to become a teacher and
pursuing teacher certiﬁcation. Some individuals removed themselves from the job market
at the end of their certiﬁcation process but before applying for any jobs. Individuals who
stayed in the teacher job market applied for at least one teaching job and may or may not
have been offered job interviews by a school or district. Individuals who received no
interview offers did not progress further in the job search. Teacher candidates with
interview offers may or may not have been offered teaching jobs. Individuals with a

teaching job offer could then decide to accept one of their offers or could decide to reject

159

all of their offers. Individuals who accepted one of their teaching offers ended up
teaching while those who did not receive any offers or who rejected all of their offers did

not end up teaching.

 

 

it

 

Initial survey, N=289 Follow-up survey, N=171

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Remain in labor Remove self from labor
market, N=161 market, N=10
Offered interviews, No interviews, N=15
N=1 46
Offered at least one No job offer,
job, N=96 N=50
Accept an offer, Reject all job
N=90 offers, N=6
v I
End with a End with no :
job, N=90 job, N=8l

 

 

 

 

 

 

Figure 2 - Steps in the Job Search

This model of the job search highlights the later stages of the teacher pipeline
(Goldhaber & Liu, 2003; Henke, Chen, Geis, & Knepper, 2000). All individuals in the

sample were prepared to teach and have thus already entered the teacher pipeline. By

160

focusing on where teachers applied and where they received job offers, this chapter sheds
light on how teacher candidates handle the later steps in the job search. Teacher
candidates made decisions about where to apply and whether to accept or reject job
offers. Likewise, schools and districts made decisions about whom to interview and
whom to offer jobs. As such, the overall sorting of teachers among schools is inﬂuenced
by both teacher and school or district factors (Boyd et al., 2006).

The model in Figure 2 only considers which teacher candidates ended up getting
any teaching job; it does not account for any potential sorting of teachers between
schools. A more complex model includes such sorting, and this chapter will develop such
a model. For example, in making decisions about whether to remove oneself from the
teacher labor market altogether or to apply for teaching jobs, individuals may decide to
apply for teaching jobs in only particular types of schools or to apply for jobs across a
wide range of schools. Teacher candidates’ preferences for particular teaching jobs may
inﬂuence both where they apply and which jobs they accept once offers are made.
Likewise, teacher candidates may get job offers only from a particular type of school or
from a range of school types. The range of job offers teacher candidates received depends
upon the range of jobs to which they applied, but it may also depend on differences in the
hiring decisions of schools.

Number of Applications

This section explores the number of districts to which teacher candidates applied.

Table 22 presents the means and standard deviations for the variables used in this

chapter. It shows descriptive statistics for all respondents to the follow-up survey as well

161

as the means and standard deviation for only those teacher candidates who remained in

the job market because most of the analysis is conducted using only these respondents.

Table 22 - Descriptive Statistics for Chapter 5

 

Teacher candidates

 

 

 

All teacher who stayed in the

candidates job market
Variable N Mean SD N Mean SD
Stayed in labor market 171 0.942 0.235 161 - -
African American 169 0.065 0.247 159 0.063 0.244
Other racial minority 169 0.112 0.317 159 0.113 0.318
Male 169 0.083 0.276 159 0.075 0.265
Mid-career changer 169 0.349 0.478 159 0.346 0.477
Mother's educationl 169 2.12 1.03 159 2.14 1.04
ACT score 171 24.70 3.19 161 24.65 3.17
ACT score missing 171 0.322 0.468 161 0.317 0.467
Named an urban or low-income
suburban school as top choice school 151 0.185 0.390 141 0.199 0.400
Named having many racial-ethnic
minority students as one of most
important characteristics 171 0.088 0.284 161 0.087 0.283
Was offered at least one teaching job 171 0.561 0.498 161 0.596 0.492
Accepted job offer 96 0.938 0.243 96 0.938 0.243
Month of ﬁrst job offer2 85 6.54 1.41 85 6.54 1.41

 

1 Mother’s education is coded as 1=high school or less; 2=some college; 3=college

graduate; 4=master’s degree or above

2 Month of ﬁrst job offer uses numerical codes for each month (i.e., 5=received ﬁrst

job offer in May; 6=June)

Table 23 shows which potential teachers removed themselves from the teacher
labor market prior to applying for any teaching job and the number of districts to which
teacher candidates applied. Overall, 5.9% of potential teachers did not apply for any
teaching job, thereby removing themselves from the teacher job market. There was some
evidence that individuals who did not take the ACT were more likely to remove

themselves from the teacher job market while those who had mothers with at least a

162

master’s degree were more likely to stay in the labor market, but no difference was
statistically signiﬁcant. Thus, there did not appear to be strong relationships between
potential teachers’ social and cultural background and their decision to remain in the

teacher labor market.

Table 23 - Percent of Potential Teachers Who Removed Themselves From the Teacher

Labor Market and the Number of Applications

 

 

Number of applications
Removed self One to Eleven Twenty- More

Variable from market 10 to 20 one to 40 than 40 N
All 5.9% 41.0% 15.5% 14.3% 29.2% 161
White 5.8 37.4 14.5 15.3 32.8 131
Aﬁican American 9.1 70.0 20.0 10.0 0.0 10
Other minority 5.3 50.0 22.2 11.1 16.7 18
New college graduate 5.5 33.7*** 13.5*** 13.5*** 39.4*** 104
Mid-career changer 6.8 54.6*** 20.0*** 16.4*** 9.1*** 55
ACT score
Top ACT quartile 6.3 33.3 22.2 15.6 28.9 45
Lower ACT quartiles 3.0~ 37.1 15.5 13.4 34.0 97
Did not take ACT 17.4* 79.0** 0.0** 15.8** 5.3** 19
Mother’s education
High school or below 7.7 45.0 13.3 15.0 26.7 60
Some college 5.7 36.4 18.2 12.1 33.3 33
Bachelor’s degree 5.7 42.0 14.0 20.0 24.0 50
Master’s or above 0.0 31.3 18.8 0.0 50.0 16

 

~p<.1, * P<-059 an: P<-01, =1"ka p<.001

There were differences in the number of districts to which teacher candidates
applied, however. Forty-one percent of teacher candidates applied to ten or fewer

districts, while 15.5% applied to 11 to 20 districts. This is notable given the labor market

163

conditions in the Smithton area of excess supply of teachers. Just less than 30% of
teacher candidates applied to more than 40 districts. There is some evidence that African
American teacher candidates applied to fewer districts, but this difference was not
statistically signiﬁcant, possibly due to the small sample size among African American
teacher candidates. When all teacher candidates who applied to more than 10 jobs were
grouped together, African American teacher candidates were signiﬁcantly less likely to
apply to more than 10 districts. Teacher candidates who were just graduating from
college applied to more districts than those who were switching into teaching from
another career. There was no evidence that teacher candidates with higher ACT scores
applied to fewer districts than those with lower ACT scores, although teacher candidates
who did not take the ACT applied to fewer districts. As those who did not take the ACT
also tended to be mid-career changers, it is possible the difference in application rates
were driven by the ﬁndings for mid-career changers.
Progression in the Job Search and the Distribution of Teachers

There were also differences in the characteristics of districts to which teacher
candidates applied. This section presents the student characteristics, employment
conditions, and location of the districts to which teacher candidates applied, received
offers, and accepted jobs. By tracing the patterns from where teacher candidates applied
to where they received job offers to where they ultimately ended up working, we can
begin to see the relative impact of teacher candidate behavior or district hiring practices
on teacher sorting. Differences in where teacher candidates applied are most likely due to
teacher candidate behaviors, thus we could describe it as teacher candidates self-sorting

into schools. Yet if teacher candidates applied to districts with similar characteristics, but

164

received offers from districts with varying characteristics in disproportionate rates, then
that suggests that school and district hiring practices were sorting teachers among
schools. Teacher candidate behavior—and possible self-sorting—becomes more
important again in differences between where they had job offers and where they
accepted jobs.

Student Characteristics

Table 24 presents the percentage of students eligible for free or reduced-price
lunch (F RL) where teacher candidates applied, received offers, and ended up working.
African American teacher candidates applied to districts that had higher percentages of
F RL students. The average district to which African American teacher candidates applied
had 47.3% FRL students, compared to 24.6% for White teacher candidates and 33.2% for
teacher candidates from other racial minority backgrounds. There was some evidence that
Aﬁican American teacher candidates were more likely to apply to low-income suburban
districts.

Both White and African American teacher candidates bad job offers from districts
with higher concentrations of low—income students than the average of those to which
they applied. This suggests that districts with more FRL students made more job offers.
Nevertheless, there were still large differences in the percentage of FRL students in
districts where African American teacher candidates had offers and those districts where
White and other racial minority teacher candidates had offers. The magnitude of the
difference was similar to the magnitude of the difference to where they applied. Thus,
there was no evidence that district hiring practices contributed to sorting teachers based

upon race into districts with varying student socioeconomic backgrounds. This difference

165

was most likely the result of where teacher candidates applied. African American teacher
candidates ended up teaching in districts that had an average of 56.9% FRL students. This
was somewhat higher than White and other minority teacher candidates. Also, White and
African American teacher candidates received offers and accepted jobs in schools that
had similar characteristics to the overall district averages, but teacher candidates from
other racial minority backgrounds received offers from schools with more FRL students
than the district average.

There were few differences in where teacher candidates applied when looking at
teacher candidate career status. Mid-career changers and new college graduates applied to
districts with similar percentages of F RL students, however, mid-career changers
received offers from districts with greater proportions of FRL students. Mid-career
changers had job offers from schools with an average of 61 .8% FRL students, compared
to new college graduates who had offers from schools with 33.0% FRL students. Mid-
career changers also ended up working in schools with greater concentrations of F RL
students compared to their peers. Mid-career changers ended up working in schools with
59.9% FRL students, compared to new college graduates who taught in schools with
31.1% FRL students. This difference was likely due to school and district preferences.
There were no differences in the average student socioeconomic characteristics of the
districts where teacher candidates of different career status applied, although there were
differences in where they had offers and where they ended up teaching. This indicates
that school and district hiring practices contributed substantially to the sorting of mid—

career and new college graduates into different schools.

166

Save r... .58 r. wove ._. .3.... 2
.Q engage. E N< 053. com docs—88 03 959838 38m 030 b Bonﬁre 8a 038 £5 E 33mm: 05. “802

 

 

 

 

167

 

ow K on mm Go of >2 2
1.1.2 3...”: aims— **N.o~ mom ...m.N~ mom 26% .8 @5532
ode Nﬁm ﬂee o.wm mﬁm ...o.eN o.me oopwou mac—050mm
mdm QwN m.mm w.wm men edm o.om 03:8 oEom
...oem ...oSe *m.mm ...ooe eon .Imde 203 323 Ho 30:8 :32
233.6%» a. Loos:
2m. 5 Now ...edo NR o.wm mom Imom HU< 8:3 Ho: ED
m.wm mem _.oe m.mm mom mom 25. S 8:55: 8304
Nem mgm oem m.mm o.mm :m oem 05.83 ooh
mucousbw
fluodm .3.. _ .ov ***w.~© ...ode mﬁm Z©.me eom gown—MAO 50.80-22
$1.: team 2.5.? twom QR 4.8 3% saga» hum8:8 >$2
ﬁne oﬁm mom mom Nmm wﬁm o.om .3858 350
0.0m Zmom mom 2m.mm ..:_.._.m.he 2o.oo o.oe 538:3. §QE<
mom mem Ohm o.mm oem mem mom BE?
odm o.mm woe odm mum o\om.mm .x. e.mm =<
323m 3th 383m. 3.22.5 henna? zthﬁm. new: toiaﬁm. Lam: ﬂoats“
oothmwooe. ENS 33.88% 3.235. 38.63 ELSE 2:85
5E Becca :82 is 8 Emma? «m3 8 33%

 

32. 3383‘ new .mcobo

vozooom “@2394 8822.80 Enough 22:5 3&5 50:54 02568:qu no can So o_£w=m magnum mo owﬂcoobm - em 03¢?

There were differences in where teacher candidates applied based upon the
teacher candidate’s socioeconomic status. While there was no evidence that mother’s
education was related to the average percentage of F RL students in the districts in which
teacher candidates applied, there was a negative relationship between a teacher
candidate’s mother’s education and the likelihood they applied to public school districts
in low-income suburbs around Smithton. In other words, teacher candidates with more
highly educated mothers were less likely to apply to low-income suburban districts
around Smithton. On the other hand, teacher candidates whose mother never attended
college were more likely to apply to low-income suburban districts. Teacher candidates
whose mother never attended college also received job offers from districts with greater
percentages of FRL students than those with more educated mothers. They also accepted
jobs in districts with greater concentrations of FRL students. Teacher candidates whose
mother had a master’s degree or above accepted jobs in districts with fewer F RL
students. The sorting of teacher candidates from varied socioeconomic backgrounds into
districts with different compositions of students by socioeconomic background seems to
be due to both teacher candidate behaviors and hiring practices. There was some evidence
of sorting in where teacher candidates applied, more sorting in where they received
offers, and even more sorting in where they accepted jobs.

Table 25 shows similar patterns in the percentage of minority students in districts
where teacher candidates applied, received offers, and accepted jobs. African American
teacher candidates applied to districts with higher percentages of minority students than
White teacher candidates or those from other racial minority backgrounds. Similar to the

ﬁndings for percentage of FRL students, both White and African American teacher

168

candidates received offers and accepted jobs from districts with more minority students
than the average district to which they applied. But the large difference in student racial
composition between White and African American teacher candidate remained. This
suggests that the sorting of teachers into schools with different racial compositions was
due primarily to differences in where they applied.

The initial survey included measures of teacher candidate espoused preferences.
In particular, it asked teacher candidates to choose which school features were the three
most important factors in their job search decisions. It was expected that these
preferences inﬂuenced where teacher candidates applied and accepted jobs. Teacher
candidates who said working in a school with many racial minority students was one of
the three most important factors in their job search applied to districts with greater
proportions of minority students than those who did not list this as one of the most
important factors. Thus, teacher candidates did seem to be acting on this espoused
preference in where they applied. However, for teachers who preferred to teach in
schools with many racial minority students, there was no difference in student
composition in where these teacher candidates received offers or ultimately ended up

working than in teachers without this preference.

169

Table 25 - Percentage of Minority Students Where Teacher Candidates Applied,

Received Offers, and Accepted Jobs

 

 

 

Received offer Accepted job

Variable Applied District School District School
All 30.8 47.1 43.3 43.3 41.2
White 27.3 45.1 39.8 41.9 38.3
African American 62.0*** 73.3* 77.5** 68.2~ 79.2"‘
Other minority 38.4 39.2 43.2 37.5 41.7
New college graduate 30.8 39.3*** 32.9*** 36.1” 31.5***
Mid-career changer 30.9 66.9*** 73.2*** 61 .5** 68.6***
ACT score
Top quartile 29.3 44.7 39.9 41.7 39.7
Lower quartiles 31.5 46.6 43.6 44.0 41.1
Did not take ACT 30.6 56.3 54.6 44.4 48.1
Mother’s education
High school or below 28.4 48.6 45.1 44.0 44.3
Some college 26.3 39.3 33.9 34.1 30.4
Bachelor’s degree 34.0 54.0 50.5 51.9~ 49.5~
Master’s or above 37.2 38.9 34.1 33.9 29.3~
Many ram“ ‘9'“th Students 28.7*** 46.2 42.9 42.4 40.9
are not most important
Many ”9‘31 mm‘m‘y Students 51.6*** 54.7 46.4 50.7 43.7
are most important
N 153 81 85 71 84

 

~ p<.1, * p<.05, ** p<.01, *** p<.001

Employment Conditions

Employment conditions such as salary and working conditions can refer to many

different aspects of work-life in a school. Unfortunately, measures of many of these

working conditions are not routinely collected. This section focuses on two speciﬁc types

of working conditions—pupil-teacher ratios and instructional expenditures—and on

salary. Table 26 presents the average pupil-teacher ratios where teacher candidates

170

applied, received offers, and accepted jobs. Overall, there were few differences in pupil-
teacher ratios and no evidence that teacher candidates were sorted among schools with
larger or smaller pupil-teacher ratios. One explanation for teacher sorting by student race
and socioeconomic background is that teachers sought out schools with better working
conditions. Yet these data suggest teacher candidates were sorted across schools with
varying student characteristics but not among schools with varying pupil-teacher ratios.
Thus if this sorting is due to working conditions, then it is a working condition other than
pupil-teacher ratios. Further, teacher candidates who said having a small class size was
one of the most important factors in their job search did not apply to or end up working in
districts with statistically signiﬁcantly lower pupil-teacher ratios. While the pupil-teacher
ratio is not an exact measure of class size, it is an indicator available to teacher candidates

on the Internet.'7

 

'7 In the interviews, many teacher candidates said they used www.greatschools.net to gather information
about schools. This website has pupil-teacher ratios.

171

Table 26 - Mean Pupil-Teacher Ratios Where Teacher Candidates Applied, Received

 

 

 

 

Offers and Accepted Jobs
Received offer Accepted job

Variable Applied District School District School
All 17.3 18.1 18.1 17.7 18.2
White 17.2 17.8 18.1 17.6 18.1
African American 17.6 19.5 17.0 17.5 17.1
Other minority 17.5 18.9 18.9 18.7 19.2
New college graduate 17.2 17.7 17.7 17.6 17.9
Mid-career changer 17.3 19.1 19.3 17.9 18.8
ACT score

Top quartile 17.7 18.0 18.6 17.9 18.6
Lower quartiles 17.0 18.2 17.7 17.6 17.8
Did not take ACT 17.3 17.5 19.3 17.6 19.2
Mother’s education

High school or below 17.4 19.2 18.2 18.3 17.7
Some college 17.1 18.1 18.2 18.3 18.7
Bachelor’s degree 17.5 18.3 18.7 17.8 18.6
Master’s or above 16.7 15.8* 16.7 16.2* 17.3
Class size is not most important 17.3 18.1 17.9 17.7 18.0
Class size is most important 16.8 18.0 19.1 17.9 19.2
N 146 68 74 64 70

 

* p<.05

Similar to the ﬁndings for pupil-teacher ratios, there were few differences in the
per-pupil instructional expenditures among teacher candidates in where they applied,
received offers, and accepted jobs (see Table 27). Teacher candidates whose mother had
at least a master’s degree applied to districts with higher per-pupil instructional
expenditures, suggesting they applied to districts with greater resources. They were also
more likely to receive offers and accept jobs in districts with higher instructional

expenditures, but the difference was present in where they initially applied. Teacher

172

candidates whose mother only had a high school education received offers and accepted

jobs in districts with lower instructional expenditures. As they applied to districts with

similar expenditures as their peers from somewhat higher socioeconomic backgrounds,

this difference could be due to district hiring practices.

Table 27 - Mean Per-Pupil Instructional Expenditures Where Teacher Candidates

Applied, Received Offers, and Accepted Jobs

 

Received Accepted

 

 

Variable Applied oﬂer job
All $5,163.14 $4,795.75 $4,906.14
White 5,181.10 4,785.10 4,873.17
African American 5,237.71 4,712.95 4,654.28
Other minority 5,013.36 4,928.04 5,255.12
New college graduate 5,213.08 4,884.63 5,028.23
Mid-career changer 5,072.94 4,575.49 4,586.39
ACT score
Top quartile 5258.39 4635.78 4,716.42
Lower quartiles 5104.80 4875.20 4,979.91
Did not take ACT 5232.35 4834.05 5,102.51
Mother ’s education
High school or below 5,072.61 4,391 .61~ 4,410.32*
Some college 5,007.17 4,472.38 4,592.93
Bachelor’s degree 5,131.21 4,932.85 5,010.57
Master’s or above 5,990.30~ 5,700.14~ 5,854.19*
Availability of materials 18 5,147.00 4,842.1 4,912.25
not most important
Ava‘l?b‘1‘ty Of mammals ‘5 5,210.31 4,594.8 4,883.25
most important
N 153 80 76
~ p<.1, * p<.05

173

Teacher candidates who said that the availability of textbooks and other
instructional materials was among the most important factors in their job search did not
apply to or end up working in districts with higher instructional expenditures.
Instructional materials are categorized as instructional expenditures and thus higher per-
pupil instructional expenditures should result in greater resource availability. The lack of
difference in these ﬁndings may be due to the per-pupil instructional expenditures being
too broad of a category to capture variation in instructional material purchases or to
teacher candidates not acting on these espoused preferences in where they applied or
accepted offers. Overall, the two measures of working conditions—pupil-teacher ratios
and instructional expenditures—were not related to the distribution of teachers across
schools. In particular, teachers with espoused preferences for these working conditions
were not more likely to apply for or accept jobs with these working conditions.

Table 28 presents the average salaries in districts where teacher candidates
applied, where they had offers, and where they accepted a job. Salary ﬁgures for the
districts in which teacher candidates applied and received offers were drawn from salaries
for beginning teachers in databases across several states.18 Teacher candidates provided
the salary for the job they accepted on the follow-up survey. Overall, teacher candidates
who ended up with a teaching job earned an average salary of $34,600. The average
salary teacher candidates ended up accepting did not vary substantially by teacher

candidate characteristics.

 

'8 See Chapter 2 for more information on the source of this data.

174

Table 28 - Mean Salary Where Teacher Candidates Applied, Received Offers, and

 

 

Accepted Jobs
Variable Applied Received oﬂer AcceptedL'ob

All $36,565.01 $36,510.36 $34,562.48
White 36,520.65 36,318.28 34,443.49
African American 37,159.71 38,303.64 33,511.75
Other minority 36,622.49 36,253.95 35,965.38
New college graduate 36,679.48 36,773.80 34,645.17
Mid-career changer 36,397.12 36,014.49 34,299.82
ACT score
Top quartile 36,622.23 36,334.06 34,355.29
Lower quartiles 36,466.50 36,396.72 34,597.05
Did not take ACT 36,884.82 37,685.88 35,316.17
Mother ’5 education
High school or below 36,403.86 35,207.44 33,565.52
Some college 35,766.58~ 35,559.33 34,182.33
Bachelor’s degree 36,803.61 37,078.91 35,439.68
Master’s or above 38,383.03* 39,012.84* 35,192.56
Salary is not most important 36,471.75 36,399.01 34,515.58
Salary is most important 37,223.34 37,178.51 35,070.50
N 137 49 71

 

Note: The salary where teacher candidates accepted jobs is not directly comparable to
where they applied and where they received offers. The average salaries for where they
applied and received offers come from district averages of ﬁrst-year teacher salaries. The
salary for where they accepted jobs comes from the follow-up survey.

 

~ p<.1, * p<.05

There was also little variation in the average salaries of the districts to which
teacher candidates applied and received offers. Teacher candidates whose mother had
some college education applied to districts with slightly lower salaries and teacher

candidates whose mother had at least a master’s degree applied to districts with higher

175

salaries. Teacher candidates whose mother had a master’s degree also received job offers
from higher paying districts. Yet teacher candidates’ espoused preferences for salary was
not related to where they applied, received offers, or accepted jobs. Teacher candidates
who rated a high salary as one of their most important features in the job search applied
to districts with similar salaries as those who did not rate a high salary as most important.
Location

Table 29 shows the percentage of teacher candidates who applied to, received
offers from, and accepted jobs in urban schools. African American teacher candidates
were more likely end up working in an urban school. About 66.7% of Aﬁ’ican American
teacher candidates accepted jobs in urban schools, compared to 33.8% and 27.3% of
White and other racial minority teacher candidates, respectively. This difference was not
surprising as African American teacher candidates also applied to urban districts in
greater numbers. Aﬁican American teacher candidates were also more likely than other
teacher candidates to apply to Smithton Public Schools.

Teacher candidates whose mother had at least a master’s degree were the most
likely to apply to urban districts. This appears to be due to the fact that these teacher
candidates applied to urban schools other than Smithton Public Schools. Teacher
candidates with highly educated mothers received job offers and accepted jobs from
urban districts at similar rates as other teacher candidates whose mother never went to
college or had a bachelor’s degree. Teacher candidates whose mother hadsome college
education were less likely to have job offers from urban schools. Mid-career changers
were more likely than new college graduates to have job offers from and end up working

in urban districts. As there were no differences in the districts to which mid-career

176

changers and new college graduates applied, this difference could due to urban districts

not making offers to new college graduates.

Table 29 - Percentage of Teacher Candidates Who Applied, Received Offers, and

Accepted a Job in Urban Schools

 

Any urban school

 

 

Applied Smithton Received Accepted

Variable Public Schools Applied oﬂers job
All 14.5% 65.4% 40.0% 35.3%
White 10.1 63.2 37.3 33.8
African American 600*“ 90.0~ 71 .4~ 66.7~
Other minority 16.7 70.6 36.4 27.3
New college graduate 13.7 63.4 28.6*** 25.4"
Mid-career changer 14.6 70.6 72.7*** 63.6”
ACT score
Top quartile 8.9 65.9 30.0 30.0
Lower quartiles 16.8 64.1 44.7 37.5
Did not take ACT 15.8 70.6 50.0 42.9
Mother’s education
High school or below 220* 63.6 42.9 38.1
Some college 12.1 54.6 22.2~ 22.2
Bachelor’s degree 12.0 67.4 45.5 39.4
Master’s or above 0.0~ 86.7~ 41.7 38.5
N 159 153 85 85

 

~ p<_], * p<_05, ** p<.01, *** p<001

Table 30 shows the average distance from the teacher candidate’s home at the
beginning of the job search to where they applied, where they had offers, and where they

ended up working. 19 It also shows the percentage of teacher candidates that applied for

 

'9 1 used the location of their home at the beginning of the job search rather than the location of their high
school due to missing values on high school location. Some respondents did not provide information on

177

jobs in other states. African American teacher candidates were less likely to apply to
districts in other states, but there were no statistically signiﬁcant differences between
teacher candidates from different racial backgrounds in the average distance from their
home to the districts in which they applied or ended up working. New college graduates
were more likely than mid-career changers to apply to districts in other states. The
average district to which new college graduates applied was 286 miles from where they
lived at the beginning of the job search, compared to 97 miles for mid-career changers.
Teacher candidates who indicated on the initial survey that proximity to their home was
one of the most important factors in their job search decision did not apply to districts
that were closer on average than teacher candidates who did not say proximity to home

was one of the most important factors.

 

their high school. I ran similar analyses for the distance from where the teacher candidates’ graduated from
high school to where they applied, received jobs, and ended up working. The results were similar for new
college graduates, but different for mid-career changers. Many of the mid-career changers in this study
grew up in other states and moved to the Smithton area as adults.

178

 

 

d9?“ 28x EckmotcEQ

8:8: Eoﬁmotsﬂh S83

hmuﬁxh kmﬁub Sm h~©0£bm

Nde edo mdm down Ndee osem mde hge Edd 96% 8 38882
w.me e.o ode o.mom ﬁoom _.moN e.om e.om ode cameo $88335
ddm odd ”Hm zmed Imodd o.owm zddm Iddm mde owozoo 28m
:93 iode .Ioom 05m wdmm ﬁmw— {1.3 ...NNN ..iodm 308d 8 828m awn;
8.2853 ”12:83.
03 mew e. Z. odew as? no: NNN edm mdm HU< 8—8 8: ED
o.om w.w_ m. g m ﬁmﬁe e.o~e ddem “.me ﬁme N.me 8:88 .8264
osm ﬁem oﬁm wdmm mSoe ﬂog m.mm m.mm mde 8:88 go...
8.8%. .NDV
w.mm ddm doe ndmm ﬁemm imdo odm ddm item 8w§no 88.80-22
Noe Q: m.mm dooe edde ......e.dwm ﬁme ode .........o.mm 88388 owozoo 382
o.oe o.o~ o.om oomm d. ﬁmd Esme o.o.n.. o.om e.ee 8:088 8:5
m.mm m.mm m.mm moon edmm m.mo~ ddm odm ...o.o_ 808:2 805484
ﬂee dom mdm domm mdom ﬂood odm odm wee 833
ode mom odm e.odm 3 ﬁe ddmm .xcodm AXLoM Rage :<
823 83% .823. 8.x @8ch 382%? do.\ 98%.» 82%? 833.8;
885 .8 o3 oo Tn N h To nmﬁmoow ~838an ~328ro 338.8%

 

32. was .maoto $882894 8o 8:8: 88m

393me owm8>< 98 88% 880 E 32. “8883‘ one .m8t0 83.88% 882%? on? 882280 8:289 .8 owﬁaoeom - on 838,—.

179

_oo.VQ .12.. Jova .3.. .mo.VQ _._ JVQ I

d 5888 a a... 833. 8m

888% 8:8 88.: 88288 88 38888. 88m 80 t: :88b8 88 2:8 m8 5 £38 8:,w 888 85 .8 :8 E8 888 85 8 88:8
:8: 82:2: :3 8 8:8: :8: 82888 m=o8> :85 888888 8:88 .8 8828885: 85 88 8:8: 28o 8888 888 8.: 882

 

 

 

 

ew em do :2 oo No mm : Z

. . . . . . . . . 8:88 88

m mm m mm e ee N. new d ndm e men d :m m dm e me 2 8:8: 9 ©8885

. . . . . . . . . 8:88 82: a:

m we N. o: o em d Noe o oee : 2m o oe m we d :e 8 8:8: 9 ~08:on
.828 823 82: 8.5 Soho hmzaaw 8\ 8888 nmzmﬁw 8388;

8.85 .8 o2 oo Th N m. To 838ro 838.8% hmimoow 838.88%
8?: 8:8: 5888:8me 828: $888888 :8: 888 8:8 5 8.888%

 

.3883 on 838

180

The preceding analysis discussed the average characteristics of jobs to which
teacher candidates applied and presented some evidence that teacher candidates from
different racial and class backgrounds applied to different districts. Yet teacher
candidates could have applied to several districts with varied characteristics. Another
strategy to examine the extent to which teacher candidates self-sorted themselves among
districts at the application phase involves looking at the characteristics of teacher
candidates that applied to different districts. To answer this question, I used the
restructured dataset to look at patterns in the districts to which teacher candidates applied.

Table 31 presents correlations between teacher candidate characteristics and the
districts to which they applied. There was a positive correlation between being African
American, the percentage of minority and FRL students in the district where they applied,
and applying to an urban district. While the correlations are statistically different from
zero, they are not that large. Likewise, there were positive and statistically signiﬁcant but
small relationships between coming from another racial minority background and the
percentage of minority and FRL students in the district. There were negative relationships
between being White and the percentage of minority or FRL students in the district where
they applied. There was a small negative relationship between coming from another
racial minority background and the per-pupil instructional expenditures in the district.
This suggests that non-White and non-African American teacher candidates applied to
districts with fewer resources. Teacher candidates’ socioeconomic background was also
related to district characteristics. As shown in Table 27 as well, there was a small positive
relationship between the teacher candidate’s mother’s educational level and the per-pupil

instructional expenditures in the district. There was a small negative relationship between

181

the teacher candidate’s ACT score and the percentage of F RL students in the district
where they applied, indicating that teacher candidates with higher ACT scores applied to

districts with fewer low-income students.

182

.Sovg ... .sva .. .movg . ...va 2

 

 

286 So- 3o 36 ...omd ...Ed- ...Sd 3o ...58 ...de 935

593%

2: .86- ...86. 85 ...25 25- mod- :3 So So 328-22

2: ...ad 3o- ...85- ...ed 8.0- So .86- 8.? 28.5...

Gomumozﬁo

2: 85 ...So- ..:.o ..:.o 3o- 86. mod Maﬁa:

£858

2: .85- ...E- 28.0- So .85 .23 3085.3

gotog

2: .18.? 85 :5 ...Rd ...Rd 98.5.

2: 8.0 So. ...de- ...mmd- 32>»

mogmwaoaxo

3:030ng

21 ...xd- ...omd- 3o Egg-am

Oman-n

of So- 85. sage-Ea

3:03am

2: ...Kd 5552...

$538

$858

2: 28......

xmngu 958. tote-oak» b.2252 tautmsw 2.5x: wmxzhhnmm-xm 925 2233::- Szmhzt. Saute-A
gum-Eu 50v. MESQE. 385 58.3%. Notewozhnﬁ x332 5E .8335
.3: ESQ mug-km 4.3-am ~28ka Emu-3m

 

now—n54.- »oﬁ- £033 8 $2.35 05 new 83338350 836ch cocoa”? Begum 5322.80 - R 033.

183

Timing of Job Offers

As noted above, previous research found that large urban districts and high-
poverty districts lose out on high quality teachers because they hire later than more
afﬂuent districts (Levin & Quinn, 2003). For this reason, it is necessary to consider when
teacher candidates received their job offers and whether or not they received multiple
offers. Table 32 shows the percentage distribution of teacher candidates according the
month in which they received their earliest job offer.20 Overall, 68.2% received their ﬁrst
job offer before the start of August. This suggests most teacher candidates who received a
job offer did not need to decide at the last minute. Mid-career changers were more likely
than new college graduates to receive their job offer in August or later. Half of mid-
career changers received their ﬁrst job offer in August while half of new college
graduates received their ﬁrst job offer in June or July. There was no evidence that African
American teacher candidates received job offers later than other teacher candidates.
There was limited evidence that teacher candidates who did not take the ACT received
their ﬁrst job offer later than other teacher candidates did, but the difference was not
statistically signiﬁcant. The sample size may limit the detection of small or moderate

effects.

 

20 Only two teacher candidates in the sample received job offers aﬁer the start of the school year. The
timing of the follow-up survey in late September to October may have limited the identiﬁcation of teacher
candidates who received job offers after the school year started.

184

Table 32 - Month in Which Teacher Candidates Received Their First Job Offer

 

 

Variable May or earlier June or July August or later N
All 21.2% 47.1% 31.8% 85
White 19.7 47.0 33.3 66
African American 25.0 50.0 25.0 8
Other minority 27.3 45.5 27.3 11
New college graduate 21 .3~ 54.1~ 24.6~ 61
Mid-career changer 20.8~ 29.2~ 50.0~ 24
ACT score
Top quartile 18.5 55.6 25.9 27
Lower quartiles 27.1 41.7 31.3 48
Did not take ACT 0.0 50.0 50.0 10
Mother’s education
High school or below 11.5 50.0 38.5 26
Some college 31.3 43.8 25.0 16
Bachelor’s degree 21.9 46.9 31.3 32
Master’s or above 30.0 40.0 30.0 10

 

~p<.1

Table 33 shows the correlation between school characteristics and the month in
which a job was offered. Most school characteristics were not associated with the month
in which a job offer was made. The percentage of FRL or minority students was not
related to the timing of the job offer. Neither were the pupil-teacher ratios, the
instructional expenditures, or the beginning teacher salary in the district. There was a
positive relationship between the month the job offer was made and whether the school
was an urban school. Urban schools tended to make job offers in later months than non-

urban schools.

185

Table 33 - Correlations Between School Characteristics of Job Offers and Month Offer

Was Received

 

 

Variable Month of oﬂer
Percent minority 0.08
Percent F RL 0.05
Pupil-teacher ratio 0.17~
Per-pupil instructional expenditures 0.13
Beginning teacher salary in district -0.01
Urban school 034* * *

 

~ p<.1, *** p<.001.
Preferences and Labor Market Outcomes

The previous sections showed there was a sorting of teacher candidates by race
and class among schools based upon the percentage minority and FRL students in the
school, yet it is still not clear how teacher candidates’ preferences contributed to this
sorting. Did the schools in which teacher candidates end up teaching reﬂect their initial
preferences about where they wanted to teach? Tables 25 to 30 showed that teacher
candidates who named a particular feature—many minority students, class size,
availability of materials, and proximity to home—as one of the most important features in
their job search did not end up working in schools with these features. For example,
teacher candidates who said working in a school with many racial minority students was
an important feature did not end up working in schools with more minority students than
those teacher candidates who did not say the racial composition of students was an
important factor (although they did apply to districts with greater percentages of minority
students). Likewise, teacher candidates who said class size or availability of instructional
materials were one of the most important features in their job search did not end up

working in schools with lower pupil-teacher ratios or higher instructional expenditures.

186

Chapter 4 introduced the idea of preferences in use and demonstrated how teacher
candidates’ preferences in use might not be consistent with their espoused preferences.
That is, the characteristics of schools that teacher candidates preferred might not reﬂect
teacher candidates’ stated preferences about what characteristics were important in their
job search. To examine the impact of teacher candidate preferences, we must also explore
whether teacher candidates were successful in realizing their preferences in use. Table 34
shows correlations between characteristics of the schools teacher candidates named as
their most desired places to teach in the initial survey and the schools in which they
ended up teaching. There were moderate and statistically signiﬁcant positive
relationships between the percentage of minority and FRL students in the schools that
teacher candidates named as their most desired job and those where they ended up
working. There was also a positive relationship between the instructional expenditures of
teacher candidates’ most desired schools and where they ended up teaching. There was
no relationship between the pupil-teacher ratio or beginning teacher salary in teacher
candidates’ preferences in use and their ultimate job outcome. This notion of preferences
in use may seem inconsistent with Chapter 4’s focus on familiarity, comfort, and ﬁtting
in. Conceptually they are similar as they both highlight attributes of the schools teacher

candidates preferred rather than the school characteristics they said they preferred.

187

Table 34 - Correlations Between Characteristics of Schools Where Teaching and Initial

 

 

Top Choice School
Initial top choice
school
School where teaching

Percent minority in top choice school 030’”
Percent F RL in top choice school 026*
Pupil-teacher ratio in top choice school 0.09
Per-pupil instructional expenditures in top choice school 0.3 l **
Beginning teacher salary in district .01

 

* p<.05, ** p<.01.

To examine more closely the factors associated with teaching in schools with
greater numbers of minority or F RL students or in urban schools, I ran a series of
regression analyses using the school characteristic as the dependent variable. These
school characteristics are often associated with hard-to-staff schools, thus it is useful to
know the factors that are related to who ended up working in these schools. The
estimated regression model is:

Y- = [30 + TEACHCHARR‘BI + ESPOUSED-*0; + INUSE-*B3 +

TIMINGi ”114+ a;

(1)

where Y, is the percentage of minority students in the school where the teacher ended up
working. TEA CHCHARi is a vector of teacher characteristics. As shown earlier in this
chapter, teacher candidate characteristics were related to the characteristics of the schools
in which they ended up teaching. In addition to these teacher candidate characteristics,
teacher candidates’ espoused preferences, their preferences in use, and timing of the job

search may also be related to the concentration of minority students in the school where

188

the teacher candidate ended up working. ESPOUSED; is a dummy variable indicating if
the teacher said working with many minority students was one of the most important
features in their job search. Stated or espoused preferences for a particular school
characteristic should be related to whether teacher candidates ended up in schools with
that characteristic. INUSE, is the percentage of minority students in the school the teacher
candidate named as his or her most desired school at the beginning of the job search.
Teacher candidate preferences in use may also inﬂuence where they ended up working. If
teacher candidates started their job search wanting to work in a high minority school,
then they may be more likely to have ended up in a high minority school. T IMINGi is a
vector of dummy variables indicating the month in which the teacher candidate received
his or her ﬁrst offer. The timing of the job offer may also inﬂuence the school racial
composition if some schools are more likely to make early hiring decisions.

The model for factors that are associated with the percentage of F RL students in
the school the teacher candidate ended up working was similar. I added an interaction
term for the espoused preferences. The initial survey asked teacher candidates whether
working in a school where the socioeconomic status of the students was similar to their
own background was one of the most important features in their job search. Thus, it is
reasonable to expect a negative relationship between the percentage of FRL students in
the school and teacher candidates with highly educated mothers if they have this
espoused preference.

The third set of analyses focus on the likelihood the teacher candidate ended up

working in an urban school. As the outcome is a dependent variable, I used a binomial

189

logistic regression model using a logit link function. The dependent variable is the log
odds that the teacher candidate ended up working in an urban school:
L0git[9(><)] = 108[9(X)/(1-(9(X))] = Bo + X13

(2)

The independent variables used in this model were similar to those used in the linear
models for percent minority and FRL students. It included a vector of teacher candidate
characteristics, a vector of espoused preferences, a dummy variable indicating whether
the teacher candidate named an urban or low-income suburban school as the top choice as
a measure of preferences in use, and a set of dummy variables for the timing of the ﬁrst
job offer.

Table 35 presents the results of regression analyses that examine relationships
between the percentage of minority students in a school and teacher candidate
characteristics, preferences, and hiring timeline. Column 1 includes the full model. Mid-
career changers ended up teaching in schools with about 42.0 percentage points more
racial minority students. Teacher candidates who did not take the ACT worked in schools
with fewer percentages of minority students, although a teacher candidate’s ACT score
otherwise had no relationship with the racial background of students in the school.
Neither the teacher candidates’ race nor socioeconomic status was related to the
percentage of minority students in the school where they ended up working when other

variables were controlled.

190

Table 35 - Regression Analyses Predicting the Percentage of Minority Students in School

 

 

 

 

Where Teaching
1 2

Variable Estimate Error Estimate Error
Intercept 43.46 28.26 43.57 26.56
African American 23.48 16.12 22.13 15.29
Other racial minority 11.11 1 1.31 10.12 10.76
Male -11.49 15.85 -l6.49 14.80
Mid-career changer 4203*M 9.80 44.71*** 8.62
Mother has BA or above 3.14 7.43 5.28 6.82
ACT score -1.20 1.10 -1.00 1.04
ACT score is missing -14.84~ 8.83 -14.44~ 8.36
Many racial minority students is important 1.71 12.40

Percent minority students in top choice school 045’” 0.15 0.41 ** 0.14
Received ﬁrst offer in June or July 7.89 9.88

Received ﬁrst offer in August or later 2.76 10.64

R2 0.385 0.384

N 72 79

 

Note: The results in this table are different if Oak State respondents are excluded from

the sample. See Table A4 in Appendix D.
~ p<.1, ** p<.01, *** p<.001.

Teacher candidates’ espoused preferences for working in a school with many

racial minority students was not related to the student racial composition of the school in

which they ended up working. Their preferences in use, however, were related to where

they ended up. Teacher candidates who named a school with greater proportions of

minority students as their initial top choice school ended up teaching in a school with

more minority students. For every one percentage point increase in the percent minority

students in teacher candidates’ initial preference, there was a .45 percentage point

increase of minority students in the school where they ended up. This indicates that the

student racial composition of the school where teacher candidates ended up working was

related to the composition of the school they named as their most desired school before

191

they applied for any teaching jobs, even controlling for the teacher candidate’s
background.

The timing of the teacher candidates’ ﬁrst job offer was also not related to the
student racial composition of the school in which they ended up teaching. Teacher
candidates who did not receive any job offers until June, July, August or later did not end
up in schools with greater or fewer minority students than teacher candidates who
received a job offer early in the hiring season. Further, a joint test of signiﬁcance found
that the month of ﬁrst offer variables and the espoused preference for working with
minority students did not explain much additional variance (F-value=.26, prob.=.857).

Table 36 presents regression results for the relationship between the percentage of
F RL students in a school and teacher candidate characteristics and preferences. Similar to
the ﬁndings for percentage minority students, mid-career changers ended up working in
schools with higher proportions of FRL students. The ﬁndings for the relationship
between the socioeconomic background of the teacher candidate and the percentage of
FRL students in the school in which they ultimately ended up teaching are mixed. In the
full model, the main effect for teacher candidates’ mother’s education was not related to
the socioeconomic status of the students in the school where they ended up working, but
there was an interaction with their espoused preferences. Teacher candidates with highly
educated mothers who said that working in a school with students from similar
socioeconomic backgrounds as themselves did end up working in schools with fewer
FRL students. That is, teacher candidates from higher socioeconomic backgrounds who
had an espoused preference for working with similar students did end up teaching

relatively more afﬂuent students. These coefﬁcients were large but not statistically

192

signiﬁcant. Very few teacher candidates indicated that student socioeconomic status was
one of the most important features in their job search. Thus, the lack of variation on this

variable may be affecting the results.

Table 36 - Regression Analyses Predicting the Percentage of FRL Students in School

 

 

 

 

Where Teaching
1 2 3
Variable Estimate Error Estimate Error Estimate Error
Intercept 52.13~ 28.41 64.15* 26.87 6763* 25.51
African American 5.05 15.76 4.04 15.74 5.09 15.18
Other racial minority 7.40 11.69 12.01 1 1.24 13.02 10.79
Male 6.04 16.92 5.19 16.91 -0.45 15.98
Mid-career changer 2420* 9.84 2392* 9.82 24.44" 8.97
Mother has BA or above -10.43 7.83 -l3.69~ 7.48 -12.01~ 6.96
ACT score -1.28 1.08 -1.58 1.03 -1.52 0.99
ACT score is missing -0.68 8.43 -1.85 8.36 -0.59 8.03
Students from same
socioeconomic 30.84 22.99
background is important
Students from same
soc1oeconom1c
background is -48.25 39.41
important*Mother has
BA or above
Percent FRL students in
top choice school 0.30~ 0.15 0.33* 0.14 033* 0.13
Rece‘ved ﬁrs‘ Offer ‘" 18.12~ 10.05 13.84 9.59
June or July
Recelved ﬁrst offer in 6.18 10.25 3.05 995
August or later
R2 0.330 0.302 0.293
N 61 61 67

 

~ p<.1, * p<.05, ** p<.01.

Column 2 shows results without including the variables for espoused preferences.

A joint signiﬁcance tests suggests removing these variables did not signiﬁcantly affect

193

model ﬁt (F-value=1.01, p-value=.372). Teacher candidates with highly educated
mothers ended up teaching in schools that had 13.7 percentage points fewer F RL
students. Also similar to the results for percentage minority students was that teacher
candidates’ preferences in use were related to where they ended up teaching. Each one
percentage point increase in the percent FRL students of teacher candidates’ most desired
school in the initial survey was associated with a .33 percentage point increase in the
percent FRL students where they ended up working. Finally, there was no relationship
between the month in which teacher candidates received their ﬁrst job offer and the
percentage of F RL students in the school in which they ended up working. Column 3
presents a model without dummy variables for the timing of the job offer. The timing
variables are neither individually nor jointly signiﬁcant (F -value=1 .30, p-value=.281).
Urban schools also face staffing difﬁculties and it is important to understand how
teacher candidates’ preferences inﬂuence who ended up teaching in urban schools. Table
37 presents results from logistic regression analyses predicting which teacher candidates
ended up teaching in urban schools. Mid-career changers had over 6.9 times the odds of
ending up in an urban school compared to their peers who were recent college graduates.
Other teacher candidate characteristics were not related to the likelihood that the teacher

candidate ended up teaching in an urban school.

194

Table 37 — Logistic Regression Analyses Predicting Whether a Teacher Candidate Ended

Up Teaching in an Urban School

 

 

 

 

1 2

Odds Odds
Parameter Estimate Error ratio Estimate Error ratio
Intercept -2.10 2.53 -3.53*** 1.02
Aﬁ'ican American 0.50 1.24 1.65
Other racial minority 0.07 0.96 1.07
Male 087 1.26 0.42
Mid-career changer 225* 0.89 9.46 1.93** 0.73 6.87
Mother has BA or above 0.76 0.69 2.14
ACT score -0.09 0.10 0.91
ACT score is missing 0.26 0.73 1.30
Many racial minority
students is important 0.07 1.1 1 1.08
Small class size is
important 0.09 0.96 1.09
Resource availability is
important -0.10 0.90 0.90
Close proximity to home is
important -0.51 0.90 0.60
Top choice school is urban
or low-income suburban
school 216* 0.93 8.70 1.82* 0.77 6.17
Received ﬁrst offer in June
or July 2.94** 1.13 18.91 258* 1.02 13.14
Received ﬁrst offer in
August or later 2.00~ 1.10 7.36 1.75~ 1.03 5.77
Pseudo R2 0.280 0.240
N 71 71

 

p<.1, * p<.05, ** p<.01, *** p<.001.

Because logistic regression models use maximum likelihood estimation, they

require larger samples. The relatively small sample size in this study may explain the

rather large coefﬁcients in this model. Reducing the number of independent variables

could compensate for smaller samples. Therefore, I ran a model omitting all independent

variables that were not statistically signiﬁcant in the ﬁrst model. Column 2 presents the

195

results of this model. While the coefﬁcients were slightly smaller, the ﬁndings were
similar.

The ﬁrst model included measures of teacher candidates’ espoused preferences
and preferences in use. As in the ﬁndings for percent minority and FRL students, teacher
candidates’ espoused preferences were not associated with the likelihood that the teacher
candidate ended up teaching in an urban school. Teacher candidates’ preferences in use
did seem to be related. Teacher candidates who named an urban or low-income suburban
school as their top choice school at the beginning of the job search had 6.2 times the odds
of teaching in an urban school.

The month in which teacher candidates received their ﬁrst offer was associated
with the likelihood that teacher candidates ended up teaching in urban schools. Compared
to teacher candidates who received their ﬁrst job offer in May, teacher candidates who
received their ﬁrst job offer in June or July or August or later had between 13.1 and 5.8
times the odds of working in an urban school. Teacher candidates who received their ﬁrst
job offer in June or July had the highest likelihood of ending up teaching in an urban
school. These ﬁndings support other research that suggests urban districts may lose high
quality teachers because they make job offers later in the summer.

Discussion and Conclusion

Overall, the data indicate that teacher candidates end up working in schools that
match their racial background. Further, teacher candidates who are switching into
teaching from another career end up working in schools with higher concentrations of
poor and minority students than did new college graduates. This chapter analyzed

patterns in teacher candidate applications, job offers, and labor market outcomes to shed

196

light on the job search and hiring processes that contribute to this sorting of teachers
across schools. In doing so, this chapter focused on the extent to which teacher candidate
preferences or school and district hiring practices contributed to this teacher sorting. The
evidence suggests that both teacher candidate preferences and hiring practices are
important, but the type of teacher sorting depends upon teacher candidate and school
characteristics. Teacher candidate behavior appears more important in explaining how
Aﬁican American teacher candidates and teacher candidates from various socioeconomic
backgrounds become sorted among schools, while hiring practices seem to inﬂuence why
mid-career changers and new college graduates end up teaching in different schools.

The sorting of teacher candidates ﬁ‘om different racial backgrounds seems to be
largely driven by teacher candidate decisions about where to apply. African American
teacher candidates apply to districts with more minority and low-income students than
other teacher candidates do. The schools from which both White and African American
teacher candidates receive job offers have higher proportions of low-income and minority
students than the average district in which they apply. But as African American teacher
candidates apply to districts with more minority students, they end up teaching in schools
with greater percentages of minority students. Further, the timing of job offers has no
relationship to the percentage of minority students in schools where the teacher
candidates end up teaching. Rather, teacher candidates’ preferences at the beginning of
the job search are a better predictor of the racial composition of the school in which they
ultimately teach. Thus, the sorting of teachers among schools is inﬂuenced by teacher
candidate self-sorting. This preference for teaching students of the same race is consistent

with other research on the distribution of teachers (Betts et al., 2000; Hanushek eta1.,

197

2004; Strunk & Robinson, 2006). The preferences of African American teacher
candidates and their decisions about where to apply results in them teaching in schools
with different characteristics of students than White teacher candidates.

The relative inﬂuence of teacher candidate preferences and hiring practices in the
sorting of teacher candidates across schools of varied socioeconomic contexts is more
mixed. Teacher candidates whose mother never went to college are more likely to work
in schools with the highest concentration of low-income students and in districts with
lower per-pupil expenditures, while teacher candidates whose mother has a master’s
degree are least likely to work in schools with high concentrations of low-income
students and, instead, worked in districts with higher per-pupil expenditures. Similar to
the sorting of teacher candidates by racial background, teacher candidates’
socioeconomic background and initial preferences for where to teach is signiﬁcantly
related to the social class composition of the schools in which they end up teaching, while
the timing of the job offers has no effect. This suggests that initial preferences are at least
as important as hiring practices in explaining the sorting of teacher candidates by the
socioeconomic status of students.

School and district hiring practices are related to the distribution of mid—career
changers and new college graduates among schools. Teacher candidates’ career status is
not related to the characteristics of districts in which they apply. While new college
graduates are more likely to apply to districts in other states, on average they apply to
districts with similar proportions of low-income and minority students. Yet mid-career

changers ended up working in schools with greater proportions of minority and low-

198

income students. This difference is most likely due to school and district hiring practices
and will be explored in more detail in the next chapter.

Previous research indicated that high poverty and urban districts hire lower
quality teachers because their late hiring timeline means they lose teachers to more
afﬂuent districts (Johnson et al., 2004; Levin & Quinn, 2003). The data presented here
offer mixed support for this ﬁnding. The timing of teacher candidates’ job offers is not
related to the racial or socioeconomic backgrounds of students in the school where they
ultimately teach. That is, teacher candidates who do not receive any job offers until late
in the summer are no more likely than other teacher candidates to work in schools with
higher proportions of low—income or minority students. Yet the timing of job offers is
associated with whether a teacher candidate ends up in an urban school. Teacher
candidates who receive job offers later in the summer are more likely to ultimately teach
in an urban school. Thus, as anecdotal evidence would suggest and consistent with the
literature, the stafﬁng difﬁculties in urban schools do seem to be related to their hiring
timeline. The different patterns between urban schools and schools with higher
concentrations of low-income or minority students is noteworthy. While urban districts
do tend to have more low-income and minority students, there are variations among
schools within districts. As such, the timing of hiring may not completely explain the
sorting of teachers within urban or suburban districts, and this is a question that warrants
further research.

Another important ﬁnding from these data is that the type of school in which
teacher candidates want to teach at the beginning of the job search predicts where they

ultimately end up teaching. Teacher candidates who expressed a desire to teach in a

199

school with more minority or low-income students, or an urban school, end up teaching
in schools with those characteristics, while those who do not express that desire work in
schools with lower proportions of low-income or minority students or non-urban schools.
This also points to the importance of preferences in use in teacher candidates’ job search.
Teacher candidates who espoused preferences for working in schools with more racial
minority students, small class sizes, closer to home, greater availability of instructional
materials, and a high salary do not end up in schools with these characteristics. But,
teacher candidates end up teaching in schools that are similar to the school they named as
their most desired place to work in the beginning of the job search. While few teacher
candidates end up working in the exact same school they named, they do end up in
schools with similar characteristics, especially in terms of student race and social class
composition. Thus, it is their preferences in use—not their espoused preferences—that
are associated with where they end up teaching.

Teacher candidate preferences for schools in February and March have a strong
relationship to where they end up teaching at the end of the job search. This is supported
by the qualitative analysis presented in Chapter 4 as teacher candidates ﬁltered out
potential districts on the basis of little information and a perception of where they thought
they would ﬁt. Further, these data also support the distinction made in Chapter 4 between
espoused preferences and preferences in use. While teacher candidates’ espoused
preferences for many racial minority students is not related to the racial composition of
the school in which they end up working, their preference in use as expressed by the

percentage of minority or F RL students in their most desired school or naming an urban

200

or low-income school as their most desired school is related to where they ultimately
teach.

That teacher candidates’ initial preferences in use predict their ultimate labor
market outcomes has implications for the ability of policy to affect job search decisions.
Chapter 4 demonstrated that teacher candidates knew little about the speciﬁc features of
the school they named as their top choice school and thus do not know whether these
schools match their espoused preferences. Despite this limited information about schools,
teacher candidates have ideas about where they want to teach and ﬁt in as a teacher. The
importance of teacher candidates’ initial preferences and the ﬁnding that desires to work
in particular schools predict where they ended up teaching suggests that current
frameworks for understanding how teachers come to work in different schools are
limited. In some sense, teacher candidates are not making decisions about where to work
during the job search. An initial decision about where to work, at least from the point of
view of the teacher candidate, was made before the beginning of the hiring season based
on the schools to which they would or would not apply, when teacher candidates
completed the initial survey in February or March. Indeed, that initial decision may well
have been made even earlier, as the desire to continue one’s own educational experiences
is a major reason individuals enter teaching (Lortie, 1975). Improving hiring practices or
timelines may equalize the distribution of teacher candidates somewhat, but if teacher
candidates’ initial preferences and ideas about where they ﬁt guide decisions about where
they apply, then these policies may have limited efﬁcacy. Rather, policies that try to

change the number of qualiﬁed applicants in particular districts may have larger effects.

20]

One example might be to stimulate supply in areas with high concentrations of poor or
minority students.

School and district hiring practices do affect the composition of teachers in a
school. Decisions by teacher candidates about where to apply, however, also inﬂuence
who ultimately teaches in a particular school. Teacher candidates are thus sorting
themselves into and out of schools in ways that alter the composition of the teaching
force in schools serving students of varying racial and socioeconomic backgrounds. That
minority students are more likely to be taught by African American teachers is not in
itself a cause for concern. Indeed, some may think this is advantageous for minority
students. But African Americans are under-represented in the teacher labor pool (Kirby et
al., 1999) and in a tight labor market context with a greater demand for new teachers, the
sorting of teachers by race may lead to shortages and the need to ﬁll vacancies with lower
quality teachers in predominantly minority schools. Thus, the tendency for teachers to
sort themselves into schools by race can lead to inequities in students’ access to high
quality teachers, especially those students that reside in urban districts and are from
minority and low-income backgrounds. The next chapter examines the impact of school

and district hiring practices on teacher labor market outcomes.

202

CHAPTER 6: THE DEMAND SIDE OF TEACHER HIRING

Teacher candidates did not have complete control over their labor market
outcomes. The matching of teachers to schools depends on both teacher and school
factors (Boyd et al., 2006). Once prospective teachers applied for jobs, they had to wait
for schools or districts to offer interviews and, ultimately, teaching jobs. Chapter 3
reviewed the literature on principals’ hiring decisions and examined teacher candidates’
perceptions of what inﬂuenced whether or not teacher candidates received job offers. It
found that embodied cultural capital as measured by personal characteristics and
demographics played a large role in structuring decisions on the teaching ﬁeld. Chapter 3
also highlighted the importance of a sense of ﬁt in teacher hiring as teacher candidates
thought schools hired teachers who ﬁt into the school.

This chapter explores the outcomes of hiring decisions made by schools and
districts and the effect these decisions may have on the distribution of teachers across
schools. In doing so, it extends the ﬁndings in Chapter 3 by examining the relative
importance of cultural, human, and social capital in predicting which teacher candidates
received job offers. Further, this chapter explores whether districts varied in the teacher
attributes they used in hiring. Through this analysis, this chapter extends Chapter 5 by
investigating whether districts’ decisions about whom to hire contributed to the sorting of
teacher candidates among districts by race and class.

The ﬁrst section describes the overall labor market outcomes of teacher
candidates, such as how many teacher candidates received job interviews and job offers.

The second section examines the teacher candidate characteristics that were associated

203

 

with getting more interviews. The third section explores which teacher candidates were
the most likely to receive a job offer. The fourth section considers whether school hiring
practices contributed to the sorting of teachers among schools by examining whether
particular types of schools were more likely to make job offers to particular types of
teachers. The ﬁfth section examines the plans of teacher candidates who ended up
without a teaching job by the end of my data collection period. The ﬁnal section
concludes with implications from these results.
Overall Labor Market Outcomes

This section describes the ultimate job outcomes of teacher candidates in this

study. Table 38 presents descriptive statistics. It shows characteristics of teacher

candidates and characteristics of schools to which they applied.

204

Table 38 - Descriptive Statistics for Chapter 6

 

Variable N Mean SD Min Max
Number of interviews 161 3.40 4.25 0 40
Number of job offers 160 1.20 1.76 0 11
African American 159 0.063 0.244 0 1
Other racial minority 159 0.113 0.318 0 1
Male 159 0.075 0.265 0 1
Mid-career changer 159 0.346 0.477 0 1
Mother's education1 159 2-14 1-04 1 4
Sergrfflssronal contacts who gave advrce about job 155 5.59 5.23 0 50
Friends and family who are teachers 156 3.87 3.50 0 20
Friends and family who are principals 157 1.32 2.24 0 20
Friends and family who work in the district ofﬁce 157 0.62 1.37 0 10
Contacts through teacher education program 154 2.78 3.79 0 25
ACT score 161 24.65 3.17 15 32
Did not take the ACT 161 0.317 0.467 0 1
Additional certiﬁcate in math 161 0.217 0.414 0 1
Number of applications2 161 3-32 1-23 2 5
Applied to schools out of the state 158 0.418 0.495 0 1
Applied to private schools 159 0.333 0.473 0 1
Applied to charter schools 159 0.459 0.500 0 1

 

1 Mother’s education is coded as 1=high school or less; 2=some college; 3=college

graduate; 4=master’s degree or above

2 Number of applications is coded as l=did not apply to any teaching jobs; 2=applied

 

to 1 to 10 districts; 3=applied to 11 to 20 districts; 4=applied to 21 to 40 districts;
5=applied to more than 40 districts
Table 39 presents the labor market outcomes of teacher candidates. Over 90% of

teacher candidates received at least one interview and the average teacher candidate had
3.4 interviews. Overall, these frequencies do not provide strong evidence on the relative
inﬂuence of cultural or human capital on teacher candidates’ labor market outcomes.
New college graduates received more job interviews than mid-career changers. Teacher
candidates with low ACT scores received more job interviews and those who did not take

the ACT at all received fewer job interviews than those with high ACT scores. Teacher

205

candidates whose mother never went to college received fewer job interviews than those
with more educated mothers. Slightly less than 60% of teacher candidates received at
least one job offer and just less than 25% received more than one job offer. New college

graduates were more likely to receive multiple job offers than mid-career changers.

Table 39 - Labor Market Outcomes

 

Had at least one Number of Had at least Had multiple

 

 

interview interviews one oﬂer oﬁrs
All 90.7% 3.4 59.6% 24.8%
White 90.8 3.4 58.0 22.1
African American 100.0 4.6 80.0 30.0
Other racial minority 88.9 3.1 66.7 44.4
New college graduate 91.4 3.9* 65.4~ 31.7**
Mid-career changer 90.9 2.5* 50.9~ 12.7**
Female 91.8 3.5 61.2 26.5
Male 83.3 2.7 50.0 8.3
ACT score
Top quartile 88.9 3.1 66.7 24.4
Lower quartiles 91.8 3.8~ 56.7 27.8
Did not take ACT 89.5 1.9** 57.9 10.5
Mother ’s education
High school or below 86.7 2.6* 48.3 15.0
Some college 90.9 3.6 54.6 30.3
Bachelor’s degree 94.0 3.9 68.0 22.0
Master’s or above 93.8 4.6 81.3 50.0
Note: N=l6l
~ p<.1, * p<.05, ** p<.01.

Number of Interviews

This section investigates the relationships between teacher candidate

characteristics and labor market outcomes more closely through regression analysis.

206

Several factors may be associated with whether a teacher candidate is invited to interview
or offered a teaching job. Teacher candidates’ human, social, and cultural capitals were
hypothesized to be related to the number of interviews and job offers they received. The
estimated regression model is:

Y, = [30 + CUL T CAP*BI + HUMANCAP*B; + SOCIALCAP*B3 +

HARDSTAFF *[14+ NUMAPPS"‘|34+ #i

(3)

where Y i is the number of interview offers received by teacher candidates.
C ULT CAP is a vector of characteristics measuring the teacher candidates’ cultural
capital. Measures of cultural capital include the teacher candidates’ racial background,
their mother’s educational level, and their career status. H UMANCAP is a vector of
variables indicating the teacher candidates’ human capital. Teacher candidates’ ACT
score and whether they had a mathematics endorsement on their teaching certiﬁcate were
used to measure human capital.21 SOCIALCAP is a vector of variables measuring the
teacher candidates’ social capital. The number and type of individuals the teacher
candidate asked for advice about the teacher job search was used to measure social
capital. Teacher candidates indicated whether the individuals they asked for advice were
professional contacts such as their cooperating teacher or colleagues from student
teaching, friends or relatives who were also teachers, friends or relatives who were also
principals or school administrators, friends or relatives who worked in a district ofﬁce,

and contacts from their teacher education program. Further, it was hypothesized that

 

2' Having a science or special education endorsement was included in initial models but was consistently
found to be non-statistically signiﬁcant and was dropped.

207

teacher candidates who applied to more districts or high-demand districts would receive
more interviews.

HARDSTAF F is a vector of dummy variables indicating whether the teacher
candidate applied to districts that are traditionally hard-to-staff. Given the over-supply of
teachers in the state, it was expected that districts in other states had a greater demand for
new teachers and thus teacher candidates who applied out of state would receive more
interviews and job offers. Charter schools, private schools, and urban districts were also
considered to have a high demand for new teachers. N UMAPPS is a vector of dummy
variables indicating the number of applications submitted by teacher candidates. The
number of applications is a series of dummy variables indicating whether the prospective
teacher applied to 11 to 20 districts, 21 to 40 districts, or more than 40 districts. Applying
to 10 or fewer jobs was the omitted category.

Table 40 presents the results of this regression analysis. The results suggest that
teacher candidates’ cultural characteristics and social capital, as well as the number of
applications submitted, were related to the number of interview offers they received.
African American teacher candidates received more interview offers than did White
teacher candidates. On average, African American teacher candidates received about 2.6
more invitations to interview. This may reﬂect recognition on the part of schools and
districts about the need to hire a more diverse teaching force. It is not clear how hiring
ofﬁcials could make decisions on who to interview based upon race. It is possible that
hiring ofﬁcials noticed names that are associated with being Aﬁican American (for
instance, the ﬁrst name Shanise) on application materials. Or, perhaps, they had other

personal interactions with teacher candidates through job fairs or substitute teaching.

208

While male teachers are also under-represented among the elementary teaching
population, male teacher candidates did not receive more interview offers than female

teacher candidates.

Table 40 - Regression Analyses Predicting the Number of Job Interviews

 

 

Variable Estimate Error
Intercept -0.68 2.73
African American 2.63~ 1.37
Other racial minority 0.00 1.02
Male -0.18 1.23
Mid-career changer -0.71 0.84
Mother has BA or above 0.83 0.69
Professional contacts from ﬁeld experiences 0.04 0.10
Friends who are teachers 0.17~ 0.10
Friends who are principals 0.47** 0.16
Friends who work in the district ofﬁce -0.63* 0.29
Contacts through teacher education program 019* 0.09
ACT score 0.01 0.10
ACT score is missing 0.66 0.79
Math endorsement -0.05 0.81
11 to 20 applications 1.81~ 0.96
21 to 40 applications 333*" 0.96
Over 40 applications 1.52 0.96
Applied out of state 157* 0.72
Applied charter -0.70 0.71
Applied private 0.20 0.78
Applied urban -0.17 0.69
R2 0.338

N 142

 

~ P<-1, * p<.05, ** p<.01, *** p<.001.

 

Social capital also appeared to be related to the number of interviews teacher
candidates received, although the direction depended on the type of social capital.
Teacher candidates who talked about their job search with friends or relatives who were

principals or teachers or with individuals from their teacher education program received

209

more interview offers than teacher candidates without these forms of social capital.
Teacher candidates who talked about their job search with ﬁiends or relatives who
worked in a district ofﬁce received fewer interview offers overall than candidates who
did not. It is interesting that one form of social capital may hurt teacher candidates’
employment prospects. It may be that teacher candidates who spoke with acquaintances
who worked in district ofﬁces were given incorrect advice about the job search process.

The largest effect on the number of interviews received was the number of
districts to which teacher candidates applied. Teacher candidates who applied to between
21 and 40 districts received 3.3 more interviews, and those who applied to between 11
and 20 districts received 1.8 more interviews, than did their peers who applied to 10 or
fewer districts. This is not surprising as teacher candidates must apply to districts before
they can receive interview offers and sending out more applications means these teacher
candidates can receive more interview offers. Yet teacher candidates who applied to more
than 40 districts did not receive more interviews. Teacher candidates who applied to
teaching jobs in other states received about 1.6 more interview offers than those teacher
candidates who did not apply out of state, although those who applied to urban, charter,
or private schools did not receive more interview offers.

It was expected that teacher candidates with greater amounts of human capital
would receive more job interviews, although there was no evidence in these data. Teacher
candidates with an additional endorsement in mathematics did not receive any more
interview offers than teacher candidates without this qualiﬁcation. Likewise, there was no

relationship between a teacher candidates’ general ability as measured by the ACT score

210

 

and the number of interviews received. This model explained about one-third of the
variance in the number of interview offers received.
Which Teacher Candidates Received Job Offers?

While getting an interview offer is an important step in the job search, ultimately
teacher candidates need a job offer. Less than half of teacher candidates who received
any job offer only received one offer, thus it made sense to use a logistic regression
model with the dependent variable indicating whether the teacher candidate received at
least one job offer. I used a binomial logistic regression model using a logit link function.
The dependent variable is the log odds that the teacher candidate received at least one job
offer:

L0git[9(X)] = l0819(><)/(1-(9(X))] = Bo + X13

(4)

The independent variables used in the model are similar to those used in the linear
models for the number of interviews. That is, the same characteristics are assumed to be
related to both the number of interview offers and getting a job offer. It includes a vector
of human capital characteristics, a vector of social capital characteristics, a vector of
cultural capital characteristics, a vector of variables indicating whether the teacher
candidate applied to traditionally hard-to-staff schools, and a vector of dummy variables
indicating the number of applications submitted. Table 41 presents the results of the
logistic regression analyses predicting the likelihood a teacher candidate received at least
one job offer. There were some differences between the ﬁndings for the number of
interviews received and the likelihood of getting a job offer. For example, talking about

one’s job search with one’s social network and applying to more districts did not appear

211

 

 

to be related to the likelihood that a teacher candidate received at least one job offer.
Further, all the social capital measures and the number of applications together did not
have a signiﬁcant effect on the overall model ﬁt (LR statistic=4.26, 8 degrees of freedom,
prob=.833). For this reason, the second column excludes these variables in the analysis.

Cultural characteristics were related to whether or not a teacher candidate was
offered a job. Just as African American teacher candidates received more interview offers
than their White peers, they were also more likely to receive at least one job offer.
African American teacher candidates had 5.4 times the odds of receiving at least one job
offer than their White peers. Likewise, teacher candidates who came from a higher social
class background as measured by their mother’s educational level had almost twice the
odds of receiving a job offer than those teacher candidates whose mothers did not have a
college degree.

While teacher candidates with an additional endorsement in mathematics received
a similar number of interview offers, they were slightly less likely to receive a job offer.
Similar to the ﬁndings for the number of interviews was that applying to jobs out of state
increased the likelihood a teacher candidate would get a job offer. Teacher candidates
who applied to jobs in other states had 2.5 times the odds of getting at least one offer than
those who only applied in this state. Teacher candidates who applied to urban districts
had 2.4 times the odds of getting a job offer, although applying to charter or private
schools had no relationship to the likelihood of getting a job offer. A teacher candidate’s

gender had no relationship to the likelihood of receiving at least one job offer.

212

 

Table 41 - Logistic Regression Analyses Predicting Whether a Teacher Candidate Was

 

 

 

 

Offered a Teaching Job
1 2

Odds Odds
Variable Estimate Error ratio Estimate Error ratio
Intercept -3.76* 1.77 -2.67~ 1.59
Aﬁican American l.80~ 0.93 6.07 1.68~ 0.89 5.39
Other racial minority 0.59 0.64 1.81 0.41 0.60 1.50
Male -0.29 0.80 0.75 -0.28 0.74 0.76
Mid-career changer -0.43 0.53 0.65 -0.66 0.47 0.51
Mother has BA or above 0.70 0.45 2.02 0.68~ 0.41 1.97
Professronal contacts from _0.05 0. 0 6 095
ﬁeld experiences
Friends who are teachers -0.04 0.06 0.96
Friends who are principals 0.06 0.09 1.07
Friends who work in the
district ofﬁce 0'15 0'20 1'16
Contacts through teacher 0.03 0. 0 6 104
education program
ACT score 0.11 0.07 1.12 0.09 0.06 1.10
ACT score is missing 0.08 0.50 1.09 0.07 0.45 1.08
Math endorsement -0.74 0.52 0.48 -0.94~ 0.48 0.39
11 to 20 applications 0.83 0.64 2.30
21 to 40 applications 0.09 0.60 1.10
Over 40 applications 0.47 0.60 1.60
Applied out of state 103* 0.46 2.81 0.92* 0.41 2.51
Applied charter -0.21 0.46 0.81 -0.23 0.43 0.80
Applied private 0.29 0.50 1.34 0.24 0.43 1.27
Applied urban 1.13* 0.44 3.09 0.86* 0.41 2.36
Pseudo R2 0.223 0.188
N 142 149
~ p<.1, * p<.05.

 

While the preceding analysis examines which teacher candidates received job

offers, it does not examine which applications submitted by those teacher candidates

213

resulted in the offer. By taking advantage of the restructured dataset22 we can explore
characteristics of both the teacher candidate and the district that were related to an
application resulting in a job offer. The model is similar to the model for the likelihood a
teacher candidate received at least one job offer, with additional variables included to
capture additional district characteristics. Table 42 presents the results from these
analyses. The ﬁrst model in Table 42 is similar to column 2 in Table 41 as it includes
teacher candidate characteristics and dummy variables for high-demand schools. As in
the previous model, applications by Aﬁican American teacher candidates had 3.8 times
the odds of resulting in a job offer than applications by White teacher candidates. Also
similar to the previous models, applications to districts in other states were much more
likely to result in a job offer. Applications by male teacher candidates were no more
likely to result in a job offer than applications by female teacher candidates.

Some of the ﬁndings conﬂict with the previous model, however. In particular, this
analysis shows that applications to charter and private schools were much more likely to
result in an offer. Indeed, these coefﬁcients show the strongest relationship with having
an application result in a job offer. Applications by mid-career changers were less likely
to result in a job offer, suggesting a preference by hiring ofﬁcials for new college
graduates or younger candidates. An application by a mid—career changer had less than
half the odds of resulting in a job offer than an application by a new college graduate.
The model included whether the application was to a district in another state, so this
advantage for new college graduates was in addition to the ﬁnding that they were more

willing to move outside the state for a teaching job.

 

22 As described in Chapter 2, the restructured dataset has one observation per teacher candidate-district
application match. The 171 respondents who completed the follow-up survey applied to 805 districts.

214

 

 

 

.Sovo ... .58 .. .38 .

 

 

an ok- mR Z

_o_.o wCo 026 NM £58m

m _ .v ooo ...N: 3:83. 5E 888m

o?” mo .92 3:83.. .3858 388.2

So So moo- Go «mo 33. moo omo noo- 535

86 _ Go ...mnN 8.2 woo ......an Suzi

woé ooo ...mod wmé omo ...om.m oo.~m mmo ...oo.m 8:20

woe omo ...mo._ moo wmo ...ow; o; omo ...o_.m 8% oo 30

woo mmo w_.o- oﬁo _m.o 3.? who omo mmo- 3888825 532

om; omo omo om; omo mmo on; omo mmo memes & 88. Ho...

woo woo moo- oog woo oo.o ooo woo So- 28.- HQ...

2”.— omo _m.o mm; mmo omo om; vmo So 96% .8 <m m2 8:82

3o mmo . So- ovo mmo ..moo- ovo mmo ..moo- Homage 828-22

n: mmo 2o om; mmo omo mm; mmo wmo 232

v2 omo moo om; mmo moo of Vmo So 5:038 308 $50

moo So L : ood ovo .5: who moo ...ng 50:25. SSE...

3o .om.m- 3o ..mnd- owo ..va- 30885

QCGL kQLLmN ﬁashwm QCQL LQLKN SCENEW QCGL kQLLN 365.2%” LNuNEQanN
35 3.5 .35

 

 

m

 

N

 

cote 50a m E 3::QO :o:§=&< S“ 55055 wccomoem mama—«ﬁx :Emmewom oummmoq .. NV 033-

215

One advantage to using the restructured dataset is that allows me to use additional
characteristics of the speciﬁc district as predictors of whether an application resulted in a
job offer. Some districts made more job offers than other districts. In analyzing the
ultimate labor market outcomes for teacher candidates, determining which districts were
hiring is just as important as which teacher candidates they hired. Columns 2 and 3 in
Table 42 examine whether the student composition in districts was related to whether an
application resulted in a job offer. If districts with large concentrations of low-income or
minority students had a greater demand for new teachers, then teacher candidates in these
districts would appear more likely to get a job. For example, Tables 24 and 25 in Chapter
5 showed that African American teacher candidates applied to districts with greater
concentrations of low-income and minority students. Even if districts did not use teacher
candidate race as a factor in hiring, African American teacher candidates would appear
more likely to get a job if the districts to which they applied hired more teachers. While
including measures of student composition in districts did decrease the relationship
between being African American and the likelihood of an application resulting in a job
offer, the coefﬁcient is still large and statistically signiﬁcant. Also, increasing the
percentage of minority students in a district to which a teacher candidate applied
increased the odds that the application would result in a job offer.

Information on the percentage of F RL students in private schools was not
available and thus Column 3 could not include both private school status and percent
F RL students. A separate model (not shown) included both the percent minority students

and percent FRL students. However, the high correlation between these variables created

216

 

multicollinearity concerns. While the individual effects were not statistically signiﬁcant
when they were both included, they jointly affected model ﬁt.
Hiring Decisions and Teacher Sorting

Chapter 5 showed that teacher candidate behavior contributed to the unequal
distribution of teachers among schools because teacher candidates applied to different
districts. Schools and districts may also contribute to teacher sorting by making offers to
different types of candidates. For example, districts with high concentrations of minority
students may prefer to hire minority teachers. This may lead to teaching sorting even if
teacher candidates distributed their applications among districts equally. Another
advantage of the restructured dataset is that it allowed me to investigate whether certain
types of schools were more likely to make offers to certain types of teachers.

Table 43 used the restructured dataset to explore whether districts varied in the
types of teacher candidates to which they offered jobs. The model is similar to Table 42,
but also includes interaction terms between teacher candidate characteristics and district
characteristics. I omitted non-signiﬁcant terms omitted due to sample size constraints.
Logistic regression models use maximum likelihood estimation and require larger
samples. I omitted the non-signiﬁcant terms to maintain an adequate sample size per
independent variable once interaction terms were included. The ﬁrst column includes
interactions between teacher candidate characteristics and the percentage of minority
students in the district. The second column includes interactions with the percentage of
F RL students. The third column includes interactions with whether the district is an urban

district.

217

 

Table 43 — Regression Analyses Exploring Whether Districts Varied in the Teacher

Candidates They Offered Jobs

 

 

 

 

 

 

 

Percent minority Percent F RL
students students Urban districts

Parameter Estimate Error Estimate Error Estimate Error
Intercept -2.62* * * 0.22 -2.61* * * 0.23 -2.67*** 0.21
African American 1.53 0.98 0.72 1.26 1.23~ 0.71
Other racial minority 1.46" 0.53 1.63” 0.56 1.13M 0.40
Mid-career changer -1.74*** 0.51 -l .91*** 0.56 -l .42*** 0.4]
Out of state 197*“ 0.29 2.16*** 0.28 1.91*** 0.28
Charter 220*“ 0.57 2.28*** 0.61 228*“ 0.56
Private 2.84*** 0.63 2.96*** 0.61
Urban -0.47 0.32 -0.48 0.30 -0.48 0.36
Percent minority 1.12~ 0.62 130* 0.52
students
Percent FRL students 1.00 0.67
Interaction with main
effect

African American -1 .09 1.40 0.39 2.36 -0.53 0.91

Other racial minority -2.52* 1.09 -3.27* 1.41 -2.13** 0.76

Mid-career changer 2.21 * 0.87 3.22” 1.19 1.47* 0.59
Pseudo R2 0.189 0.176 0.192
N 773 724 773

 

~ p<.1, * p<.05, ** p<.01, *** p<.001.

The results are similar across all three types of districts, suggesting that districts
with greater numbers of minority students, greater numbers of FRL students, and urban
districts made similar hiring decisions. Table 42 found that applications from mid-career
changers were less likely to result in a job offer than those from their peers who were just
graduating from college. This result was consistent when interaction terms were included
in the model, but there was evidence that districts’ preferences for mid-career changers
varied. Applications to districts with more minority or F RL students and urban districts

were more likely to result in an offer if they came from a mid-career changer. This

218

 

'Q

suggests that districts’ preferences for mid-career changer or new college graduate varies
by the student composition and urbanicity of the district. Thus, district decisions about
who to offer jobs contributed to the sorting of mid-career changers and new college
graduates between urban and non-urban schools and the schools serving different
populations of students.

There was less evidence that district hiring decisions contributed to the sorting of
teacher candidates by race into schools with students from similar racial backgrounds.
Applications to districts with greater percentages of minority students were not more
likely to result in a job offer if they came from African American teacher candidates.
Further, applications from teacher candidates from other racial minority backgrounds in
general had a higher likelihood of resulting in a job offer. Yet as the percentage of
minority or FRL students increased in a district, or if the district was urban, applications
from teacher candidates from non-White and non-African American racial backgrounds
were less likely to result in a job offer.

Candidates Who Did Not Get Job Offers

About 44% of teacher candidates did not get any job offers. Table 44 shows the
plans for the next academic year for teacher candidates who did not end up with a
teaching job, including what their main activity would be and whether they planned to
apply for teaching jobs again. Most unsuccessﬁil teacher candidates planned to remain in
the education ﬁeld and continue looking for a teaching job. Slightly less than two-thirds
of unsuccessful teacher candidates planned to work as either short or long-term substitute
teachers. Another 8% of unsuccessful teacher candidates planned to work in education-

related jobs other than teaching, such as working at a daycare or preschool, tutoring, or

219

working at a YMCA. Twelve percent of teacher candidates without jobs took jobs outside

the ﬁeld of education and 9.3% enrolled in a master’s or other degree program.

Table 44 - Future Plans of Teacher Candidates Who Did Not Get Job Offers

 

 

 

Main activity Percent
Substitute teach 65.3
Work in non-education job 12.0
Work in education-related job 8.0
Enroll in degree program 9.3
Other 5.3
Apply for teaching jobs again next year Percent
Yes 82.1
No 3.9
Not sure 14.1
Note: N=78

Most unsuccessful teacher candidates also planned to apply for teaching jobs
again. About 82% of teacher candidates who did not get jobs this year planned to apply
again next year. Few unsuccessful teacher candidates, about 4%, did not plan to apply for
teaching jobs again. About 14% of teacher candidates who did not get a job this year
were not sure if they would apply again in the future.

Discussion and Conclusion

This chapter explored the relative importance of cultural, human, and social
capital in the labor market outcomes of teacher candidates. It also examined variation in
districts’ decisions about whom to offer jobs to investigate how districts’ hiring decisions
impacts the sorting of teacher candidates among schools. Overall, the ﬁndings suggest
that cultural characteristics of teacher candidates are strong predictors of their labor
market outcomes, although social capital also plays a role. Districts did vary in the types

of teacher candidates to which they offered jobs, but the evidence does not support the

220

notion that teacher candidates receive job offers in places where they ﬁt or that district
hiring decisions signiﬁcantly contributed to the sorting of teachers among schools.

Cultural characteristics of teacher candidates play a large role in their labor
market outcomes. African American teacher candidates receive more interview offers and
are more likely to get a job offer, suggesting that hiring ofﬁcials care about the racial
background of teacher candidates. Chapter 3 showed that most teacher candidates thought
minority teachers had an advantage in the teacher labor market. The evidence in this
chapter indicates those perceptions are accurate, at least for African Americans. Further,
the tendency for districts to make job offers to African American teacher candidates is
universal. Districts’ decisions about whom to offer jobs does not appear to contribute to
the sorting of African American teacher candidates into predominantly minority schools.
That is, districts with greater proportions of minority students were no more likely than
other districts to make offers to African American teacher candidates. The growing
number of students of color in the US. and the increased focus on the achievement gap
between White and racial minority students has focused attention on the need to hire
more teachers of color. This suggests that districts are responding to the need for more
racial minority teachers.

Districts also appear to prefer younger teacher candidates who are just graduating
from college to older teacher candidates with other work experience. Applications from
new college graduates are more likely to result in job offers. Further, there is limited
evidence that teacher candidates from higher social class backgrounds are also more
likely to receive a job offer. Teacher candidates with more highly educated mothers are

more likely to have received at least one job offer, although this ﬁnding is not very

221

 

robust. As discussed in Chapter 3, the teacher hiring literature has found that principals
care about personal characteristics—such as enthusiasm, ﬂexibility, and communication
skills—when hiring teachers (Harris et al., 2006). Teachers from higher social class
backgrounds may be better able to display the appropriate personal characteristics in
interviews or new college graduates may have an advantage in appearing enthusiastic to
hiring ofﬁcials. Therefore, these characteristics may represent forms of cultural capital in
the teacher labor market.

Social capital also predicts the labor market outcomes of teacher candidates. In
particular, social capital is related to the number of interviews teacher candidates
received, but is not related to whether or not the teacher candidate received a job offer.
Perhaps having the right social contacts gives a teacher candidate an advantage in
obtaining an interview, but it does not help them get a job. Further, the type of social
capital is important as having contacts that work in a district office decreases the number
of interviews received. This ﬁnding has implications for future research on district hiring
decisions. This literature has not explored the impact of social networks on hiring
decisions, instead focusing on professional qualiﬁcations and personal characteristics.
Yet the importance of social networks in other occupations (Granovetter, 2005) and in
this study suggests more attention is needed to the role of social capital in teacher hiring.

Consistent with other research on teacher hiring, professional qualiﬁcations do
not predict teacher candidates’ labor market outcomes. In particular, districts do not show
a preference for hiring teacher candidates with higher ACT scores, consistent with
research that suggests districts do not hire teacher candidates from selective colleges

(Ballou, 1996). There is some evidence that having an endorsement for teaching math in

222

 

addition to the general elementary certiﬁcate makes it less likely that teacher candidates
will get a job offer. While mathematics is a traditional shortage area for secondary
schools, the market for elementary teachers does not give an advantage to math
specialists. This is noteworthy as many upper elementary grade teachers do specialize in
subjects and most elementary teachers lack deep knowledge of mathematics (Ball,
Lubienski, & Mewborn, 2001). Additional research is needed on the role of subject
matter knowledge in elementary teacher hiring.

Although there are overall patterns in which teacher candidates districts want to
hire, there is also variation in districts’ hiring decisions. In particular, not all districts
prefer new college graduates. Districts with more minority and low-income students, and
urban districts, are more likely to make job offers to mid-career changers. This preference
for mid-career changers is noteworthy given the proliferation of alternative certiﬁcation
programs. Many alternative certiﬁcation programs are designed to facilitate the career
change of individuals who want to move into teaching and to ease stafﬁng difﬁculties of
districts in shortage areas. All the teacher candidates in this study went through
university-based teacher preparation programs and the over-supply of new teachers in the
area should not force districts to hire teachers with less than full state certiﬁcation. Yet it
may be that districts serving large proportions of minority students have a higher demand
for mid-career changers—usually alternatively certiﬁed teachers—because they value the
experience of older candidates. The proliferation of alternative certiﬁcation programs
may reﬂect both the lack of adequate supply of teachers in high minority schools and the
preferences of hiring ofﬁcials in these schools for teachers with additional work

experience.

223

Chapter 3 emphasized the role of ﬁt in teacher hiring. It may be that urban
districts and districts with more minority and low-income students are more likely to
make job offers to mid-career changers because they are a better ﬁt with the existing
staff. The data used in this study cannot answer this question. The data do provide some
evidence on whether or not teacher candidates are more likely to get job offers in places
where they ﬁt in terms of the race and class composition of the schools. The ﬁndings
from this study do not support the notion that districts with more minority students are
more likely to make job offers to minority teacher candidates or that schools with
students from lower socioeconomic backgrounds make job offers to teacher candidates
from lower socioeconomic backgrounds. That is, the evidence does not suggest that
district decisions about whom to hire contribute to teacher sorting by the race and class
backgrounds of teachers.

Indeed, the data indicate that districts with more minority and low-income
students are less likely to make job offers to teacher candidates from non-White and non-
African American backgrounds. Additional research should examine more closely the
interaction of speciﬁc racial groups in teacher hiring. The small number of non-White
and non-African American teachers in the Smithton area limits the ability of this study to
examine this in more depth. Further, research should consider the relationship between
the characteristics of teachers already teaching in a school and teacher candidates
applying for jobs. The sense of ﬁt that teacher candidates described as important on the
teaching ﬁeld in Chapter 3 focused on ﬁt with teachers rather than students.

Finally, there is no evidence that districts are more likely to make job offers to

male teacher candidates. This is noteworthy given that elementary teachers are

224

disproportionately female. Chapter 3 found that teacher candidates expected males to
have an advantage in the elementary teaching job market because there are so few male
elementary teachers. Yet this was not the case. Conceptions of caring are central to
elementary teaching, which is often compared to mothering (N ias, 1999; Vogt, 2002).
Male elementary teachers experience tension and ambiguity in their ability to express
care for their students (Hansen & Mulholland, 2005). Therefore, it may be that male
teacher candidates are not perceived to ﬁt the idea of an elementary teacher. The results
in this chapter may reﬂect the tension between hiring more male elementary teachers to
balance the existing teaching force and hiring ofﬁcials’ perceptions of caring as female
work.

One limitation of this chapter is that the data do not provide evidence on the
processes through which district and school hiring officials made their decisions about
who to interview and who to offer a job. The data are limited to the outcomes of these
decisions; that is, which teacher candidates had interview offers and which applications
by teacher candidates resulted in a job offer. Despite this limitation, the data do provide
evidence on the relative importance of cultural characteristics on teacher candidates’
labor market outcomes and that the outcomes of districts’ decisions contribute to the
sorting of teacher candidates into different schools by teacher candidate career status but

not race or class background.

225

CHAPTER 7: CONCLUSION

Over 150,000 new teachers are hired every year throughout the country and
elementary schools have almost the highest number of annual job openings among
occupations in the state in which this study was conducted. This dissertation focused on
the processes through which new teachers come to ﬁll the job openings in particular
schools. I approached this by highlighting the strategies and processes teachers use to
ﬁnd teaching jobs and make decisions about where to teach. Using a theoretical
framework not previously applied to teacher career decisions—cultural reproduction
theory—I explored how these processes are related to teachers’ social and cultural
backgrounds. In summarizing the main ﬁndings of the dissertation and connecting these
ﬁndings to the larger body of research on teachers’ career decisions, this chapter develops
a more comprehensive framework to understand the forces that inﬂuence the current
distribution of teachers. It also suggests avenues for future research and implications for
policy and practice.

Summary

Overall, my research shows a tendency for teachers to self-sort into schools that
serve students from similar social and cultural backgrounds as themselves. Some of these
forces are deliberate—such as African American teachers’ preferences to work in a
school with many racial minority students. Others are more tacit—such as what teachers
think it takes to get a job, how they identify open positions, and how they perceive
schools during interviews. My study corroborates previous research which has shown that

district and school hiring practices also contribute to the sorting of teachers among

226

schools. Together, these forces lead to different labor market outcomes for African
American and White teacher candidates, teacher candidates from different socioeconomic
backgrounds, and mid-career changers and new college graduates.

Chapter 3 found that teacher candidates recognize the independent inﬂuence of
cultural, social, and human capital in the teacher labor market. While teacher candidates
recognize the importance of basic qualiﬁcations and certiﬁcation, most of them
emphasize the role of ﬁtting in at a school and their own personal characteristics as vital
to getting a job. These traits represent forms of embodied cultural capital because
teachers’ communication and interaction styles, the ways they present themselves, and
how teachers behave in particular social situations are traits acquired through a process of
socialization rather than instruction. This structure of the teaching ﬁeld corresponds to
what we know about principals’ hiring decisions and thus it seems that most teacher
candidates understand the “rules of the game.” A smaller number of teacher candidates,
however, downplay the importance of personal characteristics to focus on human capital
and teaching ability. These teacher candidates—who are more likely to be African
American, older mid-career changers, or from less afﬂuent backgrounds—also prefer to
teach in schools that the majority of teacher candidates consider less desirable, such as
urban or low-income suburban schools or charter schools. This suggests that similarities
exist in the structures of the teaching ﬁeld and the ﬁeld of schools, as teacher candidates
who appear to understand how principals make hiring decisions prefer to teach in schools
serving more afﬂuent communities and those that they view as selective.

Chapter 4 outlined the practical logic that teacher candidates bring to the job

search. This practical sense of the labor market emphasizes characteristics such as

227

comfort and familiarity as guiding job search decisions. The strategies teacher candidates
use to identify open positions results in them not learning about opportunities in districts
that were previously unfamiliar to them. Teacher candidates have a sense of where they
want to teach and where they do not want to teach even though they know very little
about these schools. Teacher candidates’ strategies for gathering information about
schools emphasize perceptions of comfort and ﬁt as they give equal weight to the nature
of interactions and “feeling” of a school as they do to the content of the interviews. Thus,
even though teacher candidates have espoused preferences for particular school attributes
such as beginning teacher support and a supportive principal, their preferences in use (as
suggested by the processes that guide their job search) emphasize familiarity, comfort,
and ﬁt.

Chapter 5 followed the progression of teacher candidates through the job search
as they sent out applications, went on interviews, received offers, and accepted jobs.
Teacher candidates become distributed across schools in part through a process of self-
sorting. For instance, African American teacher candidates apply to districts with greater
concentrations of minority students. Likewise, teacher candidates from higher
socioeconomic backgrounds are less likely to apply to districts in low-income
neighborhoods. This results in African American teachers ultimately teaching in schools
with greater concentrations of minority students and teacher candidates from higher
social class backgrounds teaching in schools with fewer low-income students. Further,
the attributes of the top choice school that teacher candidates initially identify are related
to the attributes of the school in which they end up teaching. If teacher candidates act on

their espoused preferences in the job search, then we would expect to see teacher

228

candidates with different espoused preferences applying to and accepting jobs in different
schools. Yet teacher candidates who say that working with minority students, being close
to home, having small classes, having readily available instructional materials, and
earning a high salary do not end up in schools that differ on these characteristics any
more than teacher candidates who do not say these features are important. Instead,
teacher candidates’ preferences in use for a school with greater numbers of minority or
low-income students, or an urban school, are associated with where they ultimately teach.

Chapter 6 focused on the school and district side of the hiring process. Cultural
characteristics predicted the likelihood that teacher candidates would get offers. In
particular, Aﬁican American teacher candidates are more likely to get job offers. Social
capital is also related to teacher candidates’ labor market success, although human capital
seems to be less important. Schools and districts do have varying preferences for the
candidates they hire. Although new college graduates are generally more likely to get a
job offer, schools with more minority or low-income students tend to hire mid-career
changers. This may explain why mid-career changers receive job offers in schools with
greater percentages of minority and low-income students than new college graduates
even though mid-career changers and new college graduates apply to schools with similar
characteristics.

Throughout these chapters, we see themes about the social and cultural context of
teachers’ career decisions. Cultural understandings and perceptions inﬂuence both how
teacher candidates view their position on the teaching ﬁeld and their opinions of schools
in which they apply and interview. A cultural reproduction framework highlights the role

of habitus in mediating between the structure of the ﬁeld and individual’s preferences,

229

perceptions, and behavior. Viewed through this perspective, it is not surprising that ﬁtting
in to a school’s culture becomes a central theme in teachers’ job search decisions.
Through interacting on the ﬁeld, teachers get a sense of the importance of ﬁt in hiring
decisions. In turn, teachers’ perceptions of whether they want to work in a school are
inﬂuenced by how they think they ﬁt, socially and culturally, with the existing staff.

Teachers do not ignore school working conditions or economic incentives that are
important in existing literature. They describe school organizational conditions such as
beginning teacher support and supportive leadership as central components in their job
search. Even though these responses may have been inﬂuenced by a social desirability
bias, I believe they were honest. Yet the ways that teachers’ socio-cultural backgrounds
inﬂuence their job search are sometimes hidden to the teacher. For example, initial
perceptions about their top choice school in which they see themselves teaching inﬂuence
their behavior about where to apply. The reliance on familiar districts and social networks
for identifying open positions inherently ﬁlter the types of schools to which they initially
apply. Further, the lack of information about salaries, professional culture, instructional
programs, and resources increase the importance of culturally informed perceptions and
(often negative) assumptions about schools.

To put it a different way, we could ask, “Did teachers end up in a school that they
wanted?” The answer to this question depends on how we deﬁne what they “want.” If, on
the one hand, we use teachers’ espoused preferences for particular characteristics of
schools in which they want to work, then teachers do not necessarily get what they want
from the job search. If, on the other hand, we deﬁne what they “want” as a school similar

to what they describe as their top choice school at the beginning of the hiring season—an

230

exemplar of what they deﬁne as a desirable place to work—then most teachers do end up
where they want to be. This distinction is important as it speaks to how researchers
conceptualize teachers’ preferences and the use of those preferences during the job
search. Teachers have a picture of the school and classroom in which they will work and
the job search process actualizes this picture despite the preferences they espouse.

This study suggests that school selection involves both rational calculations
concerning school organization, and, more fundamentally, a process of adjustment in
which teachers adjust their school selection to their perceived social position. Previous
studies have shown that teachers are distributed across schools in particular ways
(Lankford et al., 2002); this study shows that teachers themselves contribute to that
sorting process. As it illuminates this adjustment process, a cultural reproduction
framework expands our understanding of teacher labor markets to include teachers as
active participants in the process, noting that their actions are sometimes conscious,
sometimes unconscious decisions.

Limitations

1 addressed some limitations of the study in Chapter 2, including the focus on
newly prepared teachers as a source of new hires and the reliance on teachers from
traditional, university-based, teacher education programs. I want to address three
additional limitations here.

First, all data collection came from the teacher side of the hiring process. Some of
the ﬁndings, particularly in Chapter 6, explore hiring practices and focus on who schools

and districts want to hire. A more complete analysis of school and district hiring

231

preferences would collect data from the school or district, such as characteristics of their
applicant pool.

Second, the excess supply of elementary teachers in the Smithton area places
limitations on the generalizability of the study. The study took place in a metropolitan
area with a history of exporting teachers to others states due to over-supply and current
economic constraints that limited demand. Yet, I speculate that it is more likely that these
market conditions decreased, rather than increased, the impact of teacher decision-
making throughout the job search. Teachers in the interview component of this study
indicated they applied to some districts because they recognized the labor market
conditions. Thus, it is reasonable that a labor market with a greater demand or lower
supply of elementary teachers would allow teacher candidates to be more selective in
where they apply and possibly increase the degree of self-sorting.

The third limitation is related to the second as it speaks to generalizing to the
larger teaching population from a speciﬁc study of prospective elementary teachers.
Elementary teaching has—for better or for worse—oﬂen been compared to mothering
with a primary concern for the child rather than the content. Thus middle and high school
teachers, with their disciplinary background, may focus on content concerns rather than
feelings of comfort and ﬁt when thinking about where they want to teach. Secondary
teachers may make quite different decisions, although that is certainly a question that

could prove to be very interesting for study in a cultural reproduction framework.

232

Implications for Policy and Practice

Despite these limitations, this study suggests several implications for policy and
practice. I outline these implications for state and federal policymakers, teacher
educators, and school and district administrators.

There are several policy proposals to create incentives for teachers to work in
hard-to-staff schools. While these policies may improve the quality of teachers in hard-
to-staff schools given the right incentives, this dissertation suggests teachers’ job search
processes will limit the efﬁcacy of these policies. Because of their socio-cultural
backgrounds and inability to see how they are limiting their own job searches, teachers
simply do not know about all the jobs that were available in the Smithton metropolitan
area. How and where teachers look for a job inﬂuence where they end up teaching.
Developing a statewide job bank for teaching jobs and encouraging its use throughout the
state could help expose teachers to a greater range of jobs. The lack of such a resource in
this state forced teachers to rely on familiarity with districts and social networks to ﬁnd
jobs. Further, the statewide job bank should emphasize the position characteristics of the
vacancy, rather than the district. The Smithton area did have county job banks, but these
job banks were primarily organized by districts within the county so the district in which
the vacancy was located remained a prominent feature.

Teachers’ decisions are also hampered by the limited information available to
them about various schools. Their searches for information on what it is like to work in a
particular school turned up little more than school performance indicators, student
demographics, and the geographic location of the school. Creating an online database of

school characteristics could give teachers more information. This database could include

233

characteristics such as beginning and average salaries, class size, teacher turnover or
percentage of new teachers, induction policies, instructional expenditures, length of
tenure of the principal, and characteristics of other teachers in the school. As states are
creating workforce databases, this information is likely to be available. States should also
take advantage of these databases to help teachers make informed decisions about where
to teach.

But improving the hiring process would not solve all stafﬁng problems. Even with
objective data about schools, teachers formed subjective opinions about what it would be
like to teach in various schools. An effective set of policies should also include ways to
stimulate supply in traditionally under-served areas. If urban or rural areas continue to
face persistent stafﬁng difﬁculties, policymakers should ﬁnd ways to develop a pool of
individuals who want to teach in these areas. Examples include the Broad Partnership at
Michigan State University and Temple University’s Urban Education Collaborative.
Further, many alternative certiﬁcation programs are designed to prepare teachers for
hard-to-staff schools. Ensuring that these alternatively certiﬁed teachers are effective
educators is vital.

Teacher educators also have a role to play. As I mentioned above, we need
research on how teacher education practices can affect prospective teachers’ perceptions
of what it is like to teach in various schools and their preferences for where to teach.
Teacher education practices may be able to change prospective teachers’ preferences and
perceptions by helping them feel comfortable in socially distant schools. These practices
may likely include designing student teaching and pre-student teaching ﬁeld experiences

in hard-to-staff schools. But caution should also be taken so that prospective teachers are

234

successful in these environments; otherwise, teachers’ negative perceptions of these
schools may be reinforced. If prospective elementary teachers have a series of positive
ﬁeld experiences in urban schools, for example, then they may perceive urban schools as
places where they can feel comfortable and succeed as teachers. While these policies
could help, they are unlikely to solve persistent stafﬁng difﬁculties by themselves, as
teachers’ perceptions may be hard to change.

This dissertation also has implications for school and district administrators. First,
urban districts could streamline their hiring process so that their late hiring timeline does
not affect whom they ultimately hire. Second, school and districts administrators should
also be cognizant of the need to help teachers gain information on what it is like to teach
in their schools. Teachers were disappointed with the information about the schools and
salary structure available to them on district websites. Attending to the needs of potential
teachers and current applicants in designing district and school websites can help districts
manage teachers’ perceptions of what it would be like to work for them. Finally, the trend
to decentralize hiring and give more control to principals or site-based teams emphasizes
the need to gather information about the teacher candidates. Yet teachers are also using
these opportunities to gather information about the school. Principals and other hiring
officials should be aware of the messages they are sending to potential teachers through
their recruitment processes and interview situations and be conscious of using their
technology resources as marketing tools to teachers.

Extending Frameworks on Teachers ’ Career Decisions
What does this mean for how researchers think about teachers’ career decisions? I

want to highlight three main themes based on their implications for understanding job

235

search processes. First, teachers value a sense of comfort and ﬁt in choosing where, and
with whom, they want to work. Second, teachers approach the job search with a picture
of a school in which they want to teach that subsequently ﬁlters the options available to
them for where they might work. Third, teachers’ perceptions of schools are subjective
and despite the best efforts of teacher education and policy, habitus is difﬁcult to change.
Comfort and Fit

A sense of ﬁt was an important factor in teachers’ job search processes. Teachers
think a good ﬁt determines whether they receive a job offer. They also want to work in
schools where they think they ﬁt. Existing work on teachers’ career decisions highlights
the trade-offs teachers make between economic incentives, location, and working
conditions such as class size, instructional supports, and professional culture. Yet ideas
about comfort and ﬁt represent quite different conceptions of work-life in schools. They
are distinct from “working conditions” in that conditions suggest characteristics of
schools that impact an individual’s ability to perform a job while ﬁt is a more subjective,
affective response to a school itself. Yet comfort and ﬁt are central to how teachers think
about what it is like to work in a particular school.

Existing research on teachers’ career decisions also highlights the need for
teachers to feel successful in their work (Johnson & Birkland, 2003). Teachers
interviewed for this study indicate their comfort and sense of ﬁt with a school affect
whether they think they can be successful. Comfort and ﬁt relates to interactions with
both students and with other teachers. Teachers do not think they can be successful in
places where they feel uncomfortable relating to students. They also do not think they

will be supported in schools where they do not ﬁt in with the rest of the staff.

236

Some recent research has begun to explore the professional aspects of ﬁt and how
hiring practices can affect teachers’ ﬁt with a school (Liu & Johnson, 2006). My
dissertation highlights the idea that ﬁt is both professional and personal. Teachers have
both a sense of themselves as teachers and a sense of their social and cultural identity.
Both of these components affect where teachers think they ﬁt. Strunk and Robinson
(2006) found some support for the idea that teachers try to maintain their social identity
by working with teachers of similar races. Likewise, Podgursky, Monroe and Watson
(2006) found that female teachers with high ACT scores are more likely to exit teaching
if the colleagues in their school have lower ACT scores. This dissertation extends this
work by highlighting the role that social distance or proximity plays in teachers’ sense of
where they ﬁt and where they can maintain their social and cultural identity.

Initial Frames on Decisions

The importance of ﬁt and comfort on teachers’ career decisions can be
incorporated into a rational choice framework. A teachers’ utility function may include a
sense of comfort in the school and can be weighed against more traditional forms of
working conditions or economic incentives. But the importance of initial decision frames
in early stages of the job search suggests other theoretical perspectives can shed light on
these processes. Teachers bring initial guiding principles to their job search that then
structure their subsequent choices. These initial frames include:

o a sense that principals hire teachers who “fit” with the rest of the school;
0 strategies for identifying available positions—based both on social
networks and the reliance on speciﬁc district websites—that ﬁlter what

options they see as available to them;

237

o unspoken preferences that limit the schools to which they apply, such as
in-state or out of state schools, charter or traditional public schools, or
schools in unfamiliar neighborhoods; and

0 strategies for gathering information about schools that focus their attention
on student demographic characteristics and feelings of comfort and ﬁt.

My data provide some evidence that teachers do care about working conditions
and salary and do engage in some deliberative thought consistent with rational choice
theory when they were presented with multiple job offers. But the choices open to them
had been ﬁltered by the guiding frames and strategies at earlier stages in the job search.
Further, teachers’ initial ideas about where they picture themselves teaching serve as
frames for subsequent perceptions of schools. Teachers end up working in schools quite
similar to their initial preferences, even though those initial preferences are not always
consistent with their espoused preferences. Individuals enter the teaching profession with
ideas about what it is like to be a teacher and pictures of themselves as teachers which
come from many years in the “apprenticeship of observation” (Lortie, 1975). These ideas
are deeply ingrained and difficult to change through teacher education practices (Feiman-
Nemser, 1983). This dissertation suggests that teachers have ideas about what it means to
teach in particular contexts and the visions that teachers have about themselves as
teachers subconsciously place them in particular contexts.

Subjective Perceptions

Some of the research on teachers’ career decisions suggests that characteristics of

schools are objective and universally perceived as good or bad by teachers. Yet teachers’

perceptions of what it is like to work in schools are inherently subjective. There are

238

objective features of schools that lead to tangible differences between schools, but what
these features mean to teachers and whether these tangible differences are perceived to be
important are structured by the habitus. What types of interactions do teachers expect
with their students or colleagues? What reference group are they using to compare the
working conditions or salary in a particular school? Answering these questions involves
understanding the teachers’ socio—cultural background. Did teachers attend schools where
students acted in particular ways? How are they used to interacting with individuals in a
supervisory role, such as principals and teacher leaders?

Thus, teachers’ perceptions of schools as workplaces depend, in part, on the
assumptions and expectations teachers bring to the school. Some research has focused on
workers’ subjective perceptions of salary. For example, satisfaction with pay depends
upon what reference group individuals use to assess their pay and what expectations they
have for salary (Young, 1999). The ﬁndings from this dissertation suggest similarly
extending conceptions of working conditions to highlight their subjective nature. In
particular, frameworks guiding research on teachers’ career decisions should consider
how perceptions of working conditions interact with the social and cultural contexts of
schools and their students. Teachers’ sense of responsibility for student learning is
affected by the race and class status of their students (Diamond, Randolph, & Spillane,
2004) and teachers have culturally-based expectations for their interactions with other
teachers, parents, and especially students (Lareau & Horvat, 1999; Lipman, 1998;
McDerrnott, 1974).

Even the relative proximity of a school is subjectively constructed by teachers.

While teachers do recognize the objective distance and commuting time between their

239

home and potential schools, this objective distance interacts with socially and culturally
constructed ideas about the social distance to the school. As we saw in Chapter 4, longer
commutes are acceptable if they lead to schools that are socially and culturally familiar to
teachers. Thus, in trying to understand the importance of location in teachers’ job search
decisions (Boyd et al., 2005a), we should consider the interplay between social and
geographic proximity.

Future Research

The previous section outlined ways in which this work extends the current
literature on teachers’ career decisions, yet many questions remain. Here, I outline four
sets of questions for future work.

First, what happens when teacher candidates receive multiple offers? The small
sample size and labor market conditions in this study limited the ability to closely
examine the processes teachers used when deciding between multiple offers. There was
some evidence that the receipt of multiple job offers allowed teachers to engage in
rational decision-making, but that also their initial ideas about where they wanted to work
still inﬂuenced their decisions. More work is needed to understand these processes.

Second, what is the impact of different colleges or universities’ teacher education
programs on teachers’ careers? The role of the colleges was kept in the background of
this study. What can universities or colleges do to inﬂuence teacher candidates’
preferences or perceptions of schools? Nearly all the teacher candidates in this study
indicated that their teacher preparation program encouraged them to teach in urban
schools, yet this did not impact their job search decisions. Are there particular practices in

teacher education programs that can facilitate prospective teachers’ sense of comfort in

240

socially unfamiliar schools? Are there ways in which teacher education programs can
make prospective teachers familiar with hard-to-staff schools or extend their networks?
Also, do hiring ofﬁcials have preferences for teacher candidates from particular colleges
or universities? Given the current state of reform in teacher education and the growing
number of alternative certiﬁcation programs, this line of research will become
increasingly important.

Third, why do some teachers counter the tendency to choose schools that
reproduce their own schooling experiences? The sample size in this study precluded the
ability to focus on teacher candidates who were conscious about wanting to teach in
schools that are quite different from those with which they are familiar. Subsequent
research should focus on these teachers. Do they use different processes when making
decisions? Have they had signiﬁcant life events that force them to rethink their career
decisions? What about the impact of programs such as Teach For America that recruit
individuals with high social status from prestigious universities into hard-to staff schools?
How do these programs affect teachers’ decisions?

Fourth, how do teacher candidates present themselves to schools and districts?
While most teacher candidates think personal characteristics and ﬁt are important in
getting a job, some emphasize qualiﬁcations and teaching ability. Do these differences in
what teachers think guides hiring decisions affect how they present themselves to
potential schools? Analyzing teacher candidates’ resumes may help us understand what
information they think hiring ofﬁcials want to know. Further, the interview protocol in

this study focused on what teacher candidates were learning about schools during

241

interviews. A related study could focus on what teacher candidates try to convey about
themselves to schools on interviews.

Each of the policy suggestions and lines of inquiry mentioned above can improve
our knowledge of teachers’ career decisions and ultimately facilitate access to high
quality teachers for all students. For policymakers and school administrators, no one
proposal is a silver bullet; addressing the stafﬁng needs of schools serving traditionally
under-served students requires a multifaceted approach. For researchers, understanding
how teachers approach the task of ﬁnding a teaching job and the processes through which
they make decisions will inform this multifaceted approach to policymaking. This
dissertation represents the ﬁrst step in understanding teachers’ job search processes as it
expands our understanding of the social and cultural contexts in which teachers make

their job search decisions.

242

APPENDICES

243

APPENDIX A: INITIAL SURVEY INSTRUMENT

Note: The formatting on this instrument has been altered to ﬁt within the margin
constraints of the dissertation.

Section l—Preferences for schools
This section asks about your opinions of drﬂerent types of schools.
1. What will you most likely do in the 2006-07 school year?
a. Teach in a K-12 setting
b. Work as a substitute teacher
c. Work in a non-teaching job
(1. Student at a college or university
e. Other (Please specify: )

 

2. Will you apply/did you apply for teaching positions in elementary schools to start
teaching in the 2006-07 school year?
a. Yes —> Please continue ﬁlling out this survey.
b. No——> Thank you for your time. There is no need to complete this survey.
Return it in the enclosed envelope.

3. In considering where you would prefer to teach, how do you view each of the
following characteristics?

 

 

Would
prefer Would
not to Would Would strongly
have somewhat prefer prefer
Circle one in each row. this Neutral prefer this this this
a. small classes 1 2 3 4 5
b. close proximity to my home 1 2 3 4 S
:éztudents who remind me of myself at their 1 2 3 4 5
d. an effective student discipline policy . - _ '1 ' 2 ‘ ' 3 4 '1 3:5} ,1
e. good reputation for teaching practices 1 2 3 4 5
f. availability of textbooks and other 1 l ’ ...... ,1,: 3 '5 ,
instructional materials 1 - ' 1 . .- = . 11‘ "-" ‘ '
g. low- a-chieving students 1 2 p 3 , 4 5
h. high-achieving students ' ~ ‘ 11 1 r '- 1 ' , '2 1 3 -* 4 S
1. many English as a Second Language 1 2 3 4 5
students
j. many racial/ethnic minority students 1 2 3 4 5
k. a staff that talks about their instruction 1 2 3 4 5
1. teachers from the same racial/ethnic
1 2 3 4 5
background as me
m. students from the same racial/ethnic
l 2 3 4 5
background as me
n. faculty mentors/ beginning teacher support 1 2 3 4 5

244

0. parents at the school who remind me of

l 2 3 4 5
my own parents 1
p. good condition of the facilities 1 2 3 4 5
q. high salary , 1 l 2 3 4 5
r. curriculum that I like to teach I 2 3 4 5
s. a staff that gets along well 1 2 3 4 5
t. school that is similar to the one I attended

1 2 3 4 5
as a student
u. mission/vision of the school compatible 1 3 4 5
with my own
v. supportive school leadership 1 2 3 4 5
w. students from the same socio-economic 1 2 3 4 5

background as me

4. Of the features listed in question 3, what are the three features that are most important
to you?
Write in the letter corresponding to the three features that are most important to you. For
example, write in d, k, m.

First feature Second feature Third feature

Sometimes it helps people to talk about their opinions when given a speciﬁc example.
The next set of questions asks for you to name a speciﬁc school in which you would like
to work and to describe that school.

5. Write the name of a speciﬁc school in which you would most like to teach.

Note: Write the name of the school you consider to be your top choice at this time. If you
do not have a single top choice at this time, write the name of any school in which you
would be happy to teach. If you can’t think of a specific school, write the name of a
district. If your preferred school is a private or charter school, write the name of a
specific school. This information is only being collected for research purposes. The
school you name will not be contacted in any way.

a. School name:

 

b. District name: State:

 

6. How would you best describe the academic performance of this school? Circle one.
a. High achieving school
b. Average achieving school
c. Low achieving school

7. How would you best describe the community surrounding this school? Circle one.
a. Mostly urban
b. Mostly high income suburbs
c. Mostly moderate income suburbs
d. Mostly low income suburbs
e. Mostly rural or small town

245

8. Think about the school you named above in which you want to teach. Please $91; the
descriptor to indicate your level of agreement with each statement about features of this
school. The scale of responses ranges from “Strongly Disagree” (1) to “Strongly Agree”
(4). If you are unsure of some feature of the school, circle, “Don’t Know.”

 

 

Strongly Strongly Don’t

Circle one in each row. Disagree Disagree Agree Agree Know
a. This school has small classes. 1 2 3 4 DK
b. This school is close to my home. 1 2 3 4 _ DK
0. The students in this school remind me of 1 2 3 4 DK
myself at their age.
d: T1118. school has an effective student 1 2 3 4 DK
drscrplme policy.
c. This school has a good reputation for 1 2 3 4 DK
teaching practices.
f. Textbooks and other instructional materials 1 2 3 4 DK
are available when needed in, this school.
g. This school has many low-achrevrng 1 2 3 4 DK
students. 7
h. This school has many hrgh-achrevrng l 2 3 4 DK
students.
1. This school has many English as a Second 1 2 3 4 DK
Language students.
j. This school has many racial/ethnic
minority students. 1 2 3 4 DK
k. Teachers in this school talk about their 1 2 3 4 DK
Instruction.
1. Most teachers in this school come from the 1 2 3 4 DK
same racial/ethnic background as I do.
m. Most students in this school come from 1 2 3 4 DK
the same racial/ethnic background as I do. .
n. Mentors/beginning teacher support are 1 2 3 4 DK
provided to new teachers in this school.
0. Parents of students in this school remind 1 2 3 4 DK
me of my own parents. ,
p. The school building and facrlitres are in 1 2 3 4 DK
good condition.
q. This school has high teacher salaries. l 2 3 4 DK
r. This school has a curriculum that I like to l 2 3 4 DK
teach. .
3. Teachers in this school get along well. 1 2 3 4 DK
t. This school rs srmrlar to ones I attended as 1 2 3 4 DK
a student.
u. This school 5 missron or vrsron rs l 2 3 4 DK
compatible wrth my own.
v. This school has a supportive school 1 2 3 4 DK
leadership. ,
w. Most students in this school come from

. . l 2 3 4 DK
the same socro-economic background as 1 do.

246

9. Of the features listed in question 8, what are the three features that make this school
MOST desirable to you?
Write in the letter corresponding to the three features that make this school most
desirable. For example, write in d, k, m.

First feature

Second feature

Third feature

10. Do you have information about your top choice school from the following sources?
Circle Yes or No for each item.

a. My mentor/cooperating teacher Yes No
b. My university supervisor Yes No
c. Conversations/interview with the principal of this school Yes No
d. Conversations/ interview with teachers in this school Yes No
6. Experience student teaching at this school Yes No
f. Other direct personal experience with this school Yes No
g. Conversations with family/friends who have direct experience Y

. . es No
w1th thls school
h. School or district website Yes No
i. Newspapers/Media/T V Yes No

The next set of questions asks for you to name a particular school in which you would
NOT like to work and to describe that school.

1 1. Write the name of a speciﬁc school in which you would NOT want to teach.

Note: Write the name of the first school that comes to mind. If you can ’t think of a
specific school, write the name of a district. If your least desirable school is a private or
charter school, write the name of a specific school. This information is only being
collected for research purposes. The school you name will not be contacted in any way.

a. School name:

 

b. District name: State:

 

12. How would you best describe the academic performance of this school? Circle one.
a. High achieving school
b. Average achieving school
c. Low achieving school

13. How would you best describe the community surrounding this school? Circle one.
a. Mostly urban
b. Mostly high income suburbs
c. Mostly moderate income suburbs
(1. Mostly low income suburbs
e. Mostly rural or small town

247

14. Think about the school you named above in which you would NOT want to work.
Please gi_rgl_e the descriptor to indicate your level of agreement with each statement about
features of this school. The scale of responses ranges from “Strongly Disagree” (1) to
“Strongly Agree” (4). If you are unsure of some feature of the school, circle, “Don’t

 

 

Know.”

Strongly Strongly Don’t
Circle one in each row. Disagree Disagree Agree Agree Know
a. This school has small classes. 1 2 3 4 DK
b. This school is close to my home. 1 2 3 4 DK 7
c. The students in this school remind me of 1 2 3 4 DK
myself at their age.
(I. This school has an effective student 1 2 3 4 DK
discipline policy. ,
e. This school has a good reputation for 1 2 3 4 DK
teaching practices.
f. Textbooks and other instructional materials 1 2 3 4 DK
are available when needed in this school.
g. This school has many low-achieving
students. 1 2 3 4 DK
h. This school has many high-achievmg 1 2 3 4 DK
students.
i. This school has many English as a Second 1 2 3 4 DK
Language students.
j. This school has many racial/ethnic l 2 3 4 DK
minority students. _
k. Teachers in this school talk about their
instruction. . l 2 3 4 DK
1. Most teachers in this schoo1 come frOm the l 2 3 4 DK
same racial/ethnic background as I do.
m. Most students in this school come from 1 2 3 4 DK
the same racial/ethnic background as I do.
n. Mentors/beginning teacher support are l 2 3 4 DK
provided to new teachers in this school.
0. Parents of students in this school remind 1 2 3 4 DK
me of my own parents.
p. The school building and facilities are in 1 2 3 4 DK
good condition.
q. This school has high teacher salaries. l 2 3 4 DK
r. This school has a curriculum that I like to 1 2 3 4 DK
teach.
5. Teachers in this school get along well. 1 2 3 4 DK
t. This school is similar to ones I attended as 1 2 3 4 DK
a student.
u. This school’s mission or vision is l 2 3 4 DK
compatible With my own.
v. This school has a supportive school 1 2 3 4 DK
leadership.
w. Most students in this school come from 1 2 3 4 DK
the same socio-economic background as I do.

248

15. Of the features listed in question 14, what are the three features that make this school
LEAST desirable to you?
Write in the letter corresponding to the three features that make this school least
desirable. For example, write in d, k, m.

First feature

Second feature

Third feature

16. Do you have information about your least desirable school from the following
sources?
Circle Yes or No for each item.

a. My mentor/cooperating teacher Yes No
b. My university supervisor Yes No
c. Conversations/interview with the principal of this school Yes No
d. Conversations/interview with teachers in this school Yes No
e. Experience student teaching at this school Yes No
f. Other direct personal experience with this school Yes No
g. Conversations with family/friends who have direct experience

. . Yes No
With this school
h. School or district website Yes No
i. Newspapers/Media/TV Yes No

Section 2: Hiring Process
This section asks about your perception of how schools hire teachers and your
preferences for teaching in particular schools and general types of schools.

17. Overall, how would you rate your chances of securing a teaching job.
Circle one.
a. I am conﬁdent I will be offered a good teaching job.
b. I am unsure about whether I will be offered a teaching job. I might be offered a
teaching job, but it might not be the job I want.
c. I doubt I will be offered any teaching job. My job prospects are low.

18. If you ARE NOT offered a full-time teaching job, what will you most likely do
during the 2006-07 school year?
Circle one.

a. Serve as a substitute teacher

b. Work in a non-teaching job

c. Teach in a preschool program

d. Go back to school

e. Other (Please specify: )

 

249

19. How more or less likely are individuals with the following characteristics to be
offered a teaching position? Please gm the descriptor to indicate your perceptions about
what it takes to get a teaching job. The scale of responses ranges from “Much Less
Likely” (1) to “Much More Likely” (4).

 

Neither
Much Slightly more or Slightly Much
Individuals are more likely to get a teaching less less less more more
job if they. .. likely likely likely likely likely
a. have experience tutonng or mentoring l 2 3 4 5
children
b. have several years of teaching experience 1 2 3 4
e. have experience working with children in
. l 2 3 4 5
non-school settings
d. have a high GPA 1 2 3 4 5
e. have a master’s degree 1 2 3 4 S
f. have a full year of student teaching 1 2 3 4 5
experience
g. come from the same racial/ethnic 1 2 3 4 5
background as other teachers in the school
h. have high scores on teacher certiﬁcation 1 2 3 4 5
exams
i. ﬁt in well socially with the school staff 1 2 3 4 5
j. have experience teaching in urban schools 1 2 3 4 5
(including student teaching)
k. have a relative or friend who is a teacher 1 2 3 4 5
I. come from a racial/ethnic minority
1 2 3 4 5
background
m. have a relative or friend who works in a l 2 3 4 5
school district ofﬁce
n. have a relative or friend who is a school
. . 1 2 3 4 5
prmcrpal
o. are certiﬁed to teach multiple subjects or
I 2 3 4 5
grades _
p. look like other teachers in the school 1 2 3 4 5
q. are qualiﬁed to teach special education I 2 3 4 5
r. feel comfortable talking about personal
. . . . l 2 3 4 5
interests With the prinapal
s. are qualiﬁed to teach English as a Second
1 2 3 4 5
Language
t. come from a highly ranked university 1 2 3 4 5
u. are male 1 2 3 4 5
v. are White 1 2 3 4 5
w. are in the right place at the right time I 2 3 4 5

250

20. Of the features listed in question 19, what are the three features you feel will most
HELP your own chances of obtaining a teaching job?
Write in the letter corresponding to the three features you feel will most help you get a
teaching job. For example, write in d, k, m.

First feature

Second feature

Third feature

21. Of the features listed in question 19, what are the three features you feel will most
HURT your own chances of obtaining a teaching job?
Write in the letter corresponding to the three features you feel will most hurt you get a
teaching job. For example, write in d, k, m.

First feature

Second feature

Third feature

22. Will you apply (or have you applied) for teaching positions in the following types of
schools?
Circle Yes or No for each type of school.

a. Regular public schools in Detroit Yes No
b. Regular public schools in Lansing Yes No
c. Regular public schools in Flint Yes No
d. Regular public schools in Chicago Yes No
e. Regular public schools in high income suburbs near Detroit Yes No
f. Regular public schools in moderate income suburbs near Y

. es No
Detr01t
g. Regular public schools in low income suburbs near Detroit Yes No
h. Regular public schools in suburbs near Lansing Yes No
i. Regular public schools in suburbs near Flint Yes No
j. Regular public schools in suburbs near Chicago Yes No
k. Schools that serve primarily at-risk students Yes No
1. Charter schools/Public school academies Yes No
m. Private schools with a religious affiliation Yes No
n. Private schools with no religious afﬁliation Yes No
0. Schools outside of Michigan Yes No
p. Schools in rural areas Yes No
q. School(s) you attended as a student Yes No
r. School(s) where you student-taught Yes No

251

23. How much do you think your teacher preparation program emphasized teaching in the
following settings?
The scale of responses ranges from “Not emphasized at all” (1) to “Strongly emphasized”

(4)-

 

Not
emphasized Somewhat Strongly

at all emphasized emphasized
3. Schools in urban areas 1 2 3
b. Schools with many low income students 1 2 3
c. Schools with many racial/ethnic minority

1 2 3

students
d. Schools with many students who 1 2 3
struggle academically

Section 3—Teacher Preparation and Certiﬁcation Information
This section asks about your teacher preparation program and certiﬁcate areas.

24. Are you certiﬁed or do you expect to be certiﬁed to teach in an elementary school by
the start of the 2006-07 school year?

a. Yes

b. No —-> Skip to question 28

25. Which of the following endorsements do you have or will you have on your teaching
certiﬁcate?
Circle all that apply.

General Elementary certiﬁcate
Middle School certiﬁcate
Bilingual Education
Early Childhood Education
Language Arts
Social Studies
Science/Integrated Science
Mathematics
World Language
Music Education
Art Education
Visual Arts Education

. Special Education
Other (Please specify: )

P a Pr‘r'r'P‘qo rm .o-p 9‘s»

 

252

26. In what year did/ will you complete your initial teacher certiﬁcation program?

27. What is the name of the university in which you obtained/will obtain this

certiﬁcation?
a. Eastern Michigan University
b. Marygrove College
c. Michigan State University
d. University of Michigan-Ann Arbor
e. University of Michigan-Dearborn
f. Wayne State University
g. University of Detroit Mercy
h. Siena Heights University
i. Other (Please specify:

 

28. In what year did/ will you complete your ﬁrst undergraduate degree?

29. What is the name of the university in which you obtained/will obtain your ﬁrst

undergraduate degree?
a. Eastern Michigan University
b. Marygrove College
c. Michigan State University
d. University of Michigan-Ann Arbor
e. University of Michigan-Dearbom
f. Wayne State University
g. University of Detroit Mercy
h. Siena Heights University
i. Other (Please specify:

 

30. What is/was your undergraduate major and/or minor?
Mark (X) all that apply.

Major

Minor

 

Elementary education

Early childhood education

Math education

English education

Science education

Bilingual education

Special education

Other subject-speciﬁc education (e.g., social studies education, arts
education)

Other education (e.g., counseling or guidance)

Ethnic or Global studies (e.g., African-American or Latin American studies)
Business or Professional studies (e.g., Marketing, Agriculture, Law)
Arts (e.g., ﬁne arts, drama, music, design)

English (e.g., English Literature or Composition, Journalism or
Communications)

Foreign Languages

Mathematics (e.g., Mathematics or Statistics)

253

l—QF-W f'_| f—VF—Wf—‘f—i F! F—IFNP—Wf—II—HHI—N
L—u-Jh—ul H I—JHI—lL—J L—d L—Jh—JL—JI—dh‘h—dl—l

Hf—I ” HF—‘Hf—N '— Hf—‘FﬁF—VP‘HH
h—dL—l I—ul h—JL—lL—IH hi hit—J—JI—JHHH

Engineering or Computer Science

Science (e.g., Physics, Biology, Chemistry, Geology or Earth Sciences)
Social Sciences (e.g., Psychology, Economics, Sociology, Anthropology,
Political Science)

Other Humanities (e.g., History, Philosophy, Religious Studies)

[
[
[
[
General Studies or other Interdisciplinary Studies (e.g., Liberal Studies) [

£——J|.—.l |—_.l L—JL—J
l—i
L—J

31. Do you have a master’s degree, or will you have a master’s degree by the start of the
2006-07 school year? 0
a. Yes
b. No

32. For how long have you done each of the following?
Circle one in each row. Do not include activities you have done as part of your teacher
preparation program, such as ﬁeld experiences or student teaching.
Less 6 or
Not than 1 1-2 3-5 more
at all year years years years

 

a. Taught your own class in a preschool or child

care facility, either part-time or full-time 1 2 3 4 5
b. Substitute taught in either a public or private

elementary or secondary school, either part-time 1 2 3 4 5
or full-time

c. Worked as an instructional aide or

paraprofessional in either a public or private 1 2 3 4 5
school full-time 7

d. Worked in another non-teaching position in 1 2 3 4 5
the education ﬁeld full-time

e. Worked in another position outside the ﬁeld of

education full-time

f. Had primary care-giving responsibilities for
small children or older adults in need of care, 1 2 3 4 5
either part-time or full-time

Section 4: Background Information
This section asks for demographic and background information.

9’

33. What is your year of birth?

 

34. Are you male or female?
a. Male
b. Female

35. a. Are you of Hispanic or Latino origin?
a. Yes
b. No

254

b. What is your race?
Circle one or more races that you identify yourself to be.
a. White
b. Black or African-American
c. Asian
d. Native Hawaiian or other Paciﬁc Islander
e. American Indian or Alaska Native

36. What is the name of the high school that you graduated from?

 

37. In what city and state is this high school located?
City:
State:

 

 

38a. Did you take the ACT college entrance exam?
If you did not take the ACT, please go to question 39.
a. Yes
b. No l

 

b. If yes, what was your highest ACT score on the test?
Highest score on the ACT

 

39. Please write your ﬁrst and last name.
Note: You will not be identified individually in any way. As this research involves surveys
at two points in time, your name and contact information are necessary to link
information from the multiple versions of the survey. After we link the data, your name
and contact information will be deleted.

40. What is your mailing address?
Note: Please provide a mailing address that will still be valid and in use in September.
You will not be identified individually in any way. As this research involves surveys at
two points in time, your name and contact information are necessary to link information
from the multiple versions of the survey. After we link the data, your name and contact
information will be deleted.

 

41. What is your email address?
Note: Please provide an email address that will still be valid and in use in September.
You will not be identified individually in any way. As this research involves surveys at
two points in time, your name and contact information are necessary to link information
from the multiple versions of the survey. After we link the data, your name and contact
information will be deleted.

 

Thank you for completing this survey!

255

APPENDIX B: FOLLOW-UP SURVEY

Note: The formatting on this instrument has been altered to ﬁt within the margin

constraints of the dissertation.

Name:
Note: Your name is necessary so your responses to this survey and the previous survey

and be linked.

 

Section l—Additional information on your background

The initial survey you completed included a few questions about your background. To
save you time in completing the initial survey, some questions about your background
were not included on that survey.

1. What is your mother’s highest level of education? Circle one.
a. Below high school
b. High school
c. Some college
(I. Bachelor’s degree
e. Master’s degree
f. Doctorate or Professional degree

2. What is your father’s occupation and industry? If your father is no longer working,
write in his former occupation.

Father’s occupation:
Father’s industry:

 

 

3. What is your mother’s occupation and industry? If your mother is no longer working,
write in her former occupation.
Mother’s occupation:
Mother’s industry:

 

 

Section 2—Experiences on the job market
This section asks about your experiences searching for a teaching job and the types of
schools you applied to.

4. Think about the people you talked to or who gave you advice about your job search.
How many of those people ﬁt into each of the following categories? Write in a number
for each type of role.
a. Professional contacts such as your mentor or cooperating teacher, colleagues
from work or student teaching

256

 

b. Friends, family members, and acquaintances who are also teachers

c. Friends, family members, and acquaintances who are principals or other school
administrators

d. Friends, family members, and acquaintances who work in a district office __

e. Individuals who work for your teacher preparation program

 

5. How many teaching jobs did you apply for since January 2006?
a. 0
b. 1-10
c. 11-20
d. 21-40
e. More than 40

6. List 7 districts to which you applied and seriously pursued teaching jobs.
If you applied to more than 7 districts, list those you most seriously pursued. If you
applied to less than 7 districts, list all that you applied to. If you applied to charter or
private schools, list the name of the school.

District State

 

7. Did you apply for teaching jobs in the following types of schools? Circle Yes or No for
each type of school.
. a. Regular public schools in Detroit Yes No

    

c. Regular public schools in tht .

 

257

8. How many schools offered to interview you for a teaching job?
Number of schools

9. Did you turn down any interview offers?
a. No —» Go to question 11.

b. Yes —> Go to question lO.—1

10. Why did you turn down an interview offer?

 

11. Were you offered a teaching job for the 2006-07 school year?

a. No ——t Go to question 12.
b. Yes —+ Go to question 14. I

12. What will you most likely do during the
2006-2007 school year?

Circle one.

Work as a substitute teacher

Work in a non-teaching job (Specify: )
Teach in a preschool or daycare program

Go back to school, get a master’s degree
Other (Please specify: )

 

.09-.057?»

 

13. Will you be applying for teaching jobs in
other schools for next year?
a. Yes

b. No
0. Not sure at this point

If you were not offered a teaching job, your

survey is complete. Thank you for completing
this survey.

14. How many schools offered you a teaching job?
Number of teaching job offers

258

15. Please list the schools and their districts and states where you had job offers, and the
month in which you received each offer. This information is only being collected for
research purposes. These schools will not be contacted in any way.

 

 

 

 

 

 

School District State Month offer
received
Offer 1:
Offer 2:
Offer 3:
Offer 4:
Offer 5:

 

 

 

 

 

16. Did you accept any of these teaching job offers?

 

a. No —* Go to question 17. l
b. Yes —-* Go to question 20.

17. Why did you turn down your job offer?

 

18. What will you most likely do during the 2006-
07 school year? Circle one.

Work as a substitute teacher

Work in a non-teaching job (Specify: )
Teach in a preschool or daycare program

Go back to school, get a master’s degree
Other (Please specify: )

.09-99‘s»

 

19. Will you be applying for teaching jobs in
other schools for next year?

a. Yes

b. No

c. Not sure at this point

If you did not accept any teaching job offers, your
survey is complete. Thank you for completing this
survey.

 

20. Write in the name of the school and district in which you accepted a job offer and will
be teaching this year.
This information is only being collectedfor research purposes. This school will not be
contacted in any way.

a. School name:

 

b. District name: c. State:

 

259

 

21. How would you classify your position at this school, that is, the activity in which you
spend most of your time?

a. Regular full-time teacher

b. Regular part-time teacher

c. Itinerant teacher (i.e., your assignment requires you to provide instruction at

more than one school)
d. Long-term substitute teacher
e. Short-term substitute teacher

22. What is your base teaching salary?

 

23. Did you have any of the following types of experiences with this school or district?
Circle Yes or No for each type of experience.

_ a. Student taught Yes No
._b. Other ﬁeld experience during teacher preparation program _ Yes . No
0. Worked as aide or paraprofessional Yes No
(1. Worked as a substitute teacher Yes No
e. Attended the school or district as a student ,. . Yes 7 No
' f. Have a child who attends/attended this school or district Yes No
g. Have friends or family who work in this school or district Yes No
h. Other type of experience with this school or district (Specify: Y I i
) es No

 

24. How satisﬁed are you to be teaching in this school?
a. Very satisﬁed
b. Somewhat satisﬁed
0. Somewhat dissatisﬁed
d. Very dissatisﬁed

25. Will you be applying for teaching jobs in other schools for next year?
a. Yes

b. No
c. Not sure at this point

26. Did you have more than one job offer for a teaching position?

a. No —) Your survey is complete. Thank you for completing this survey.
b. Yes ——* Please continue ﬁlling out this survey.

260

27. How would you rate your chosen school compared to other schools in which you had
job offers in terms of each of the following aspects? Circle one in each row.

 

Better in Not Better in No information
chosen better or other to make this
school worse schools comparison

a. Class size 1 2 3 NA
b. Proximity, to my home 1 2 3 NA
c. Student behavior 1 2 3 NA

‘ (I. School discipline policy 1 2 3 NA ,
e. Reputation for teaching practices 1 2 ..3 NA
f. Availability of textbooks and other 1 I 2 i 3 NA
instructional materials
g. Student achievement 1 2 3 NA
h. Instructional dialogue among teachers 1 2 ' 3 NA
i. Parent involvement » . 7 1 2 3 NA
j. Support from school leadership 1 2 3 NA
k. Availability of faculty 1 2 3 NA
mentors/beginning teacher support

I. Condition of the schoolfacilities _ . l . 1 2, . 3 NA .
m. Salary 1 2 3 NA
11. School curriculum~ 1 2 3 NA
0. Social relationships among teachers 1 . 2 3 NA

’ p. Mission/vision of school . 1 2 3 . , YMNA,"
q. School leadership 1 2 3 NA

_‘ r. Student motivation, ,_ A . l 2 3 NA 1'
3. Safety of environment 1 2 3 NA

: t. Job security 7 _ , j . 1 2 _ 3 . ...,NA
11. Opportunities to make a difference in 1 2 3 NA
the lives of others . _
v. Sense of personal accompliShment_' 1 2 7 ' 3 NA
w. Teaching assignment (i.e., subject or 1 2 3 NA

grade level)

28. Of the features listed in question 27, what are the three features that most inﬂuenced
your decision to accept the teaching job at your chosen school rather than other schools in
which you bad job offers?
Write in the letter corresponding to the three features that most inﬂuenced your decision
to accept this particular teaching job. For example, write in a, e, h.

Most important feature

Second most important feature

Third most important feature

Thank you for completing this survey!

261

APPENDIX C: INTERVIEW PROTOCOLS

Initial Interview Protocol

1. Could you describe your background in education and your reasons for
enrolling in an elementary teacher education program? (Probe for why they are interested
in becoming a teacher, when they ﬁrst began considering teaching as a career, how
family responded when they expressed interest in teaching)

2. I am interested in other careers you may have considered or pursued. What
other careers did you seriously consider or actually pursue? Why are you going into
teaching as opposed to these other careers? (Probe for whether the respondent is a career
changer and whether they are applying for any non-teaching jobs)

3. Thinking back to when you ﬁrst decided to go into teaching, describe the
school in which you pictured yourself working.

4. Could you describe to me what you think schools look for when they hire
teachers? How do you think you compare to the characteristics that schools look for?
(Probe for how successful they think they will be, how they perceive job alternatives,
what do you think will hurt chances of getting a job?, Race, class, gender)

5. Please describe one of the schools to which you are applying. (Probe for salary,
beneﬁts, other features, how people respond when they say they are applying there).

6. Could you describe your job search process? (Probe for how the respondent
identiﬁes job openings and how many jobs they've applied to thus far)

7. How will you evaluate potential job openings? (Probe for the criteria used in

deciding whether to apply for a particular job, how much do you know?)

262

8. What features of a school are most important in determining whether or not to
apply for a job there? Why are these features important? When did you ﬁrst begin to
think about these features? (Probe for physical building, students, principal, other
teachers, community, geographic location)

9. How do you get information on these important features? (Probe for how
different individuals know about schools, sources and accuracy of information)

10. Do you plan to apply to any public charter schools or private schools? If so,
how will you get information about these types of schools? (Probe for whether the
respondent has a preference for public charter schools or private schools, what they think
it would be like teaching in a charter or private school)

1 1. Are there schools in which you would prefer not to work? (Probe for
characteristics of schools in which respondent would prefer not to work, how do they
learn about these characteristics)

12. Can you describe to me some of the interviews you had at different schools?
(Probe for what, if anything, made them excited, comfortable, or uncomfortable)

13. Please describe one of the schools in which you had an interview. (Probe for
salary, beneﬁts, other features, how people respond when they say they are interviewing
there, how it compares to their stated preferences, whether school is good match for
them)

14. Who do you talk to about your career decision?

15. In what ways, if at all, is (will) your preparation program likely to affect your
ability to ﬁnd a job? (Probe for effects of program reputation, relationships with mentors,

instructional supervisors, university instructors)

263

Second Interview Protocol
This interview was tailored to the teacher candidate’s progress in the job search process.

It was also based upon information learned in the ﬁrst interview.

1. Can you tell me how your job search has gone since we last spoke? (Probe for
additional schools applied to, any interviews or callbacks)

2. Can you describe to me some of the interviews you had at different schools?
(Probe for what, if anything, made them excited or uncomfortable)

3. Please describe one of the schools in which you had an interview. (Probe for
salary, beneﬁts, other features, how people respond when they say they are interviewing
there, how it compares to their stated preferences, whether school is good match for
them)

4. Can you tell me about any job offers you may have received? (Probe for
whether accepted or rejected, reasons behind decision, why this school is different)

5. If you haven’t received any job offers yet, can you tell me why you think that
is? (Probe for what they think would need to be different to get a job, what are their plans

now)

264

APPENDIX D: ANALYSIS TABLES EXCLUDING OAK STATE UNIVERSITY

RESPONDENTS

Respondents from Oak State University had low responses. To check whether this
low response rate affects the results, I ran all quantitative analyses twice: once including
Oak State respondents and one excluding Oak State respondents. Most analyses were not
substantially changed by the inclusion of Oak State respondents. The tables in this
appendix are those in which including Oak State respondents substantially change the

results.

Table A l - Regression Analyses of the Inﬂuence of Being White in Teacher Hiring,

Excluding Oak State Respondents

 

 

Variable Mean SD
Mean 3.13 0.79
Respondent characteristics Coeﬁ‘icient SE
Intercept -O.13 0.09
Master’s degree 083* 0.32
Mid-career changer 0.08 0.18
Male -O.42~ 0.22
African American O.50~ 0.25
Other racial ethnic minority 0,16 0.25
R-squared 0.107

N 206

 

Note: The dependent variables for these regression equations are teacher candidates’
perception of the impact of being White and of coming from a racial minority
background on their likelihood of getting a job offer. Results including Oak State
respondents can be found in Table 13.

~ p<.1, * p<.05.

265

Sovg t... .Svg ...... .mova .. avg z
.3. 2an 5 250m B :8 macoucomme 88m v30 menus—2: 3.33% 682

 

266

 

 

 

 

om mm um me 0: mg _N_ 2
5.3 1.52 tuﬁ ...mdm mSN ...m.m_ m.mm 30% co £832
wsm _.mm w.wm New Qom mom N._m uocwou mac—30mm
2m Dom qmm odm NAN 0.3. 33 owe—Bo 28m
n? 0.2 an. 4.2“ 2a to? n G 323 a Boss :2:
Schauxxuw a. L2§QV<
m.~m 2K mew m. _ m m. _ m oNv sham HU< 8.8 8: ED
m.mm cdm 9mm wém ﬁwm Bum w.wm who—ESE 8304
ﬁwm mew me 2mbm 5mm m.mm 04mm 05.55 aoH
muggy. .NDV
2.de ...owv *N.mm ._b.? odm ...m. _ m 3% commune 6030.22
ham :3 ...Em ...de EN 1.; 1mm 033sz 03:8 Ez
mom mém mom wSN m.mm wdm Qmm broad: 850
53 2v. _ m 9mm 2.on 1.1.3. 9mm m.mm 535:2 §o£<
m._m 52 9mm w._m cém 9mm 9mm 33>»
$6.? $0. _ m famwm {amend $03 $0.0m $53 =<
393m. BEEQ 333% 822.5 Roam“? aeEtEm. Lam: toitﬁm Lam: mEBES
cebeﬁmvoow xm\\e Bmémumx 3553. 2:85 3535. 2:39:
5E $928: new: as S 333% ﬁg S 33%

 

mucoccommom 88m $5 wins—8Q .32. 338044 28 Jumbo

3383* 6239a 3:66ch 8:089 9855 3m”: 5053 outméoozwom .5 moi 5m Bnmwzm 3535 Mo owﬂcoocom - m < 293.

 

n. :. 52 A. :. 38 2mm 32 6:. :3 25.8 25% 8 use“:
*3... *3. 8%. mg 4.3% 3.3 no... 2:. 9% same. {2053
new we a: 3% :38 E: Qom odm odm 03:8 meow

......a.._~ it? :3: 38 2nd: $2 1.: 5.2 team 323 s 328 swam
29:83:» a. $303
$6 298 $3 3.5.: in: ......L..mm 2; 3m 3m 5». 93 so: 29
9mm of 9% was. 3.9. 3% N3. an. at. 8:23.. 533
cam in 0.9. an... 2.? 38 mom M2; me. 2:32. a;
~335an
>2 mm 3% a: : 1&3 1.4% to: .32 team 53% 5030-22
9% m: mam 36:. :23. ...:Ssm he. in. tum 228cm 03:8 52
v.3. 2 _ 3.x. 3% ~80 «New o? 9% 3m arose 850
3:. SUN 2:. 3mm 3mm 38 m.mm 3m :.: cause... 505...
0.3. 0.2 53 0. am $2 $2 9mm 9mm 3,. 353
$3.4 $3: «1.2 no? can” 33 $3... $2 «1.3. =<
8:5 3:: :2: .3 SB SEE... ﬁx 2%. 323% asses
PBS .8 g N g Tm. N m. Ts 32563.. @3382 Beaumoow hmémomm

 

267

£ch 223 ESNSEEQ SEE EoﬁVotSﬂh 233% 83:. SSE S £85m.

mwcowcommom 88m me wcmvioxm .33. use Hobo .mcozmozma< 5% 2:03 805 8:8me

owﬁo>< wan moaﬁm 850 E 33 82.33.. can Jumbo Biooom .wo__&< 0:3 3:36ch 853% go owﬁqoaom - m < 2an

sova 2.. .Svg .. .mova .. .3“. 2
.0m 059—: cm 938 on ado masoﬁcoamoh 38m wa maﬁa—05 mzsmom .033 05 mo 50 USN 3.3m 05 E 200:8
Son 2.205 LPN 8 080: 88.. 82836 32:; 53> 332980 8:88 No owﬁcoeoa 05 98 0:5: 80¢ 85me :38 2:. ”8oz

 

ON. E. K w: m on NNN Z

. . . . . . . . . “Sagan: H88
m mm m M: m S. 0 Rm 0 com o mmm m mm v mm m 9. mm 080: 8 meNon

“antenna“ 508 8:

 

 

 

m9. 3: Nmm 5.23 @63. m.~m~ as». odm mi. a 0:8: 8 >35on
$28 823 3:3 ﬁx fmﬁo .3.?wa QQN memxo 3&3.me ﬁesta;
32: .8 3 N g N- m N nNé 32393.. 33.38% NumRNmuuV 33.38%
$.on was: Eokmuwﬁﬁq 253 Eckmunﬁmﬁ 23$ .83: ESQ £ ”383%

 

5:53 m... 2an

268

Table A 4 - Regression Analyses Predicting the Percentage of F RL Students in School

Where Teaching, Excluding Oak State Respondents

 

 

 

 

 

1 2 3
Variable Estimate Error Estimate Error Estimate Error
Intercept 8635* 32.22 104.03” 29.52 87.19“ 29.34
African American 18.63 17.44 18.68 17.45 17.39 17.91
Other racial minority -6.92 12.24 -0.81 11.46 3.21 11.64
Male 16.70 19.94 12.02 19.67 7.14 19.55
Mid-career changer 4.78 12.79 6.86 12.70 7.74 12.73
Mother has BA or above -7.60 7.99 -11.71 7.45 -1 1.20 7.32
ACT score -2.37~ 1.22 -2.89* 1.15 -2.23~ 1.13
ACT score is missing -7.24 9.01 -10.12 8.75 -7.71 8.99
Students from same
socioeconomic background
is important 28.90 21.85
Students from same
socioeconomic background
is important*Mother has
BA or above -47.66 37.33
Percent F RL students in
top choice school 0.28~ 0.16 034* 0.16 034* 0.15
Received ﬁrst offer in June
or July 17.36~ 10.19 13.34 9.78
Received ﬁrst offer in
August or later -6.34 10.11 -9.58 9.78
R2 0.385 0.350 0.231
N 49 49 54

 

~ p<.l, * p<.05, ** p<.01.

269

APPENDIX E: DISTRIBUTION OF TEACHER CANDIDATES BY GENDER

This appendix includes data on the districts and schools to which male and female
teacher candidates applied, from which they received offers, and where they accepted
jobs. Table A5 presents characteristics on the districts to which male and female teacher
candidates applied. Overall, there were few differences in the characteristics of districts
where male and female teacher candidates applied. Male teacher candidates applied to
districts that were closer to where they lived at the beginning of the job search and
districts with higher pupil-teacher ratios. There were no differences in the student

composition of the districts to which men and women applied.

Table A 5 - District Characteristics Where Male and Female Teacher Candidates Applied

 

 

District characteristic Female Male
High income suburbs near Smithton 55.2% 50.0%
Low income suburbs near Smithton 35.9% 25.0%
Mean percent F RL 27.3 25.1
Mean percent minority 31.3 25.6
Mean pupil-teacher ratio 17.1* 18.7*
Mean per-pupil instructional expenditures $5,186.48 $4,904.25
Mean salary $36,626.12 $36,054.4
Smithton Public Schools 13.8% 16.7%
Any urban district 66.7% 54.6%
Schools in other states 43.1% 33.3%
Distance from home 2392*” 26.1***

 

* p<.05, *** p<.001.

Table A6 presents characteristics on the districts and schools from which male
and female teacher candidates received job offers. As male teacher candidates applied to

districts that were closer to where they lived than female teacher candidates, they also

270

received jobs in schools closer to their home. They were also less likely to get a job offer
from a school in another state. Male teacher candidates received job offers from districts

with higher percentages of FRL students than female teacher candidates.

Table A 6 - District and School Characteristics Where Male and Female Teacher

Candidates Received Job Offers

 

 

Characteristics Female Male
Mean percent FRL

District 34.7~ 58.3~

School 39.6 67.0
Mean percent minority

District 46.1 66.3

School 42.3 64.9
Mean pupil-teacher ratio

District 18.0 19.4

School 18.0 21.3
Mean per-pupil instructional expenditures $4,784.36 $5,012.28
Mean salary $36,573.60 $35,799.30
Urban 38.3% 75.0%
Schools in other states 41 .9%* 0.0%*
Distance from home 4336*” 45.6***

 

~ p<.1, * p<.05, *** p<.001.

Table A7 presents characteristics on the districts and schools in which male and
female teacher candidates accepted jobs and ended up working. The results for where
male and female teacher candidates ended up working are similar to where they received
offers. Female teacher candidates ended up working in schools that were farther from
where they lived at the beginning of the job search than male teacher candidates. They

were also more likely to end up teaching in schools out of state. Male teacher candidates

271

ended up working in districts with greater percentages of F RL students than female

teacher candidates.

Table A 7 - District and School Characteristics Where Male and Female Teacher

Candidates Accepted Jobs

 

 

Characteristics Female Male
Mean percent F RL

District 33.8~ 62.2~

School 38.0 56.0
Mean percent minority

District 42.5 58.9

School 40.6 50.3
Mean pupil-teacher ratio

District 17.7 18.8

School 18.2 18.3
Mean per-pupil instructional expenditures $4,918.02 $4,617.19
Mean salary $34,650.53 $32,566.67
Urban 35.0% 40.0%
Schools in other states 41.2%~ 0.0%~
Distance from home 3904* ** 38.0** *

 

~p<.1, *** p<.001.

272

REFERENCES

Akerlof, G. A., & Kranton, R. E. (2005). Identity and the Economics of Organizations.
Journal of Economic Perspectives, 19(1), 9-32.

Antos, J. R., & Rosen, S. (1975). Discrimination in the market for public school teachers.
Journal of Econometrics, 3, 123-150.

Argyris, C. (1985). Strategy, Change, and Defensive Routines. London: Pitman
Publishing.

Argyris, C., & Schon, D. A. (1975). Theory in Practice: Increasing Professional
Effectiveness. San Francisco, CA: Jossey-Bass.

Argyris, C., & Schon, D. A. (1978). Organizational Learning: A Theory of Action
Perspective. Reading, MA: Addison-Wesley Publishing.

Arsen, D., Clay, T., Davis, T., Devaney, T., Fulcher-Dawson, R., & Plank, D. N. (2005).
Adequecy, Equity, and Capital Spending in Michigan Schools: The Unﬁnished
Business of Proposal A. East Lansing, MI: The Education Policy Center at
Michigan State University.

Baker, B. D., & Cooper, B. S. (2005). Do Principals With Stronger Academic
Backgrounds Hire Better Teachers? Policy Implications for High Poverty
Schools. Educational Administration Quarterly, 41(3), 449-479.

Ball, D. L., Lubienski, S. T., & Mewborn, D. S. (2001). Research on Teaching
Mathematics: The Unsolved Problem of Teachers‘ Mathematical Knowledge. In
V. Richardson (Ed.), Handbook of Research on Teaching (fourth ed., pp. 433-
456). Washington, DC: American Educational Research Association.

Ballou, D. (1996). Do Public Schools Hire the Best Applicants? Quarterly Journal of
Economics, 111(1), 97-134.

Banerjee, A. V. (1992). A Simple Model of Herd Behavior. Quarterly Journal of
Economics, 107(3), 797-817.

Baugh, W. H., & Stone, J. A. (1982). Mobility and Wage Equilibrium in the Educator
Labor Market. Economics of Education Review, 2(3), 253-274.

Bempah, E. O., Kaylen, M. S., Osburn, D. D., & Birkenholz, R. T. (1994). A
Econometric Analysis of Teacher Mobility. Economics of Education Review,
13(1), 69-77.

Berry Cullen, J., & Loeb, S. (2004). School Finance Reform in Michigan: Evaluating
Proposal A. In J. M. Yinger & W. D. Duncombe (Eds), Helping Children Left

273

Behind: State Aid and the Pursuit of Educational Equity. Cambridge, MA: MIT
Press.

Betts, J. R., Rueben, K. S., & Danenberg, A. (2000). Equal Resources, Equal Outcomes?
The Distribution of School Resources and Student Achievement in California. San
Francisco: Public Policy Institute of California.

Bikhchandani, S., Hirshleifer, D., & Welch, I. (1998). Learning from the Behavior of
Others: Conformity, Fads, and Informational Cascades. Journal of economic
Perspectives, 12(3), 151-170.

Bourdieu, P. (1977a). Cultural Reproduction and Social Reproduction. In J. Karabel & A.
H. Halsey (Eds), Power and Ideology in Education (pp. 487-511). New York:
Oxford University Press.

Bourdieu, P. (1977b). Outline of a Theory of Practice. Cambridge: Cambridge University
Press.

Bourdieu, P. (1980). The Logic of Practice (R. Nice, Trans). Stanford, CA: Stanford
University Press.

Bourdieu, P. (1984). Distinction: A Social Critique of the Judgement of Taste (R. Nice,
Trans). Cambridge, MA: Harvard University Press.

Bourdieu, P. (1986). The Rorrns of Capital. In J. G. Richardson (Ed.), Handbook of
Theory and Research for the Sociology of Education (pp. 241-258). New York:
Greenwood.

Bourdieu, P. (1998). Practical Reason: On the Theory of Action. Stanford, CA: Stanford
University Press.

Bourdieu, P., & Wacquant, L. J. D. (1992). An Invitation to Reﬂexive Sociology. Chicago:
University of Chicago Press.

Boyd, D., Lankford, H., Loeb, S., & Wyckoff, J. (2002). Initial Matches, Transfers, and
Quits: Career Decisions and the Disparities in Average Teacher Qualifications
Across Schools. Working paper.

Boyd, D., Lankford, H., Loeb, S., & Wyckoff, J. (2005a). The Draw of Home: How
Teachers' Preferences for Proximity Disadvantage Urban Schools. Journal of
Policy Analysis and Management, 24(1), 113-132.

Boyd, D., Lankford, H., Loeb, S., & Wyckoff, J. (2005b). Explaining the Short Careers of
High Achieving teachers in Schools With Low-Performing Students. American
Economic Review.

274

Boyd, D., Lankford, H., Loeb, S., & Wyckoff, J. (2006). Analyzing the Determinants of
the Matching of Public School Teachers to Jobs: Estimating Compensating
Differentials in Imperfect Labor Markets.Unpublished manuscript.

Brewer, D. J. (1996). Career Paths and Quit Decisions: Evidence from Teaching. Journal
of Labor Economics, 14(2), 313-339.

Cable, D. M., & Turban, D. B. (2003). The Value of Organizational Reputation in the
Recruitment Context: A Brand—Equity Perspective. Journal of Applied Social
Psychology, 33(11), 2244-2266.

Clotfelter, C. T., Ladd, H. F., & Vigdor, J. L. (2004). Teacher Sorting, Teacher Shopping,
and the Assessment of Teacher EffectivenessUnpublished manuscript, Duke
University, Durham, NC.

Cohn, E., & Geske, T. (1990). Human Capital and the Role of Government in Education.
In The Economics of Education (3rd ed.). New York: Pergamon Press.

Coleman, J. S. (1988). Social Capital in the Creation of Human Capital. American
Journal of Sociology, 94(Supplement), 95-120.

De Tray, D. N., & Greenberg, D. H. (1977). On Estimating Sex Differences in Earnings.
Southern Economic Journal(October), 348-353.

Delli, D. A., & Vera, E. M. (2003). Psychological and Contextual Inﬂuences on the
Teacher Selection Interview: A Model for Future Research. Journal of Personnel
Evaluation in Education, I 7(2), 137-155.

Diamond, J. B., Randolph, A., & Spillane, J. P. (2004). Teachers' Expectations and Sense
of Responsibility for Student Learning: The Importance of Race, Class, and
Organizational Habitus. Anthropology and Education Quarterly, 35(1), 75-98.

Dillman, D. A. (2000). Mail and Internet Surveys: The Tailored Design Method. New
York: Wiley.

Dolton, P., & Makepeace, G. H. (1993). Female Labour Force Participation and the
Choice of Occupation. European Economic Review, 3 7(7), 1393.

Dolton, P., & van der Klaauw, W. (1999). The Turnover of Teachers: A Competing Risks
Explanation. The Review of Economics and Statistics, 81(3), 543-552.

Eberts, R. W., & Stone, J. A. (1985). Wages, Fringe Beneﬁts, and Working Conditions:
An Analysis of Compensating Differentials. Southern Economic Journal, 52(1),
274-279.

Education Commission of the States. (2005). State Financial Incentive Policies for
Recruiting and Retaining Effective New Teachers in Hard-to-Staﬂ Schools, from

www.ecs.orchlearinghouse/61/61/6161.htm

275

Emley, K., & Ebmeier, H. (1997). The Effect of Employment Interview Format on
Principals' Evaluations of Teachers. Journal of Personnel Evaluation in
Education, I I, 39-56.

Falch, T., & Strom, B. (2005). Teacher Turnover and Non-Pecuniary Factors. Economics
of Education Review, 24, 611-631.

Feiman-Nemser, S. (1983). Learning to Teach. In L. S. Shulman & G. Sykes (Eds),
Handbook of Teaching and Policy (pp. 150-170). New York: Longman.

Ferguson, R. F., & Ladd, H. F. (1996). How and Why Money Matters: An Analysis of
Alabama Schools. In H. F. Ladd (Ed.), Holding Schools Accountable:
Performance-Based Reform in Education. Washington, DC: Brookings Institute.

Gatewood, R. D., Gowan, M. A., & Lautenschlager, G. J. (1993). Corporate Image,
Recruitment Image, and Job Choice Decisions. The Academy of Management
Journal, 36(2), 414-427.

Gilbert, D. (1998). The American Class Structure in an Age of Growing Inequality (5th
ed.). Belmont, CA: Wadsworth.

Glaser, B. G. (1965). The Constant Comparative Method of Qualitative Analysis. Social
Problems, 12(4), 436-445.

Goldhaber, D., & Brewer, D. J. (2000). Does Teacher Certiﬁcation Matter? High School
Teacher Certiﬁcation Status and Student Achievement. Educational Evaluation
and Policy Analysis(2), 129-145.

Goldhaber, D., & Liu, A. Y.-H. (2003). Occupational Choices and the Academic
Proﬁciency of the Teacher Workforce. In W. J. Fowler (Ed.), Developments in
School Finance: 2001-2002, NCES 2003-403. Washington, DC: US. Department
of Education, Institute of Education Sciences, National Center for Education
Statistics.

Granovetter, M. (2005). The Impact of Social Structure on Economic Outcomes. Journal
of Economic Perspectives, 19(1), 33-50.

Greenwald, R., Hedges, L. V., & Laine, R. D. (1996). The Effect of School Resources on
Student Achievement. Review of Educational Research, 66(3), 361 -3 96.

Grissmer, D., & Kirby, S. N. (1997). Teacher Turnover and Teacher Quality. Teachers
College Record, 99(1), 45-56.

Guin, K. (2004). Chronic Teacher Turnover in Urban Elementary Schools. Education
Policy Analysis Archives, 12(42), 1-30.

Hansen, P., & Mulholland, J. A. (2005). Caring and Elementary Teaching: The Concens
of Male Beginning Teachers. Journal of Teacher Education, 56(2), 119-131.

276

Hanson, S., & Pratt, G. (1992). Dynamic Dependencies: A Geographic Investigation of
Local Labor Markets. Economic Geography, 68(4), 373-405.

Hanushek, E. A., Kain, J. F ., & Rivkin, S. G. (1999). Do Higher Salaries Buy Better
Teachers? Cambridge, MA: National Bureau of Economic Research, NBER
Working Paper 7082.

Hanushek, E. A., Kain, J. F., & Rivkin, S. G. (2004). Why Public Schools Lose Teachers.
Journal of Human Resources, 39, 326-354.

Harris, D., Rutledge, S., Ingle, W. K., & Thompson, C. (2006). When Supply Meets
Demand: Principal Preferences and Teacher Hiring. Paper presented at the
Annual Meeting of the American Educational Research Association, San
Francisco, CA.

Henke, R. R., Chen, X., Geis, S., & Knepper, P. (2000). Progress Through the Teacher
Pipeline: 1 992-93 College Graduates and Elementary/Secondary School
Teaching as of 1 99 7, NCES 2000-I52. Washington, DC: US. Department of
Education, National Center for Education Statistics.

Horng, E. L. (2005). Teacher Tradeoffs: Retaining teachers at hard-to-stajf schools by
examining their preferences for workplace characteristics. Paper presented at the
Annual Meeting of the American Educational Research Association, Montreal,
Canada.

Ingersoll, R. M. (2001). Teacher Turnover and Teacher Shortages: An Organizational
Analysis. American Educational Research Journal, 38(3), 499-534.

Johnson, S. M., & Birkland, S. E. (2003). Pursuing a "Sense of success": New Teachers
Explain Their Career Decisions. American Educational Research Journal, 40(3),
581-618.

Johnson, S. M., Kardos, S. M., Kauffman, D., Liu, E., & Donaldson, M. (2004). The
support gap: New teachers’ early experiences in high income and low-income
schools. Retrieved June 10, 2005, from http://epaa.asu.edu/epaa/v12n61/

Johnson, S. M., & Landman, J. (2000). "Sometimes bureaucracy has its charms": The
working conditions of teachers in deregulated schools. Teachers College Record,
102(1), 85-124.

Kardos, S. M., Johnson, S. M., Peske, H. G., Kauffman, D., & Liu, E. (2001). Counting
on Colleagues: New teachers encounter the professional culture of their schools.
Educational Administration Quarterly, 37(2), 250-290.

Kirby, S. N., Berends, M., & Naftel, S. (1999). Supply and Demand of Minority Teachers
in Texas: Problems and Prospects. Educational Evaluation and Policy Analysis,
21(1), 47-66.

277

Kirby, S. N., & Grissmer, D. (1993). Teacher Attrition: Theory, Evidence, and Suggested
Policy Options. Santa Monica, CA: RAND.

Kirby, S. N., & Hudson, L. (1993). Black Teachers in Indiana: A Potential Shortage?
Educational Evaluation and Policy Analysis, 15(2), 181-194.

Knapp, M. S., Elfers, A. M., Plecki, M. L., Loeb, H., & Zahir, A. (2005). Teachers
Count: Support for T eachers' Work in the Context of State Reform. Seattle, WA:
University of Washington.

Lamaison, P., & Bourdieu, P. (1986). From Rules to Strategies: An Interview with Pierre
Bourdieu. Cultural Anthropology, 1(1), 110-120.

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

Lankford, H., & Wyckoff, J. (2005). Why Are Schools Racially Segregated? Implications
for School Choice. In J. T. Scott (Ed.), School Choice and Diversity: What the
Evidence Says (pp. 9-26). New York: Teachers College Press.

Lareau, A. (1987). Social-Class Differences in Family-School Relationships: The
Importance of Cultural Capital. Sociology of Education, 60, 73-85.

Lareau, A., & Horvat, E. M. (1999). Moments of social inclusion and exclusion: Race,
class, and cultural capital in family-school relationships. Sociology of Education,
72(January), 37-53.

Levin, J ., & Quinn, M. (2003). Missed Opportunities: How We Keep High-Quality
Teachers Out of Urban Classrooms: The New Teacher Project.

Levinson, A. M. (1988). Reexarnining teacher preferences and compensating wages.
Economics of Education Review, 7(3), 357-364.

Lipman, P. (1998). Race, Class, and Power in School Restructuring. Albany, NY: State
University of New York Press.

Lin, B, & Johnson, S. M. (2003). New teachers' experiences of hiring: Late, rushed, and
information-poor. Cambridge, MA: Project on the Next Generation of Teachers.

Liu, E., & Johnson, S. M. (2006). New teachers' experiences of hiring: Late, rushed, and
information-poor. Educational Administration Quarterly, 42, 324-360.

Liu, E., Johnson, S. M., & Peske, H. G. (2004). New teachers and the Massachusetts
signing bonus: The limits of inducements. Educational Evaluation and Policy
Analysis, 26(3), 217-236.

278

Loeb, S., Darling-Hammond, L., & Luczak, J. M. (2005). How teaching conditions
predict teacher turnover in California schools. Peabody Journal of Education,
80(3), 44-70.

Lortie, D. C. (1975). Schoolteacher : A sociological study. Chicago: University of
Chicago Press.

MacLeod, J. (1995). Ain't no makin' it. Boulder: Westview Press.

Maxwell, J. A. (1992). Understanding and Validity in Qualitative Research. Harvard
Educational Review, 62(3), 279-300.

McDerrnott, R. P. (1974). Achieving school failure: An anthropological approach to
illiteracy and social stratiﬁcation. In G. Spindler (Ed.), Education and cultural
process (pp. 82-118). New York: Holt, Rinehart and Winston.

Miles, M. B., & Huberman, A. M. (1994). Qualitative Data Analysis: An Expanded
Sourcebook. Second Edition. Thousand Oaks, CA: Sage Publications.

Mont, D., & Rees, D. I. (1996). The inﬂuence of classroom characteristics on high school
teacher turnover. Economic Inquiry, 34(1), 152-167.

Mumane, R. J ., & Olsen, R. J. (1989). The Effects of Salaries and Opportunity Costs on
Length of Stay in Teaching: Evidence from Michigan. The Review of Economics
and Statistics, 71(2), 347-352.

Mumane, 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, 25(1), 106-124.

Mumane, R. J., Singer, J. D., Willett, J. B., Kemple, J. J., & Olsen, R. J. (1991). Who Will
Teach? Policies That Matter. Cambridge, MA: Harvard University Press.

Nakao, K., & Treas, J. (1992). The 1989 Socioeconomic Index of Occupations:
Construction from the 1989 Occupational Prestige Scores (N o. GSS
Methodological Report 74). Chicago, IL: National Opinion Research Center.

Nias, J. (1999). Primary Teaching as a Culture of Care. In J. Prosser (Ed.), School
Culture. London: Paul Chapman.

Place, A. W., & Kowalski, T. J. (1993). Principal Ratings of Criteria Associated with
Teacher Selection. Journal of Personnel Evaluation in Education, 7(291-300).

Podgursky, M., Monroe, R., & Watson, D. (2006). The Academic Quality of Public
School Teachers: An Analysis of Entry and Exit Behavior. Economics of
Education Review, 23, 507-518.

279

Ralph, E. G., Kesten, C., Lang, H., & Smith, D. (1998). Hiring New Teachers: What Do
School Districts Look For? Journal of Teacher Education, 49(1), 47-56.

Rivkin, S. G., Hanushek, E. A., & Kain, J. F. (2005). Teachers, Schools, and Academic
Achievement. Econometrica, 73(2), 471.

Rosen, S. (1986). The Theory of Equalizing Differences. In 0. Ashenfelter & R. Layard
(Eds), Handbook of Labor Economics (Vol. I, pp. 641-692). Amsterdam: Elsevier
Science Publishers.

Rynes, S. L., & Lawlor, J. (1983). A Policy-Capturing Investigation of the Role of
Expectancies in Decisions top Pursue Job Alternatives. Journal of Applied
Psychology, 68(4), 621-631.

Sanders, W. L., & Rivers, J. C. (1996). Cumulative and residual eﬂects of teachers on
future student academic achievement. Knoxville, TN: University of Tennessee
Value-Added Research and Assessment Center.

Scaﬁdi, B., Sjoquist, D. L., & Stinebrickner, T. R. (2007). Race, poverty, and teacher
mobility. Economics of Education Review, 26(2), 145-159.

Schultz, T. (1977). Investments in human capital. In J. Karabel & A. H. Halsey (Eds),
Power and ideology in education (pp. 313-324). New York: Oxford University
Press.

Smith, T. M., & Ingersoll, R. M. (2004). What are the effects of induction and mentoring
on beginning teacher turnover? American Educational Research Journal, 41 (3),
681-714.

Soelberg, P. O. (1967). Unprogrammed Decision Making. Industrial Management
Review, 8(2), 19- 29.

Stevens, C. K. (1998). Image Theory and Career-Related Decisions: Finding and
Selecting Occupations and Jobs. In L. R. Beach (Ed.), Image Theory: Theoretical

and Empirical Foundations (pp. 227-239). Mahwah, NJ: Lawrence Erlbaum
Associates.

Stinebrickner, T. R. (2001). A Dynamic Model of Teacher Labor Supply. Journal of
Labor Economics, 19(1), 196-330.

Stockard, J ., & Lehman, M. B. (2004). Inﬂuences on the Satisfaction and Retention of
lst-Year Teachers: The Importance of Effective School Management.
Educational Administration Quarterly, 40(5), 742-771.

Strunk, K. O., & Robinson, J. P. (2006). Oh, Won't You Stay: A Multilevel Analysis of
the Difﬁculties in Retaining Qualiﬁed Teachers. Peabody Journal of Education,
81(4), 65-94.

280

Theobald, N. D. (1990). An Examination of the Inﬂuence of Personal, Professional, and
School District Characteristics on Public School Teacher Retention. Economics of
Education Review, 9(3), 241-250.

Toder, E. J. (1972). The supply of public school teachers to an urban metropolitan area:
A possible source of discrimination in education. The Review of Economics and
Statistics, 54(4), 439-443.

Trimble, S. (2001). The Teacher Applicant Pool. Principal Leadership, 44-47.

US. Department of Education. (2000). Teacher Supply in the United States: Sources of
Newly Hired Teachers in Public and Private Schools, (N o. NCES 2000-309, by
Stephen P. Broughman and Mary R. Rollefson). Washington, DC: National
Center for Education Statistics.

Vogt, F. (2002). A Caring Teacher: Explorations into Primary School Teachers'
Professional Identity and Ethic of Care. Gender and Education, 14(3), 251-264.

Walden, M. L., & Sogutlu, Z. (2001). Determinants of Intrastate Variation in Teacher
Salaries. Economics of Education Review, 20(4), 63-70.

Weiss, C. H. (1995). The four "I's" of school reform: How interests, ideology,
information, and institution affect teachers and principals. Harvard Educational
Review, 65(4), 571-592.

Winter, P. A. (1996). Applicant Evaluations of Formal Position Advertisements: The
Inﬂuence of Sex, Job Message Content, and Information Order. Journal of
Personnel Evaluation in Education, 10, 105-116.

Winter, P. A., & Melloy, S. H. (2005). Teacher Recruitment in a School Reform State:
Factors that Inﬂuence Applicant Attraction to Teaching Vacancies. Educational
Administration Quarterly, 41(2), 349-372.

Wise, A. E., Darling-Hammond, L., & Berry, B. T. (1987). Effective Teacher Selection:
From Recruitment to Retention. Santa Monica: RAND Corporation.

Young, 1. P. (1999). Salary Discrimination: A Test of the Paradoxical Female
Hypothesis. Educational Administration Quarterly, 35(3), 379-397.

Young, 1. P., & Delli, D. A. (2002). The Validity of the Teacher Perceiver Interview for
Predicting Performance of Classroom Teachers. Educational Administration
Quarterly, 38(5), 586-612.

Young, 1. P., & Fox, J. A. (2002). Asian, Hispanic, and Native American Job Candidates:
Prescreened or Screened Within the Selection Process. Educational
Administration Quarterly, 38(4), 530-554.

281

Young, 1. P., Place, A. W., Rinehart, J. 8., Jury, J. C., & Baits, D. F. (1997). Teacher

recruitment: A test of the similarity-attraction hypothesis for race and sex.
Educational Administration Quarterly, 33(1), 86.

Young, 1. P., Rinehart, J. S., & Heneman, H. G. (1993). Effects of Job Attribute
Categories, Applicant Job Experience, and Recruiter Sex on Applicant Job
Attractiveness Ratings. Journal of Personnel Evaluation in Education, 7, 55-66.

Young, 1. P., Rinehart, J. S., & Place, A. W. (1989). Theories for teacher selection:
Objective, subjective, and critical contact. Teaching and Teacher Education, 5,
329-336.

Zajonc, R. B. (1968). The Attitudinal Effects of Mere Exposure. Journal of Personality
and Social Psychology, 9, 1-27.

282

 

   

' 1011111111111111