Handbook of the Sociology of Education in the 21st Century

1  Family, Schooling, and Cultural Capital 17

Montgomery County, Maryland:  The report card effort/motivation/persistence.” In Richland they
has separate sections for grading each of the aca- include that the student “understands effort and
demic subjects, plus one for grading what are perseverance directly impact learning, listens
called Learning Skills. This is divided into two attentively in different learning situations, and
sets of items. The first, called Work Habits, con- manages materials and time.” What these have in
tains the following: common is that they all describe aspects of good
academic work habits. They are the traits needed
• Rules and Procedures to be academically successful while not reducing
• Task Completion the success of the other students in the class.
These are the behaviors that teachers are most
The second, called Thinking and Academic focused on rewarding, not knowledge of classi-
Success Skills, contains the following: cal music or fine arts. Teacher “gatekeeping”
rewards effective and cooperative10 academic
• Analysis work habits, and punishes their opposite—low
• Collaboration effort, poor organization, inattention, sloppiness,
• Effort/Motivation/Persistence disrespect, and disruptiveness. A quick perusal
• Fluency of a larger number of district report card formats
• Intellectual Risk Taking available online suggests that teacher judgment
• Metacognition of these aspects of students’ academic work hab-
• Originality its is widespread.11
• Synthesis

Richland, Washington:  The report card, in addi- 1.4.2 P utting It All Together
tion to grades for the separate academic subjects,
also has grades for what are called Social and A focus on academic skills and work habits was
Learning Skills. These are the following: the basis for the empirical study of cultural capi-
tal undertaken by Farkas and colleagues more
• Engages effectively with others than 25  years ago (Farkas et  al. 1990; Farkas
• Understands effort and perseverance directly 1996). In this work, a representative sample of
Dallas Independent School District (DISD) 7th
impact learning and 8th grade social studies teachers responded
• Listens attentively in different learning to a “student work-ethic characteristics question-
naire” regarding up to six of their students
situations selected by stratified random sampling. The
• Respects individual differences/rights of teachers rated the students on homework, class
participation, effort, organization, disruptiveness,
others assertiveness, and appearance and dress. The first
• Takes responsibility for choices and actions four of these had correlations between 0.80 and
• Manages materials and time 0.95, and were combined into a scale of work
• Advocates for self habits. One of the variables—assertiveness—
showed little relationship with the other (inde-
All of these districts give grades in each of the pendent or dependent) variables and was omitted
academic subjects. But what sets of behaviors, from the study. A student’s days absent as
explicitly identified for grading, do these dis-
tricts have in common? The answer is—habitual 10 But note that Richland also judges whether the student
behaviors that facilitate learning in the American “advocates for self.”
classroom. In Sarasota these include “works 11 And these teachers’ values likely benefit females more
well in class, is courteous, respectful and coop- than males. See Dumais (2002), Morris (2008).
erative; works independently, without disturbing
others.” In Montgomery County these include
“rules and procedures, task completion, and

18 G. Farkas

recorded by the district was also included as a cultural capital theory, in which the student’s
behavioral variable, as were disruptiveness and habitus, strongly influenced by parents and peers
appearance and dress. Basic skills were measured in the home and neighborhood, and by the child’s
by student scores on the Iowa Test of Basic Skills preschool experiences before kindergarten entry,
(ITBS), which includes both Language and then evolves via the student’s interaction with
Mathematics totals, as well as subskill scores for family, peers, and teachers as the student moves
each of these variables. Farkas and colleagues up the grade-levels?
operationalized student skills, habits, and styles
as the student’s ITBS score, work habits, days Farkas (2003) reviewed the literature on cog-
absent, disruptiveness, and appearance and dress. nitive and noncognitive skills developed by econ-
omists and sociologists and related it to the
This research was also able to profit from an “skills, habits, and styles” version of cultural
unusual initiative undertaken by the DISD in capital theory discussed above. Economists’
response to the Texas Education Reform Act of research in this area can be traced back to the
1984. Groups of teachers in each of the subject- work of Bowles and Gintis (1976), whereas
matter areas were assembled over the summer to related work by sociologists dates from the book
create test items representative of the course sub- by Jencks and colleagues (Jencks et al. 1979).
ject matter. These curriculum-referenced tests
were then administered uniformly to DISD stu- Bowles and Gintis argued that “in capitalist
dents at the end of the appropriate semester. The America,” variation in the design and manage-
resulting scores provide an objective measure of ment of schools exists to create those worker per-
each student’s coursework mastery in the subject. sonality traits needed by different jobs in the
industrial system, largely based on the jobs held
The authors then estimated a causal model in by the student’s parents, thereby leading to social
which student and teacher sociodemographics are reproduction. Thus, the children of working-class
regarded as determining the student’s basic skills parents typically obtained no more than a high
and the teacher’s judgment of the student’s habits school degree, perhaps with an emphasis on
and styles, and these in turn are related to the stu- vocational training, and became factory workers
dent’s actual coursework mastery. All of these whose obedience to authority was their most
variables together are then related to the teacher- desired trait. Accordingly, such obedience was
assigned course grade. This model is summarized emphasized by K–12 teachers. By comparison,
in Fig. 1.1. It shows the key relationships involved the children of middle- and upper-class parents
as students from different social backgrounds went on to college, where creativity and indepen-
interact with teachers from different social back- dence received greater rewards, since these are
grounds, resulting in the teacher-­gatekeeper’s final the skills needed for middle-class management
judgment on the student for the semester—the and professional employment.
course grade. This is the closest that empirical
research has come to implementing a quantitative To provide evidence for these assertions,
and testable version of Lareau and Weininger’s Bowles and Gintis empirically tested their asser-
(2003) suggestion that cultural capital studies tion that the personality trait they labeled “sub-
focus on the interaction of students with their mission to authority” was, along with cognitive
teacher-gatekeepers, and how this interaction skills, the principal determinant of course grades
results in different schooling outcomes for stu- in high school. Their empirical work supported
dents from different social backgrounds. this assertion, but crucially, they defined such
submission as including the following character-
I will defer discussion of the empirical find- istics of a student’s academic work habits: perse-
ings from this work until the following section, verance, dependability, consistency, identifies
where the detailed findings from prior empirical with school, empathizes orders, punctuality, and
work are reviewed. However, the question arises, defers gratification. As we shall see throughout
what has been done since this work by Farkas and this review, these are indeed the habits and behav-
colleagues to implement and test this version of iors graded positively by K–12 teachers.

1  Family, Schooling, and Cultural Capital 19

However, for most teachers and many other social class differences are replicated, says
researchers, myself included, these traits do not Lareau, when parents interact with teachers. In
deserve the pejorative label “submission to such situations she describes working-class and
authority.” Instead, they simply constitute “good poor parents as “baffled, intimidated, and
work habits” whose effects are to be measured subdued.”
empirically, and which may be desirable at all
levels of the occupational structure. Other sociologists have undertaken related
analyses, both quantitative and qualitative, seek-
This is the approach taken by Jencks et  al. ing to discover which parent and student behav-
(1979), who conducted extensive analyses of the iors are most strongly associated with student
roles played by individual cognitive skills and success. At the same time, economists have pro-
non-cognitive (personality) traits on school and duced a quantitative literature on the effects of
employment success. Using multiple data sets they cognitive and non-cognitive skills on school and
measured the effects of self-assessed personality employment success. Prominent here is a paper
traits as well as what they considered to be indirect by Heckman and Kautz (2014) seeking to esti-
personality measures involving self-r­eports of var- mate the empirical importance of cognitive skills
ious behaviors possibly reflecting underlying per- and non-cognitive traits in determining schooling
sonality. A principle components analysis of 14 outcomes. Findings from these literatures will be
questions identified a construct they referred to as reviewed in the following section.
“study habits.” They also analyzed data in which
teachers rated students on each of nine personality To summarize, the “skills, habits, and styles”
traits. Results of these analyses are summarized in paradigm has been widely used to investigate
the following section. how the actions of parents, children, and teachers
lead to the differential school success of children
Other researchers continued the analysis of from middle- and upper-class children, compared
the effects of cognitive and noncognitive skills on to those from the working class. It seems to rea-
school success. Within sociology, Lareau (2011) sonably capture Bourdieu’s intentions for the
echoed the distinction between working-class habitus (underlying) and cultural capital (enacted)
and middle-class parenting orientations dis- concepts to serve as mediators between family
cussed by Bowles and Gintis, referring to the background and schooling success. Indeed, after
working-class style as “the accomplishment of the dominance of this research area by cultural
natural growth” and the middle-class style as sociologists focused on elite cultural activities,
“concerted cultivation.” She repeats the Bowles this research approach brings back an emphasis
and Gintis observation that working-class parents on the daily actions and interactions involving
tend to want their children to follow directives, students and teachers that ultimately determine
while middle-class parents tend to encourage the schooling and social class attainment of the
their children to ask questions and to reason. students. It also brings back the concern with
Rather than emphasizing the social class differ- finding a sociological equivalent of the human
ences in academic work habits likely resulting capital paradigm advanced by economists, and
from these parenting differences, Lareau instead employed so successfully to apply economic rea-
emphasized that the middle-class parenting style soning to almost every field of human endeavor.
teaches the child to develop an individualized Both James Coleman and Pierre Bourdieu were
sense of self, including a sense of comfort, enti- explicitly in interaction with economists, and
tlement, and agency when dealing with adult were inspired to create their formulations by the
organizations such as the school, where they world-wide success of the human capital para-
learn to present themselves and perform (Lareau digm. Bringing this research area back to a place
2011, pp. 242–243). Lareau asserts that, by con- where economists and sociologists speak to one
trast, the working-class parenting style leaves another, and empirically test their theories, sim-
children feeling uncomfortable and constrained ply puts this research area back on a developmen-
when dealing with these same institutions. These tal trajectory consistent with its beginning.

20 G. Farkas

1.5 E mpirical Findings 1.5.1 S ocial Class Differences
in Parenting and Their
A schematic model of cultural capital’s causal Consequences
effects was presented in Fig. 1.1. This is a media-
tion model, in which parenting, habitus, and aca- Duncan and Magnuson (2011, Fig. 3.1) provide a
demic skills and habits mediate the relationship schematic model of how genes, families, schools,
between SES and course grades. The SES of each and peer groups combine to determine the trajec-
student implies the parenting they receive. This tories of children’s cognitive skills and behaviors
parenting helps determine the student’s habitus, from birth to grade 12, which in turn determine
his/her disposition (including skills) toward vari- the individual’s subsequent educational and labor
ous behaviors and strategies of action. These dis- market attainment. For a variable to play a role in
positions then lead to the academic skills and creating social class differences in children’s
work habits that the student presents to the school success, two conditions must be met.
teacher in the classroom. These skills and habits First, it must significantly differ across social
are then employed by the teacher to assign a class groupings. And second, it must significantly
course grade to the student. Where quantitative affect schooling outcomes, such that when it is
empirical work is concerned, researchers are able controlled, the relationship between parental
to find measures of SES, parenting, academic social class and student success in school is
skills, work habits, and course grades (or teach- reduced or eliminated. In this section we examine
ers’ judgements of students’ skills) on many of empirical tests of the extent to which parenting
the large, nationally representative data sets col- meets these conditions.
lected by the National Center for Education
Statistics and that are widely available to 1.5.1.1 Measuring Parenting:
researchers (these include the ECLS-K, the The HOME Score
ECLS: 2011, the NELS, and ELS). Other data
sets, including the 28-nation PISA, have also That working-class parents have different parent-
been used in empirical studies. ing styles from middle- and upper-class parents is
a perennial finding of sociologists, psychologists,
The habitus, conceived as a collection of and economists. These differences have been
underlying dispositions, including skills, habits, conceptualized and measured in a number of
identities, worldviews, preferences, or values, ways.
can typically not be measured directly, so that its
characteristics are inferred by the academic skills Particularly widely used is the Home
and habits it gives rise to. (However, as we shall Observation for Measurement of the Environment
see, Gaddis (2013) seeks to measure it by using (HOME). Separate versions of this measurement
two attitudinal scales.) Thus, empirical work has instrument have been created to measure parent-
typically included some subset (or all) of the ing quality for children of different ages, but all
variables SES, parenting, academic skills and versions are similarly structured. As modified for
work habits, and course grades shown in Fig. 1.1. use in the National Longitudinal Survey of Youth
The result has been empirical studies in which (NLSY), the HOME produces two parenting
parenting is regressed on SES, skills and work measures—one for cognitive stimulation and the
habits are regressed on SES and parenting, and other for emotional support. It is useful to
course grades are regressed on some or all of ­examine the behavioral items typically included
SES, parenting, and skills and work habits.
Empirical studies of these types are the ones priate to use teacher-assigned course grades, because only
reviewed here.12 these represent the teacher-gatekeeper judgments that are
so central to cultural capital theory. (Of course standard-
12 A subset of studies use standardized test scores as their ized test scores should be one of the predictors of the
ultimate outcome measures. But it would be more appro- teacher-assigned course grade, since test scores measure
the academic knowledge and skills that the student dis-
plays to the teacher.)

1  Family, Schooling, and Cultural Capital 21

in these scales in order to understand which –– Mother caressed, kissed, or hugged child at
parental behaviors researchers consider most least once?
important for children’s development. To take
one example, for children aged 3–5, the follow- –– Mother introduced interviewer to child by name?
ing items are used to measure parental cognitive –– Mother physically restricted or (shook/
stimulation and emotional support:
grabbed) child? Coded non-supportive
Cognitive Stimulation Scale: –– Mother slapped or spanked child at least once?

–– How often read stories to child? Coded non-supportive
–– How many children’s books does child have? –– Mother’s voice conveyed positive feeling
–– How many magazines family gets regularly?
–– Child has use of CD player? about child?
–– Do you help child with numbers?
–– Do you help child with alphabet? We see that the cognitive stimulation scale
–– Do you help child with colors? is focused on direct parental instruction and
–– Do you help with shapes and sizes? the materials useful for learning. That is, this
–– How often is child taken on any kind of outing? scale emphasizes parental actions that foster
–– How often is child taken to museum? cognitive readiness for school. The emotional
–– Child’s play environment is safe? responsiveness scale focuses on warm, positive
–– Interior of the home is dark or perceptually parent–child interaction, and gives a lower
score when the parent employs physical pun-
monotonous? ishment. The elements of the HOME score
–– All visible rooms of the home are reasonably listed above encompass many of the items that
quantitative studies have used to measure par-
clean? enting. However, some studies, particularly
–– All visible rooms of the home are minimally those associated with the original notion of
cultural capital defined as knowledge of and
cluttered? participation in high (elite) cultural activities
(e.g., classical music and museum quality art)
Emotional Support Scale: advanced by DiMaggio (1982) and of “con-
certed cultivation” (e.g., scheduled activities
–– If child got so angry that s/he hit you, what including sports, music and dance classes)
would you do? Respondent is offered multiple advanced by Lareau (2011) have added or sub-
responses. If either “hit him/her back” or stituted these activities for the items in the
“spank child,” item is scored “not emotionally HOME above.
supportive.”
1.5.1.2 Social Class Differences
–– How much choice is child allowed in deciding in HOME Parenting Measures
foods s/he eats at breakfast & lunch?
Reeves and Howard (2003) used longitudinal
–– About how many hours is the TV on in your HOME scores from the Children of the NLSY to
home each day? >4 is scored “not emotionally create measures of “strong versus weak parent-
supportive.” ing.” That is, for each child they measured
whether the HOME score was in the bottom or
–– How often does child eat a meal with you and top 25% of parents at each of the three stages—
his/her father/stepfather/father-figure? infancy (age 0–2), early childhood (age 3–5), and
middle childhood (age 10–15). Parents scoring in
–– About how many times, if any, have you had the bottom 25% during at least two of these
to spank child in the past week? >1 is scored stages were considered to be the weakest parents;
“not emotionally supportive.” those scoring in the top 25% during at least two
of these stages were considered to be the stron-
Interviewer observed: gest parents. (This resulted in 20.9% of parents
being categorized as weakest and 17.6% as
–– Mother conversed w/child >1 time (no scold-
ing or suspicious comments)?

22 G. Farkas

­strongest.) The researchers then computed the these skills and habits when children enter
percent of each type of parent among families in kindergarten.13
either the bottom or the top quintile on family
income. They found that, for families in the bot- 1.5.1.4 Concerted Cultivation
tom income quintile, almost 50% were among In a widely discussed study, Lareau (2011) focused
the weakest parents whereas fewer than 5% were on a somewhat different set of parenting behaviors
among the strongest parents. By contrast, for on which working-class and middle-class parents
families in the top income quintile, about 35% differ. These are the formalized out-o­f-h­ome
were among the strongest parents, whereas only activities that middle-class parents typically
about 5% were among the weakest. Thus, parent- schedule for their children, contrasted with the
ing quality as measured by the HOME scale var- more around the home and neighborhood, self-
ies strongly and significantly across social organized activities of working-class children.
classes. But to what extent do these social class Lareau referred to the latter as “the accomplish-
differences in parenting quality account for social ment of natural growth” and the former as “con-
class differences in children’s cognitive and certed cultivation.” As described by Lareau (2011,
behavioral outcomes? pp. 238–239), in middle-class families

1.5.1.3 HOME Parenting Affects parents actively fostered and assessed their chil-
Cognitive and Behavioral dren’s talents, opinions, and skills. They scheduled
Outcomes their children for activities. They reasoned with
them. They hovered over them and outside the
This question has been addressed by a number of home they did not hesitate to intervene on the chil-
empirical studies. Morgan et al. (2009) replicated dren’s behalf. They made a deliberate and sus-
the findings of Reeves and Howard, reporting tained effort to stimulate children’s development
that mothers in the lowest educational quintile and cultivate their cognitive and social skills.
displayed HOME scores approximately one stan-
dard deviation lower than those in the highest By contrast, Lareau says,
educational quintile. They also found that these
parenting scores significantly affected children’s working-class and poor parents viewed children’s
learning-related behaviors, and explained a sig- development as unfolding spontaneously, as long
nificant portion of the social class differences in as they were provided with comfort, food, shelter,
these behaviors. Hoff (2003) followed up on and other basic support…Parents who relied on
work by Hart and Risley (1995), showing that natural growth generally organized their children’s
social class differences in mothers’ speech to lives so they spent time in and around home, in
their 2-year-olds fully explained social class dif- informal play with peers, siblings, and cousins…
ferences in these children’s vocabularies. Farkas Instead of the relentless focus on reasoning and
and Beron (2004) found that parenting measures negotiation that took place in middle-class fami-
partially explained social class differences in the lies, there was less speech (including less whining
oral language skills of children. Bradley et  al. and badgering) in working-class and poor homes…
(2001) showed the significant effects of HOME Directives were common. In their institutional
parenting scores on children’s cognitive and encounters, working-class and poor parents turned
behavioral development. Smith et  al. (2006) over responsibility to professionals; when they did
showed that maternal responsiveness to the child try to intervene, they felt they were less capable
positively affected cognitive development. In and less efficacious than they would have liked.
sum, the cognitive stimulation and emotional
support activities measured by the HOME are Lareau’s mention of middle-class parents
significantly and positively associated with the actively fostering their children’s “talents, opin-
skills and habits of children, and explain a por- ions, and skills” is reminiscent of Swidler’s
tion, but not all, of the social class differences in
13 There is a large literature on parental involvement with
their child’s school work, teacher, and school activities
more generally, and how this involvement is related to stu-
dent achievement. For examples, see Van Voorhis et  al.
(2013) and Nunez et al. (2015).

1  Family, Schooling, and Cultural Capital 23

“skills, habits, and styles.” Yet in Lareau’s discus- effects for student academic work habits and

sion of the consequences of these social class dif- prior achievement. Sticking relatively closely

ferences in parenting, she emphasizes the with Lareau’s definition of concerted cultivation,

organized activities that middle-class children Dumais et  al. (2012) found no positive signifi-

experience—for example, sports and summer cant relationship between (a) parents’ cultural

camps—and the way these help the child to activities with their child and/or parents’ school

develop an “individualized sense of self.” She involvement and (b) teachers’ evaluations of

goes on to describe these experiences as assisting students’ language and literacy skills, academic

middle-class children to develop a sense of enti- work habits, or interpersonal skills. Similar

tlement and agency when dealing with adults and results were reported by De Graaf et al. (2000).

their institutions, such as teachers and schools. They used both elite cultural activities and

By contrast, she says, the working-class child parental reading to their children to predict the

rearing style does not foster such a sense of self child’s ultimate educational attainment. They

(Lareau 2011, pp. 241–43). Lareau’s emphasis on found that reading to the child, but not elite cul-

scheduled activities and the development of a tural activities, significantly predicted educa-

sense of entitlement in middle-class children tional attainment.

tends to de-emphasize the importance of those Bodovski and Farkas (2008) used ECLS-K

direct, academic skill building activities that data for first grade to estimate the association

m­ iddle-class parents also devote time to fostering between both social class and parenting quality

(although she does mention language use as a key (with an emphasis on the concerted cultivation

component of concerted cultivation). While it is parenting style) on the one hand and students’

no doubt true that middle- and upper-class­ academic work habits, academic performance,

parents provide their children with both a sense and the teacher’s judgment of the student’s per-

of entitlement and agency and with the concrete formance on the other. The authors employed a

skills and behaviors needed to succeed in school, more general definition of concerted cultivation

it is important to know which of these plays the that added parental instructional and interac-

larger role in the greater school success of tional activities to the measures of participation

­middle-class students compared with those from in organized activities and parental involvement

the working class. Thus, although the report cards with the schools. The result was three dimen-

I sampled emphasized academic work habits, at sions of parental activities for first graders, mea-

least one, from Richland WA, included an item sured in three separate scales and then combined

about the student’s agency, namely “advocates into a single scale. The first dimension is paren-

effectively for self.” tal perceptions of their responsibilities towards

1.5.1.5 Determinants their child, with a particular focus on instruction

and Consequences and interaction. The following variables were

of Concerted Cultivation used to construct this scale: tell a child stories,

sing songs, do art, play games, teach about

Quantitative studies of the determinants and nature, build blocks, do sports, practice numbers

consequences of concerted cultivation have and letters, read to a child, listen to a child even

yielded mixed results. Roscigno and Ainsworth- if busy, foster the child’s opinion, help with

Darnell (1999) found a relatively strong positive homework.

relationship between SES and each of cultural The second dimension is how children spend

trips, cultural classes, and household educational their leisure time, particularly their participation

resources. However, when they employed these in organized activities. These were measured as

parenting variables to predict course grades, music, arts and crafts, dance lessons, clubs, orga-

either with or without controlling prior grades nized performing arts and athletic activities, edu-

and test scores, they found insignificant or small cational trips to the library, museum, zoo, concert,

effects. By contrast, they found much larger or live show.

24 G. Farkas

The third dimension was conceptualized as ficient of SES on reading test scores was 0.31.
parents’ relationships with social institutions, Adding parental expectations and concerted cul-
particularly schools. This was measured as par- tivation reduced this by 26% to 0.23, showing
ticipation in parent–teacher conferences, attend- that concerted cultivation can explain at most a
ing an open house or back-to-school night, portion of SES differentials in cognitive perfor-
participating in PTA, attending a school event, mance. The direct effect of concerted cultivation
volunteering at school, and participating in fund- on reading test scores was 0.09. Finally, aca-
raising. The authors also added another vari- demic work habits were added to the equation.
able—number of children’s books in the This reduced the SES effect to 0.18, slightly more
home—providing an additional measure of than half of its total effect. The direct effect of
parental efforts to enrich their children’s lives and academic work habits on reading test scores was
understanding, as well as assist with pre-reading a very substantial 0.38, showing once again that
and reading skills. these behaviors appear to strongly affect
learning.
Bodovski and Farkas restricted their analysis
sample to White children in order to avoid con- Finally, these variables were used in sequen-
troversies regarding whether or not race func- tial regressions to predict the teacher’s judgment
tions as a stratifying factor in addition to of the student’s language and literacy skills. In
SES.  They first ran regressions using SES and the first regression, with only SES and demo-
other demographics to predict the concerted cul- graphics controlled, the total effect of SES was
tivation measure. They found a medium stan- 0.24. As the variables were added sequentially,
dardized coefficient of 0.40 for the path from by far the strongest predictors of the teacher’s
SES to concerted cultivation. This validates the judgment were academic work habits and read-
observations of Lareau and others regarding ing test scores. By the final regression, with all
strong social class differentials in the parenting predictors in the equation, the effect of the read-
activities measured by this variable. ing test score was 0.62, that of academic work
habits was 0.32, and the SES effect on the teach-
Next, Bodovski and Farkas used SES and con- er’s judgment of the student’s language and lit-
certed cultivation in sequential regressions to eracy skills had been fully explained. I conclude
predict the student’s teacher-judged academic that, at least in this nationally representative data
work habits—persistence at tasks, eagerness to set of first grade students, the teacher-assigned
learn, attentiveness, learning independence, flex- course grade is determined about 2/3 by actual
ibility, and organization. With only SES and performance and 1/3 by student work habits. This
demographics as predictors, the authors found gives a smaller role to work habits than was found
that SES had a standardized coefficient of 0.19 by Farkas (1996) for the Dallas schools (see
with academic work habits. When parental Fig. 1.2). However, this may be accounted for by
expectations for the child’s educational attain- differences in the subjects examined and the
ment and concerted cultivation were added to the available data. In particular, the 1996 Farkas
equation, the coefficient of SES declined 26% to study predicted the actual grade assigned for 7th
0.14; the direct effect of concerted cultivation and 8th grade social studies, whereas the 2008
was 0.07. This shows once again that direct mea- Bodovski and Farkas study predicted the teach-
sures of parenting activities are able to explain a er’s judgment of first grade student’s language
portion, but only a portion, of the effect of SES and literacy skills. The latter study likely showed
on the child’s academic work habits. a stronger effect of test scores since it was the
skills tested that the teacher was asked to judge.
Following this, Bodovski and Farkas used The fact that even in this case, with standardized
SES, demographics, parental educational expec- test scores controlled, student work habits had an
tations for the child, concerted cultivation, and effect size as large as 0.32 in predicting student
academic work habits in sequential regressions to skills demonstrates the importance of these work
predict the student’s reading test score. With only
demographics controlled, the standardized coef-

1  Family, Schooling, and Cultural Capital 25

habits in the teacher’s judgment of student cational topics with the child, helping with
performance. homework, managing the child’s time on literacy
and nonliteracy activities, and the parent’s educa-
Several additional studies have employed tional expectations for the child. (Note that
quantitative measures of concerted cultivation, Bodovski and Farkas included this last measure
typically testing for its role as a mediator in in their analyses, but did not consider it to be part
explaining the relationship between SES and of concerted cultivation.) The dependent variable
achievement measured by test scores, but without was academic achievement, measured as a com-
attention to either the academic work habits of posite including the teacher-assigned grades in
students or to teacher’s judgment of these and the reading and math as well as teacher reports of
role of this judgment in the assignment of a grade whether the child was above or below grade level
for the course. An example is Cheadle (2008), in reading and math. This use of grades and
who uses ECLS-K data to test the role of con- teacher judgments as outcomes puts the study
certed cultivation as a mediator between SES and more directly in the cultural capital field.
math and reading test score trajectories from kin-
dergarten through third grade. Cheadle uses The authors found a positive relationship
many of the same variables as Bodovski and between parental social class and concerted culti-
Farkas to measure concerted cultivation. These vation. Lee and Bowen also found that parental
comprised elite cultural activities, participation involvement at school and expectations for the
in school activities such as parent–teacher con- child’s educational attainment were positively
ferences, and the number of the child’s books, but associated with achievement, and partially medi-
omitted the direct instructional activities included ated the effect of social class on this outcome.
by Bodovski and Farkas, such as time spent read- These findings are generally consistent with
ing to a child or helping with homework. Cheadle those of other researchers. This study also found
finds that concerted cultivation explains about some significant interactions (moderation)
20% of the effect of SES on test scores. He also between elements of their measure of concerted
finds that concerted cultivation is most strongly cultivation and some of the demographic mea-
associated with race gaps in achievement at kin- sures. However, these did not follow any mean-
dergarten entry, and appears to play a smaller role ingful pattern.
in achievement growth as children move up to
first and third grade. Overall, the conclusion is Gaddis (2013) uses data from youth who par-
that the concerted cultivation parenting style ticipated in the Big Brothers/Big Sisters of
plays a modest role in mediating the effect of America program to test whether a measure of
SES on achievement. Cheadle might have found habitus mediates the relationship between a con-
larger effects if he had included direct instruc- certed cultivation parenting style and course
tional activities in his measure of concerted culti- grades. He operationalizes cultural capital using
vation. However, since this study employs test three measures of elite cultural participation plus
scores rather than course grades as the outcome, weekly hours spent reading. This paper is one of
it does not test for the determinants of teacher the few to claim to quantitatively measure habi-
judgments which are so central to cultural capital tus, which Gaddis does using two scales—a
theory. youth’s belief that she/he can succeed in school
and a scale measuring the youth’s belief that edu-
Other studies have used concerted cultivation cation is valuable to her/his success in life. Using
measures that partially overlap with those used first difference models, he first regresses change
by Bodovski/Farkas and Cheadle. Bodovski in grades on change in each of his four elements
(2010) found that, contrary to Lareau, even after of cultural capital (museum visits, play atten-
controlling SES, Black parents were less support- dance, cultural lessons, and time spent reading).
ive of their children’s school success than Whites. Two of these (museum visits and time spent read-
Lee and Bowen (2006) used measures of the par- ing) show significant positive effects on
ent physically visiting the school, discussing edu- GPA.  Second, he adds change in the habitus

26 G. Farkas

­variables (the two attitude measures) to the equa- with the a­ cademic work habits that teachers use
tion. They are both significantly associated with in determining course grades. Indeed, when
GPA, and with these variables controlled the assigning course grades, teachers had no knowl-
effects of the cultural capital variables become edge of the student’s scores on these attitude
smaller and lose significance. Gaddis concludes scales. Their only opportunity to observe differ-
that habitus mediates the effect of cultural capital ences in these attitudes across students was due
on GPA. He finds that museum visits and reading to their observation of the student’s academic
both have effect sizes of 0.05; the habitus attitude work habits.
variables both have effect sizes of 0.15. These are
small to modest in size. Comparing the way Gaddis operationalized
the cultural capital theory with the way it was
How can we compare Gaddis’ work where operationalized by Farkas (1996) and Bodovski
habitus is measured by two schooling attitude and Farkas (2008) is instructive. Gaddis opera-
scales with that of Farkas (1996) or Bodovski and tionalized the habitus with two attitudinal scales
Farkas (2008) where habitus is not explicitly closely related to the student’s positive feelings
measured, but academic work habits and test about her/his schoolwork, and used these as
scores measuring cultural capital are taken to be mediators between concerted cultivation and
the variables that teachers consider when assign- course grades. He did not use a measure of actual
ing course grades? Clarification is attained by student academic performance. Farkas (1996)
looking at the items comprising each of Gaddis’ did not seek to measure the habitus, which is the-
scales. The “I can succeed at school” scale may orized to be dispositions and skills internal to the
measure habitus, since it shows how the student student. Instead, he measured the academic work
sees herself in the school setting. But it is likely habits partially determined by the student’s habi-
also measuring the student’s actual success at tus, and estimated how the teacher-assigned
schoolwork. It is not surprising that positive course grade was affected by the student’s aca-
changes in school performance would be associ- demic performance (measured by both basic
ated with positive changes in the student’s reports skills and curriculum referenced tests) and the
of her school performance. However there is a student’s academic work habits. Similarly,
danger of reverse causality, where school perfor- Bodovski and Farkas (2008) did not attempt to
mance is driving attitudes rather than the other measure the habitus, but again tested the extent to
way around. which academic work habits and test score per-
formance affected the teacher’s assessment of the
As for the second scale, described by Gaddis student’s competency at the subject. They also
as a measure of “the youth’s belief that educa- tested the extent to which these work habits and
tion is valuable to her success in life,” it does test scores mediated the relationship between
contain items such as “How valuable do you concerted cultivation and the teacher’s judgment
think your education will be in getting the job of the student. Gaddis used many of the same
you want?” However, it also contains items such parenting variables used by others, but chose to
as the following: Do you think your school work refer to these as “cultural capital.” Bodovski and
is boring? Do you think your homework is fun to Farkas employed similar variables (although con-
do? Do you think the things you learn in school taining more about the parent’s direct instruction
are worthless? Do you care about doing your of the child) and, instead of viewing these as
best in school? How upset would you be if you measures of habitus, tested for the effects of work
got a low grade for one of your subjects? Change habits and test scores as mediators between par-
in these items could also be expected to be posi- enting and the teacher’s judgment of the child.
tively correlated with changes in grades, but The largest difference between the two research
once again, there may be reverse causality, where approaches is that Gaddis uses survey questions
positive change in grades leads to positive about attitudes toward school to measure habitus
change in these measures of feelings toward and tests for it as a mediator without controlling
school. Further, these items are likely correlated

1  Family, Schooling, and Cultural Capital 27

test scores. By contrast, Bodovski and Farkas use effect of mother’s education on reading test
academic work habits as expressions of the stu- scores, and about 18% of the mother’s education
dent’s cultural capital, and employ both work effect on math test scores. This is generally con-
habits and test scores as mediators. Since Gaddis’ sistent with prior work, although the use of test
survey questions appear to be closely related to scores rather than grades makes these results less
work habits, the most consequential difference of a true test of the cultural capital theory. It
between the two studies may be that Gaddis does appears that, in general, explicitly measured par-
not control test scores. enting activities of the type available on large
nationally representative data sets can explain
Using ECLS-K data, Bodovski (2014) opera- about 1/4 of the relationship between parental
tionalized students’ emerging habitus using 8th social class and student grades or test scores. This
grade students’ educational expectations, internal estimate is quite similar to the findings reported
locus of control, and general and area-specific by Bodovski and Farkas (2008) and Cheadle
self-concepts. She examined how early parental (2008).
practices and educational expectations (measured
during kindergarten and first-grade years) affect Tramonte and Willms (2010) take a similar
students’ emerging habitus and academic approach, but analyze PISA data containing
achievement when they reach adolescence (mea- information on more than 200,000 students
sured in eighth grade). The findings revealed that across 28 OECD countries. They operationalize
students from higher-SES families had more pos- cultural capital along two dimensions. They
itive general and area-specific self-concepts, measure “static cultural capital” by combining
higher educational expectations, internal locus of responses to nine questions about elite (“high-
control, and higher academic achievement. brow”) cultural activities. They measure “rela-
Higher parental educational expectations were tional cultural capital” by responses to six items
positively associated with all studied outcomes. concerning conversations between parents and
The findings provided only partial support for the the child covering topics such as social issues,
effects of early parental practices and highlighted books, films, television programs, how well the
the role of gender and race/ethnicity in shaping child is doing at school, as well as whether the
adolescents’ habitus. child herself enjoys talking with other people
about books or going to the bookstore or library.
Potter and Roksa (2013) also analyzed the The authors run regressions, separately for each
ECLS-K, emphasizing the over-time nature of country, estimating the effects of relational and
concerted cultivation, and the effects of contem- cultural capital on the student’s reading test
poraneous and cumulative concerted cultivation score and sense of belonging at school, control-
on student test scores in reading and math, esti- ling parental education, occupation, and sex.
mated with growth curve models. Their measure They find that both cultural capital measures
of concerted cultivation combines child activities are positively and significantly associated with
(e.g., dance, music, athletics), parental school reading test scores for each of the 28 countries,
involvement, parental educational expectations, with the relational measure association slightly
the number of books in the household, and stronger than that of the static measure for a
parent-­to-parent contact. They find that the moth- majority of the countries. The associations of
er’s education is positively associated with each these variables with sense of belonging is also
of these parenting behaviors, and that, with the generally positive, more consistently so for the
exception of parent-to-parent contact, cumulative relational cultural capital measure. However,
measures of each of these behaviors are posi- once again, this study used test scores rather
tively associated with increasing social class gaps than grades as the outcome. For a related study
in both reading and math test scores as children focused on the countries of Eastern Europe see
move up the grade levels. When entered as con- Bodovski et al. (2016).
trols, these behaviors explain about 23% of the

28 G. Farkas

1.5.2 Social Class Differences SES groups, so that the magnitude of the
in Cognitive Skills 36-month SES gap persists at least through to
and Academic Work Habits 13 years of age. As discussed earlier, large social
class gaps in cognitive performance are found at
Studies reviewed in the previous section focused kindergarten entry, and persist as children move
on the role of parenting as a mediator of the rela- up through the grades. These school readiness
tionship between SES and educational outcomes, and persistent social class differences in chil-
perhaps involving cognitive skills and work hab- dren’s cognitive performance are likely due to
its as additional mediators. In this section we combinations of parenting, environmental, and
focus on studies that do not consider parenting, biological differences between children from
but simply consider cognitive skills and work lower- and higher-SES families.
habits as mediators between social class back-
ground and schooling success. 1.5.2.1 A cademic Work Habits
As with cognitive skills, social class differences
If cognitive skills and academic work habits in task-related work habits are observed very
are to mediate the relationship between SES and early in children’s development. Morgan et  al.
course grades, they must first be shown to differ (2009) estimated SES differences in behaviors at
across social classes, with middle- and upper- 24  months of age, using data collected from
class students showing greater cognitive skills administration of the Bayley Scales of Infant
and academic work habits than students from the Development. They found that when mother and
working and lower classes. I now turn to the child were given simple tasks to do, children
empirical evidence on these issues. from mothers in the lowest education quintile
were more than twice as likely as those from
Cognitive Skills  A relatively large body of mothers in the highest education quintile to not
empirical research has demonstrated that social persist at tasks, to be inattentive, to show no
class differences in cognitive skills begin very interest, to be uncooperative, and to be frustrated.
early in life, are of relatively large magnitudes at Since mother and child performed as a dyad,
kindergarten entry and are, in general, main- these outcomes are suggestive of mother–child
tained through to high school education. Fernald interaction differences across social classes.
et al. (2013) found that significant disparities in
vocabulary and language processing efficiency By kindergarten entry, the academic work
were already evident at 18  months between habits of children in the top SES quintile are 0.6
infants from higher- and lower-SES families, and standard deviation above those of children from
that by 24  months there was a 6-month gap the bottom SES quintile (Duncan and Magnuson
between SES groups in processing skills critical 2011, p. 56). By 5th grade this behavior gap has
to language development. That is, it was not until widened slightly. By 8th grade these gaps have
24  months of age that the less advantaged chil- decreased to about 0.4 standard deviation, and by
dren reached the same level of processing speed 12th grade to 0.3 standard deviation (Farkas
and accuracy displayed by the more advantaged 2011, p. 79) In kindergarten, children from the
children at 18  months. Hart and Risley (1995) lowest SES quintile show antisocial behaviors
and Hoff (2003) showed that higher social class (externalizing problem behaviors) that are 0.3
parents speak a very much greater number and standard deviation worse than those from the
variety of words to their infants and toddlers than highest SES quintile. By 5th grade this gap has
do working-class parents, and these differences increased to 0.5 standard deviation but it
partially explain the larger vocabularies of mid- decreases thereafter, to 0.3 SD in 12th grade.
dle and upper-class children. Farkas and Beron However, this may be at least partly due to the
(2004) found large SES oral vocabulary gaps at higher school dropout rate among students with
36  months of age, and subsequent vocabulary the worst behaviors, particularly those from
growth rates that were similar across different lower- and working-class homes.

1  Family, Schooling, and Cultural Capital 29

In sum, there is ample evidence showing that most strongly determined by Basic Skills and
family social class background is a powerful Work Habits. The path model in Fig. 1.2 shows
determinant of academic skills and work habits. the results of putting these effect estimates
If these are found to strongly determine the together into a single model. Basic skills has a
course grades a student receives, then the basic direct effect of 0.22 on course grades plus an
tenets of the cultural capital theory presented indirect effect of 0.38 × 0.27 = 0.10 via course-
here will have been supported. work mastery, for a total effect of 0.32. Work
habits has a direct effect of 0.53 on course grades
1.5.3 S kills and Habits Determine plus an indirect effect 0.32 × 0.27 = 0.09, for a
Course Grades total effect of 0.62. Coursework mastery itself
has a direct effect of 0.27. Other effects are much
Farkas et al. (1990) and Farkas (1996) used data smaller, with the largest of these being days
collected from the Dallas School District to esti- absent, with a direct effect of −0.15. In sum, aca-
mate portions of the model presented in Fig. 1.1. demic work habits exert the strongest effect on
These studies contained measures of poverty, teacher-assigned course grades in 7th and 8th
academic skills and work habits, and course grade social studies, with a total effect size of
grades. They lacked measures of parenting, but 0.62. That is, increasing these work habits by 1
they did have separate measures of basic aca- standard deviation would lead to a course grade
demic skills (measured by the Iowa Test of Basic increase of 0.62 of a standard deviation. By con-
Skills) and of the actual coursework mastery of trast, basic skills have an effect only about half
the students in the 7th and 8th grade social stud- this size, and the effect of coursework mastery is
ies classes from which the study sample was smaller still.
drawn (this measure is drawn from a curriculum
referenced test administered uniformly within the Group differences in work habits also
Dallas schools). accounted for large portions of race gaps in aca-
demic achievement. For example, other findings
These researchers found that when it comes to included the fact that Asian children, scoring
predicting social studies course grades assigned high on academic work habits, received a double
in 7th and 8th grade, the direct effect of course- benefit from these behaviors. First, these work
work mastery had an effect size of 0.27, and the habits strongly and positively affected course-
direct effect of basic skills (measured by lan- work mastery, which raised their grades.
guage arts and math scores from the Iowa Test of However, over and above this effect via course-
Basic Skills) was 0.22. The largest direct effect work mastery, Asians’ good work habits earn an
was that of academic work habits, with a stan- extra reward by further raising their grades.
dardized coefficient of 0.53. Absenteeism, dis-
ruptiveness, and appearance and dress also had These are striking findings. It has been widely
significant direct effects, but of much smaller believed that during the early elementary grades,
magnitude. The striking finding is that despite when children are being trained to have good aca-
controls for two types of cognitive skills, work demic learning habits, these habits form a signifi-
habits still had such a large effect size, even as cant portion of the teacher-assigned course grade.
late as middle school, when one might expect But it has also been believed that in middle and
cognitive performance to have become much high school, where students have different teach-
more important than the student’s work habits. ers for different academic subjects, and the focus
is on learning the assigned material, tests and
These are direct effects, with all variables con- other objective measures of such learning play
trolled. But in addition, there are indirect effects the largest role in course grade assignment. Yet,
in which causally prior variables affect course this is not what we have found for 7th and 8th
grades through their effects on mediators. One grade social studies. Of course these data are
such mediator is coursework mastery. This is from the late 1980s, in only one city. It would be
valuable to have research updating these findings

30 G. Farkas

to a more recent time period and to the nation as unfortunately the effect sizes are small, and fade
a whole. More generally, a structural equation out by second grade (Puma et al. 2010). In addi-
model could be estimated in which the habitus is tion, many higher-income families also send their
a latent variable, with test scores and academic children to child care centers, which are often of
work habits as indicators. Or, perhaps a better higher quality than those utilized by low-income
model would involve two latent habitus variables, families, thereby exacerbating rather than reduc-
one for cognitive ability and the other for habits ing SES differentials in the cognitive stimulation
and behaviors. Then test scores would be the and support provided to preschoolers. Further,
indicators of cognitive skills, and teacher reported research has shown that longer time periods in
judgments of student work habits and other out-of-home child care tend to be associated with
behaviors as the indicators of the latent habits more conflictual relationships between the child
and behaviors variable. This would seem to be and both teachers and the child’s mother, although
the appropriate operationalization of a model in this effect is reduced when the care is of higher
which the student’s habitus is not directly quality (Early Child Care Research Network
observed. 2005). Overall, and particularly for cognitive
skills, preschool programs can play a role in
Research by Blanchard and Muller (2015) fur- complementing or even substituting for the
ther supports the importance of academic work efforts of parents to prepare children for kinder-
habits in determining the teacher-assigned course garten entry. There is a very large research litera-
grade. This study analyzes ELS:2002 data to test ture on this, which I do not have the space to
whether teacher-perceived student work habits consider here. For a useful starting point, see the
mediate the relationship between being an immi- meta-analysis by Duncan and Magnuson (2013).
grant student and the course grade received in
10th grade math. The authors find that the teach- 1.5.5 Peer Effects
er’s perception that the student “works hard” is
positively related to the student’s course grade, In addition to the family and teachers, the peer
with (after controls) an effect size of 0.62 group has been found to exert significant effects
SD. This is a very strong effect, which is likely at on the educational success of students. That
least partly inflated by the authors’ failure to con- working- and lower-class peer groups, particu-
trol test scores in the analysis. larly among males, can create a culture antithet-
ical to school achievement has long been
1.5.4 C hild Care reported by ethnographic studies. This has been
reported within both White and Black low-
Parenting activities are not the only way that chil- income peer groups (Ogbu 1978, 2003; Willis
dren’s school-related habitus and cultural capital 1977; Macleod 1995; Anderson 1999; Tyson
may be shaped. Federal and state preschool pro- et al. 2005) and has led to a spirited controversy
grams for low-income children were designed to regarding the existence of an “oppositional cul-
compensate for SES differences in the stimulat- ture,” in which, among both male and female
ing, nurturing, and healthful aspects of home Black students, striving for academic achieve-
environments. Head Start, and most recently ment is denigrated as “acting White” (Fordham
state-run preschool programs, serve many, but and Ogbu 1986; Ainsworth-Darnell and Downey
not all, low-income children, since Head Start is 1998; Downey and Ainsworth-Darnell 2002;
not fully funded. The best of these programs Farkas et  al. 2002; Carter 2005; Fryer and
operate in child care centers utilizing a “whole Torelli 2010). The reality of this effect may be
child” model of comprehensive service provi- inferred from the well-established finding that,
sion, including health- and family-related ser- all other things equal, the higher the percentage
vices. Research has shown that these programs of Black students in a school, the lower the aver-
do increase cognitive performance, although

1  Family, Schooling, and Cultural Capital 31

age academic a­chievement of students in the of social class differences in the school-related

school (Mickelson et al. 2013). Of course other habitus of these children. (For additional reading

explanations, including lower-quality teachers, see Currie and Reichman (2015), and the litera-

are also possible. ture cited there.)

But what about peer effects of having a high

percentage of working- and lower-class students Academic Work Habits
as Personality Traits
in a school? Palardy (2013) found that even 1.6

among otherwise similar students, attending a

school where the average student comes from a

high-SES family significantly increases the prob- Once we moved past studies restricting cultural

ability of high school graduation and college capital to behaviors and skills associated with

enrollment. He concludes that these effects are elite “high culture” we found a great commonal-

largely explained by peer effects, which tend to ity among the skills and habits reported by eth-

be negative in low-SES schools. Once again, the nographers as being central to different

likely mediating mechanism is lower levels of subcultural repertoires, those included by psy-

academic work habits where the student peer chologists in scales of quality parenting such as

group comes largely from working- and lower-­ the HOME, those explicitly listed on report cards

class homes. Similar findings have been reported to be graded by teachers, and those work habits

by Anderson (1999), Carrell and Hoekstra (2010), that are empirically found to join cognitive per-

Hanushek et al. (2003), Morris (2008), and Willis formance as being most predictive of the grades

(1977) among others. assigned by teachers. As noted by Farkas (2003),

these are the same characteristics included in the

1.5.6 Biological Make-Up concept of “conscientiousness” that industrial

and Health psychologists find to be the only one of the “big

five” personality characteristics to predict job

performance and wages. These are the same char-

Beginning even before birth, children from low- acteristics that the Knowledge is Power Program

SES households experience lower-quality health (KIPP n.d.) schools, the charter school network

than higher-SES children. Low-SES children are with the most well-documented positive effects,

more likely to experience growth retardation and uses as the basis of their “contract” with

inadequate neurobehavioral development in students.

utero. These children are also more likely to be These conscientious academic work habits

born prematurely, at low birth weight, with a dis- have been somewhat neglected by sociologists of

ability, or with fetal alcohol syndrome or education, even as economists and psychologists

AIDS. These outcomes are typically due to poor have concentrated on them, in some cases claim-

prenatal care, poor nutrition and maternal sub- ing that they hold the key to improving the school-

stance use during pregnancy, and living in an ing and life outcomes of children from

environment where violence is common and con- low-income households. Thus, Borghans et  al.

taining toxins such as lead and airborne pollut- (2008) and Heckman and Kautz (2014) empha-

ants. Further, when low-income children size personality traits, particularly conscientious-

experience a health problem or disability they are ness, as the key to success in school and life.

less likely than higher-SES children to receive These authors refer to the work of psychologist

adequate health care (Bradley and Corwyn 2002). Roberts (2009), who states that “conscientious-

There is insufficient space here to review this ness is a personality trait, which is defined as a

very large literature. But there is little doubt that ‘tendency to respond in certain ways under cer-

the biological and health differences between tain circumstances,’…the tendency to think, feel,

children from low and middle social class back- and behave in a relatively enduring and consistent

grounds play a significant role in the d­ evelopment fashion across time in trait-affording situations.”

32 G. Farkas

Note that this is very close to the definition of The Knowledge is Power (KIPP n.d.) charter
habitus discussed earlier. Heckman and Kautz go schools appear to have done just that. First devel-
on to list the American Psychology Dictionary oped by two Teach for America teachers in 1994,
description of conscientiousness, its facets, this network of charter schools now numbers
related skills, and analogous childhood tempera- more than 180 schools across the country. Their
ment skills. The word is defined as the tendency highly structured program for children from low-
to be organized, responsible, and hardworking. It income households includes commitment state-
includes competence (efficient), order (orga- ments that must be agreed to by teachers, parents,
nized), dutifulness (not careless), achievement and students. That for students reads as follows:
striving (ambitious), self-discipline (not lazy),
and deliberation (not impulsive). Related skills • I will always work, think, and behave in the
are grit, perseverance, delay of gratification, best way I know how, and I will do whatever it
impulse control, achievement striving, ambition, takes for me and my fellow students to learn.
and work ethic. Analogous childhood tempera- This also means that I will complete all my
ment skills are attention/(lack of) distractibility, homework every night, I will call my teachers
effortful control, impulse control/delay of gratifi- if I have a problem with the homework or a
cation, persistence, and activity. problem with coming to school, and I will
raise my hand and ask questions in class if I do
These traits and behaviors are similar to the not understand something.
academic work habits we have emphasized
throughout this chapter. Almlund et  al. (2011) • I will always behave so as to protect the safety,
report effect sizes for intelligence and each of the interests, and rights of all individuals in the
big five personality traits in their effects on years classroom. This also means that I will always
of education attained. The largest effect is for listen to all my KIPP teammates and give
conscientiousness, with an effect size of 0.25. everyone my respect.
The next largest effect is for intelligence. The
other personality traits either have no or much • I am responsible for my own behavior, and I
smaller effects. This finding, in which academic will follow the teachers’ directions.
work habits have even stronger effects on educa-
tional attainment than test scores, is reminiscent This is nothing other than the academic work
of Farkas’ (1996) findings on the relative strength habits discussed throughout this chapter.
of effect of test scores and work habits on course Similarly, the pledge that must be signed by par-
grades. For a wide-ranging discussion of the ents reads as follows:
importance of grit in life success see Duckworth
(2016). Here we see another example of the con- We will make sure our child arrives at KIPP every
vergence of viewpoints in sociology, economics, day by 7:25 a.m. (Monday–Friday) or boards a
and psychology. KIPP bus at the scheduled time. We will always
help our child in the best way we know how and
1.7 P olicy Implications we will do whatever it takes for him/her to learn.
This also means that we will check our child’s
What are the policy implications of the finding homework every night, let him/her call the teacher
that teacher-judged academic work habits are a if there is a problem with the homework, and try to
major mediating factor for the strong positive read with him/her every night. We will always
relationship between family social class back- make ourselves available to our children and the
ground and student success in school? Can this school, and address any concerns they might have.
finding be employed to increase the school suc- This also means that if our child is going to miss
cess of children from lower- and working-class school, we will notify the teacher as soon as pos-
families? sible, and we will carefully read any and all papers
that the school sends home to us.

Here the emphasis on checking homework
and reading with the student every night reflects
the kinds of good parenting behaviors embodied
in the HOME score instrument.

1  Family, Schooling, and Cultural Capital 33

What has been the impact of KIPP schools on enrollment and completion, leading to more

the students attending them? The answer is that rewarding (in both the pecuniary and non-­

they have shown significant positive effects on pecuniary sense) employment careers.

reading and math achievement at elementary, I consider this narrative to be consistent with

middle and high school levels (Angrist et  al. the work of economist Gary Becker, who brought

2010, 2012; Nichols-Barrer et  al. 2015; Tuttle great attention to the development and output

et al. 2015). These results appear to be the bright- from human skills, and of sociologist James

est spot in a great variety of school structure Coleman, who emphasized the importance of

experiments that have been unleashed by the social networks, trust, and the individual’s posi-

charter schools movement. This is perhaps the tion within a social structure as determinants of

strongest evidence yet for the overwhelming human capital development and deployment.

importance of student skills, habits, and styles in Sociologist Pierre Bourdieu added a focus on

the determination of student outcomes, and the how the individual’s position in the social struc-

possibility of fostering increased school success ture affects her habitus, which helps determine

for students from low-income and working-class the individual’s enacted educational cultural cap-

families by creating a schooling environment ital (skills and behaviors) that are judged by

within which these students can improve these teacher-gatekeepers whose feedback and

skills, habits, and styles. assigned grades help determine the student’s edu-

cational attainment and thus subsequent occupa-

1.8 Summary and Discussion tional employment and earnings. In this chapter I
have tried to show that cultural capital theory, by

introducing student strategies of action con-

I began this chapter by discussing social repro- strained by their habitus, producing classroom

duction, arguably the most important empirical cultural capital (skills and work habits) judged by

finding in the sociology of education. Seeking to teachers, offers an integrative focus in which the

understand the mechanisms by which the chil- study of educational stratification can be

dren of middle- and upper-class parents attain advanced in a way consistent with the visions of

greater school success than lower- and working-­ Becker, Coleman, and Bourdieu, as well as many

class children, I explicated Bourdieu’s theory of other sociologists, economists, and psychologists

cultural capital, which supposes that parents from working on these issues today.

different social classes imbue children with dif- The epigraph was a quotation from Paula

ferent sorts of habitus, or dispositions (including England’s ASA Presidential Address (2016),

skills) toward action. The resulting habitus dif- where she defined personal characteristics as

fers across social classes, so that children from “things individuals carry across situations, such

middle and higher social class families tend to as skills, habits, identities, worldviews, prefer-

present the cultural capital (cognitive skills and ences or values.” England is a gender scholar, and

academic work habits enacted in the classroom does not generally undertake research in the soci-

and homework) that are pleasing to, and rewarded ology of education. She writes about skills and

by, teachers, whereas this is less common among habits because she is treating them as central to

children from lower- and working-class families. the “social structure and personality” theorizing

Teachers respond by giving higher report card that, she argues, offers an important vantage

grades to the middle- and upper-class students, point for understanding a very wide variety of

leading them to experience more successful aca- outcomes across the social world. She concen-

demic trajectories and to attain greater academic trates on two examples. One is the finding that

skills and knowledge as they progress up through more women than men report being bisexual. The

the elementary, middle, and high school grade second is that disadvantaged women use contra-

levels. These more successful K–12 trajectories ception less consistently than more advantaged

then translate into more successful postsecondary women, even when they do not want to get

34 G. Farkas

­pregnant. She argues that in each case, the struc- improved understanding of those portions of
turally disadvantaged position of the members of working- and lower-class family and neighbor-
a group, gay men in the first case, disadvantaged hood life that are most determinative of student
women in the second, has caused them to inter- academic skills and work habits. We have already
nalize particular skills, habits, identities, world- seen that the hypothesis that elite cultural activi-
views, preferences, or values. For a gay man, this ties are central to the school success of middle-
is a straight identity, which he feels constrained class children has been empirically rejected. We
to present because of the stigma attached to gay- have also seen that the parenting activities mea-
ness. For the disadvantaged woman, this is a sured by instruments such as the HOME explain
lesser sense of efficacy—the ability to align your only a modest portion of the better academic
identity with your goals—which is the result of skills and work habits of middle- and upper-class
the constrained resources available at her place in children. We expect that children’s academic
the social structure. A principal point of England’s work habits evolve continuously over time, so
paper is to argue against the long held view that that behavior in kindergarten likely reflects pre-
any study involving the personal characteristics school behavior. And we have also learned that
of a group that is disadvantaged by the social greater time in lower-quality preschool is associ-
structure involves “blaming the victim” (Ryan ated with lower attention skills and greater exter-
1971), a point of view arguing that focusing on nalizing behavior (McCartney et  al. 2010). Yet
the personal characteristics of disadvantaged research is only beginning on how parenting,
groups shifts the discussion away from the social social structure, and peers shape preschool
structure and instead makes the individual’s situ- behavior, and the four together shape student
ation “their own fault.” But instead, England behaviors in kindergarten. (For examples of this
argues, examining the personal characteristics of work see Henry and Rickman 2007; Neidell and
disadvantaged groups needn’t direct attention Waldfogel 2010.) This is just one of many areas
away from the social structure. Instead, it merely where it would be useful to learn more about par-
shifts the social structure one step back in the enting, peers, skills, and behaviors and their joint
causal chain, from which it leads to the creation variation across the social structure. In this
of the personal characteristics (habitus) which in regard, recent research has suggested that the test
turn lead to less than desirable (constrained) score achievement gap between children from
behaviors. That is, the social structure constrains families in the top and bottom income quintile
the individual to become a person who produces increased significantly in the 1970s and 1980s
less than desirable behaviors. As England quotes (Reardon 2011), but appears to have modestly
Wacquant (2005, p. 316), “the society becomes narrowed between 1998 and 2010 (Reardon and
deposited in persons in the form of lasting dispo- Portilla 2015), and these most recent changes
sitions, or trained capacities and structured pro- may be at least partly due to narrowing of the
pensities to think, feel and act in determinant income–parenting gap (Bassok et al. 2016). Such
ways, which then guide them.” Thus, the vision over-time change in social class differences in
of cultural capital theory presented here is built parenting and test scores indicate that social
upon the now well-demonstrated notion that to reproduction is dynamic rather than static, and
understand the lower academic performance of should be studied as a dynamic system subject to
working- and lower-class students we need to a wide variety of forces, importantly including
understand the social psychology of both the aca- government policy and public media dissemina-
demic performance and the academic work hab- tion of information about families and parenting.
its they bring to the school, as well as the
student–teacher interactions and course grades Another area ripe for investigation is social
that result from these interactions. class differences in the detailed patterns of aca-
demic work habits within each grade level, and as
There are many promising directions for students move up the grade levels. Our current
future research in these areas. One is to seek measures of student academic work habits are

1  Family, Schooling, and Cultural Capital 35

typically restricted to a few questions asked of extent than has heretofore been demonstrated
the teacher at a single point in time. More detailed by other programs, policies, or interventions.
data might provide insights that could be used to Efforts to better understand the detailed mech-
develop interventions, programs, or policies to anisms by which student skills and habits
improve the academic work habits of working- determine educational attainment, and how
and lower-class children. Other promising schools can be managed so as to increase all
research areas include greater attention to how three for children from working- and lower-
student course grades evolve over time, and how class households, should be high on the
these are related to outcomes such as dropout, research agenda of sociologists of education
high school graduation, college enrollment, and for many years to come.
employment. To the greatest extent possible these
studies should attempt to move beyond merely References
correlational evidence, and incorporate evidence
from experimental or quasi-experimental Ainsworth-Darnell, J.  W., & Downey, D.  B. (1998).
research designs. If non-experimental data (e.g., Assessing the oppositional culture explanation for
those in large national data sets like the ECLS-K) racial/ethnic differences in school performance.
are used, researchers should at least attempt to American Sociological Review, 63, 536–553.
use methods such as teacher fixed effects that at
least partially control for possible selection bias. Almlund, M., Duckworth, A., Heckman, J.  J., & Kautz,
T. (2011). Personality psychology and economics. In
Also worthy of investigation is the way that E. A. Hanushek, S. Machin, & L. Wofsmann (Eds.),
cognitive skills and academic work habits pro- Handbook of the economics of education, (Vol. 4, pp.
vide an advantage to children from higher social 1–181). Amsterdam: Elsevier.
class backgrounds in higher education and the
labor market. Empirical research has established Anderson, E. (1999). Code of the street: Decency, vio-
that positive attitudinal/behavioral traits have lence, and the moral life of the inner city. New York:
effects on wages that are at least as large as those Norton.
of cognitive skills (see Hall and Farkas (2011)
and the studies cited there). But the detailed Angrist, J. D., Dynarski, S. M., Kane, T. J., Pathak, P. A.,
mechanisms of these effects across varied occu- & Walters, C. R. (2010). Inputs and impacts in char-
pations and industries are unknown. There is ter schools: KIPP Lynn. American Economic Review:
much to study here. Papers & Proceedings 100 (May). http://www.aeaweb.
org/articles.php?doi=10.1257/aer.100.2.1
I began with the question, what are the
mechanisms by which children from middle- Angrist, J.  D., et  al. (2012). Who benefits from KIPP?
and higher social class parents tend to achieve Journal of Policy Analysis and Management, 31(4),
greater school success than those from lower- 837–860.
and working-c­lass parents? The evidence
shows that the greater school success of Bassok, D., Finch, J.  E., Lee, R., Reardon, S., &
m­ iddle- and upper-class children is due to Waldfogel, J.  (2016). Socioeconomic gaps in early
their stronger cognitive skills and academic childhood experiences, 1998 to 2010. AERA Open,
work habits. These are in turn strongly affected 2(3), 1–22.
by parenting, peers, and genetics, as well as
teachers and school climate. Fortunately, Becker, G. (1964). Human capital. New  York: National
schools such as KIPP have demonstrated that Bureau of Economic Research.
by creating a culture focused on developing
positive academic work habits and related val- Blanchard, S., & Muller, C. (2015). Gatekeepers of the
ues, with buy-in from both teachers and par- American Dream: How teachers’ perceptions shape
ents, children from lower- and working-c­ lass the academic outcomes of immigrant and language-­
schools can succeed at school to a greater minority students. Social Science Research, 51,
262–275.

Bodovski, K. (2010). Parental practices and educa-
tional achievement: Social class, race, and habitus.
British Journal of Sociology of Education, 31(2),
139–156.

Bodovski, K. (2014). Adolescents’ emerging habitus:
The role of early parental expectations and practices.
British Journal of Sociology of Education, 35(3),
389–412.

Bodovski, K., & Farkas, G. (2008). “Concerted cultiva-
tion” and unequal achievement in elementary school.
Social Science Research, 37, 903–919.

36 G. Farkas

Bodovski, K., Jeon, H., & Byun, S.  Y. (2016). Cultural dents: Exploring the intersection of race and parents’
capital and academic achievement in post-socialist educational attainment. Sociological Perspectives,
Eastern Europe. British Journal of Sociology of 55(1), 17–42.
Education, 38(6), 1–18. Duncan, G.  J., & Magnuson, K. (2011). The nature and
impact of early achievement skills, attention skills,
Borghans, L., Duckworth, A.  L., Heckman, J.  J., & ter and behavior problems. In G.  J. Duncan & R.  J.
Weel, B. (2008). The economics and psychology Murnane (Eds.), Whither opportunity? Rising inequal-
of personality traits. Journal of Human Resources, ity, schools, and children’s life chances (pp.  47–70).
XLIII(4), 972–1059. New York: Russell Sage.
Duncan, G.  J., & Magnuson, K. (2013). Investing in
Bourdieu, P. (1986). The forms of capital. In J.  E. preschool programs. The Journal of Economic
Richardson (Ed.), Handbook of theory and research Perspectives, 27(2), 109–132. https://doi.org/10.1257/
for the sociology of education (pp.  241–258). jep.27.2.109.PMC.
Westport: Greenwood. Duncan, G.  J., et  al. (2005). The apple does not fall far
from the tree. In S. Bowles, H. Gintis, & M. E. Groves
Bowles, S., & Gintis, H. (1976). Schooling in capitalist (Eds.), Unequal chances: Family background and eco-
America. New York: Basic Books. nomic success (pp. 23–79). New York: Russell Sage.
England, P. (2016). Sometimes the social becomes per-
Bradley, R. H., & Corwyn, R. F. (2002). Socioeconomic sonal: Gender, class, and sexualities. American
status and child development. Annual Review of Sociological Review, 81, 4–28.
Psychology, 53, 371–399. Farkas, G. (1996). Human capital or cultural capital?

Bradley, R. H., Corwyn, R. F., Burchinal, M., McAdoo, Ethnicity and poverty groups in an urban school dis-
H. P., & Coll, C. G. (2001). The home environments trict. New York: Aldine de Gruyter.
of children in the United States Part II: Relations with Farkas, G. (2003). Cognitive skills and noncognitive
behavioral development through age thirteen. Child traits and behaviors in stratification processes. Annual
Development, 72(6), 1868–1886. Review of Sociology, 29, 541–562.
Farkas, G. (2011). Middle and high school skills, behav-
Carrell, S., & Hoekstra, M. (2010). Externalities in the iors, attitudes, and curriculum enrollment, and their
classroom: How children exposed to domestic vio- consequences. In G. J. Duncan & R. J. Murnane (Eds.),
lence affect everyone’s kids. American Economic
Journal: Applied Economics, 2(1), 211–228. Whither opportunity? Rising inequality, schools, and
children’s life chances (pp. 71–90). New York: Russell
Carter, P. L. (2005). Keepin’it real: School success beyond Sage.
Black and White. New York: Oxford University Press. Farkas, G., & Beron, K. (2004). The detailed age trajec-
tory of oral vocabulary knowledge: Differences by
Cheadle, J.  (2008). Educational investment, family con- class and race. Social Science Research, 33, 464–497.
text, and children’s math and reading growth from kin- Farkas, G., Grobe, R., Sheehan, D., & Shuan, Y. (1990).
dergarten through third grade. Sociology of Education, Cultural resources and school success: Gender, ethnic-
81, 1–31. ity, and poverty groups within an urban school district.
American Sociological Review, 55, 127–142.
Coleman, J., & Hoffer, T. (1987). Public and private high Farkas, G., Lleras, C., & Maczuga, S. (2002). Does
schools: The impact of communities. New York: Basic oppositional culture exist in minority and poverty
Books. peer groups? American Sociological Review, 67,
148–155.
Coleman, J., et al. (1966). Equality of educational oppor- Fernald, A., Marchman, V.  A., & Weisleder, A. (2013).
tunity. Washington, DC: U.S.  Government Printing SES differences in language processing skill and
Office. vocabulary are evident at 18 months. Developmental
Science, 16, 234–248. https://doi.org/10.1111/
Currie, J.  & Reichman, N., (Eds.), (2015). Policies desc.12019.
to promote child health. In The future of children. Fordham, S., & Ogbu, J. (1986). Black students’ school
Washington, DC: Princeton-Brookings. success: Coping with the “burden of ‘acting White’”.
Urban Review, 18, 176–206.
De Graaf, N.  D., De Graaf, P.  M., & Kraaykamp, G. Fryer, R. G., & Torelli, P. (2010). An empirical analysis of
(2000). Parental cultural capital and educational “acting White”. Journal of Public Economics, 94, 380–
attainment in the Netherlands: A refinement of the cul- 396. https://doi.org/10.1016/j.jpubeco.2009.10.011.
tural capital perspective. Sociology of Education, 73, Gaddis, M. (2013). The influence of habitus in the rela-
92–111. tionship between cultural capital and academic
achievement. Social Science Research, 42, 1–13.
DiMaggio, P. (1982). Cultural capital and school success: Goffman, E. (1961). Asylums: Essays on the condition
The impact of status culture participation on the grades
of U.S. high school students. American Sociological of the social situation of mental patients and other
Review, 47, 189–201. inmates. New York: Doubleday.

Downey, D., & Ainsworth-Darnell, J. (2002). The search
for oppositional culture among Black students.
American Sociological Review, 67, 156–164.

Duckworth, A. (2016). Grit: The power of passion and
perseverance. New York: Scribner.

Dumais, S. (2002). Cultural capital, gender, and school
success: The role of habitus. Sociology of Education,
75(1), 44–68.

Dumais, S., Kessinger, R.  J., & Ghosh, B. (2012).
Concerted cultivation and teachers’ evaluations of stu-

1  Family, Schooling, and Cultural Capital 37

Goldthorpe, J. H. (2007). “Cultural capital”: Some critical McCartney, K., Burchinal, M., Clarke-Stewart, A., Bub,
observations. Sociologica, 2, 1–22. K.  L., Owen, M.  T., & Belsky, J.  (2010). Testing a
series of causal propositions relating time in child care
Greenstone, D. (1991). Culture, rationality, and the to children’s externalizing behavior. Developmental
underclass. In C.  Jencks & P.  Peterson (Eds.), The Psychology, 46, 1–17. https://doi.org/10.1037/a0017886.
urban underclass (pp.  399–408). Washington, DC:
Brookings Institution. Mickelson, R.  A., Bottia, M.  C., & Lambert, R. (2013).
Effects of school racial composition on K–12 math-
Hall, M., & Farkas, G. (2011). Adolescent cognitive ematics outcomes: A metaregression analysis. Review
skills, attitudinal/behavioral traits and career wages. of Educational Research, 83, 121–158. https://doi.
Social Forces, 89, 1261–1285. org/10.31102/0034654312475322.

Hannerz, U. (1969). Soulside: Inquiries into ghetto culture Mincer, J.  (1958). Investment in human capital and
and community. New York: Columbia University Press. personal income distribution. Journal of Political
Economy, 66, 281–302.
Hanushek, E., Kain, J., Markman, J., & Rivkin, S. (2003).
Does peer ability affect student achievement? Journal Mincer, J.  (1974). Schooling, experience, and earnings.
of Applied Econometrics, 18, 527–544. New York: National Bureau of Economic Research.

Hart, B., & Risley, T. (1995). Meaningful differences in Morgan, P., Farkas, G., Hillemeier, M., & Maczuga, S.
everyday experiences of young American children. (2009). Risk factors for learning-related behavior
Baltimore: Brookes. problems at 24 months of age: Population-based esti-
mates. Journal of Abnormal Child Psychology, 37,
Heckman, J., & Kautz, T. (2014). Achievement tests 401–413.
and the role of character in American life. In J.  J.
Heckman, J.  E. Humphries, & T.  Kautz (Eds.), The Morris, E.  W. (2008). “Rednecks,” “Rutters,” and
‘Rithmetic: Social class, masculinity, and schooling
myth of achievement tests: The GED and the role in a rural context. Gender and Society, 22, 728–751.
of character in American life (pp.  1–56). Chicago:
University of Chicago Press. Neidell, M., & Waldfogel, J. (2010). Cognitive and non-
Henry, G. T., & Rickman, D. K. (2007). Do peers influence cognitive peer effects in early education. Review of
children’s skill development in preschool? Economics Economics & Statistics, 92(3), 562–576.
of Education Review, 26(1), 100–112.
Hoff, E. (2003). The specificity of environmental influ- NICHD Early Child Care Research Network. (2005).
ence: Socioeconomic status affects early vocabulary Early child care and children’s development in the
development via maternal speech. Child Development, primary grades: Follow-up results from the NICHD
74, 1368–1378. Study of Early Child Care. American Educational
Jencks, C., Bartlett, S., Corcoran, M., Crouse, J., Research Journal, 42(3), 537–570.
Eaglesfield, D., et  al. (1979). Who gets ahead?
The determinants of economic success in America. Nichols-Barrer, I., Gleason, P., Gill, B., & Tuttle,
New York: Basic Books. C.  C. (2015). Student selection, attrition, and
Jenkins, R. (1992). Pierre Bourdieu. New  York: replacement in KIPP middle schools. Educational
Routledge. Evaluation and Policy Analysis, 38, 5–20. https://doi.
Kingston, P. (2001). The unfulfilled promise of cultural org/10.3102/0162373714564215.
capital theory. Sociology of Education, 74(Extra
Issue), 88–99. Nunez, J.  C., Suarez, N., Rosario, P., Vallejo, G., Valle,
KIPP. (n.d.). http://www.kipp.org/our-approach/five- A., & Epstein, J.  L. (2015). Relationships between
pillars#sthash.xvyNMjIm.dpuf. Accessed on 28 Jan perceived parental involvement in homework, student
2016. homework behaviors, and academic achievement:
Lamont, M., & Lareau, A. (1988). Cultural capital. Differences among elementary, junior high, and high
Sociological Theory, 6, 153–168. school students. Metacognition and Learning, 10,
Lareau, A. (2011). Unequal childhoods. Class, race, and 375–406.
family life (2nd ed., with an update a decade later).
Berkeley: University of California Press. Ogbu, J.  (1978). Minority education and caste: The
Lareau, A., & Weininger, E.  B. (2003). Cultural capital American system in cross-cultural perspective.
in educational research: A critical assessment. Theory Orlando: Academic.
and Society, 32(5), 567–606.
Lee, J., & Bowen, N. (2006). Parent involvement, cultural Ogbu, J. (2003). Black American students in an American
capital, and the achievement gap among elementary suburb: A study of academic disengagement. Mahwah:
school children. American Educational Research Erlbaum.
Journal, 43(2), 193–218.
Lucas, S. (2001). Effectively maintained inequal- Palardy, G. J. (2013). High school socioeconomic segre-
ity: Educational transitions and social background. gation and student attainment. American Educational
American Journal of Sociology, 106, 1642–1690. Research Journal, 50, 714–754. https://doi.
Macleod, J. (1995). Ain’t no makin’it: Leveled aspirations org/10.3102/0002831213481240.
in a low-income neighborhood (2nd ed.). Boulder:
Westview. Patterson, O. (Ed.). (2015). The cultural matrix:
Understanding Black youth. Cambridge, MA: Harvard
University Press.

Potter, D., & Roksa, J.  (2013). Accumulating advan-
tages over time: Family experiences and social class
inequality in academic achievement. Social Science
Research, 42, 1018–1032.

38 G. Farkas

Puma, M., Bell, S., Cook, R., & Heid, C. (2010). Head who vary in birth status. Pediatrics, 117, 1608–1617.
Start Impact Study: Final Report. Washington, DC: https://doi.org/10.1542/peds2005-1284.
U.S. Department of Health and Human Services. Swidler, A. (1986). Culture in action: Symbols and
strategies. American Sociological Review, 51,
Raftery, A. E., & Hout, M. (1993). Maximally maintained 273–286.
inequality: Expansion, reform, and opportunity in Swidler, A. (2008). Comment on Stephen Vaisey’s
Irish education. Sociology of Education, 66, 41–62. “Socrates, Skinner, and Aristotle: Three
ways of thinking about culture in action”.
Reardon, S. F. (2011). The widening socioeconomic status Sociological Forum, 23, 614–618. https://doi.
achievement gap: New evidence and possible explana- org/10.1111/j.1573-7861.2008.00080.x.
tions. In R. J. Murnane & G. J. Duncan (Eds.), Whither Tramonte, L., & Willms, J.  D. (2010). Cultural capital
and its effects on education outcomes. Economics of
opportunity? Rising inequality and the uncertain life Education Review, 29, 200–213.
chances of low-income children. New  York: Russell Tuttle, C.  C., Booker, K., Gleason, G., Chojnacki, G.,
Sage Foundation. Knechtel, V. et  al. (2015). Understanding the effect
Reardon, S.  F., & Portilla, X.  A. (2015, September).
of KIPP as it scales: Volume I, impacts on achieve-
Recent trends in socioeconomic and racial school ment and other outcomes. Final report of KIPP’s
readiness gaps at kindergarten entry (Working Paper). investing in innovation grant evaluation. Princeton:
Stanford University. Mathematica Policy Research.
Reeves, R. V., & Howard, K. (2003). The parenting gap. Tyson, K., Darity, W., & Castellino, D.  R. (2005). It’s
Washington, DC: Center on Children and Families, not “a Black thing”: Understanding the burden of act-
Brookings Institution. ing White and other dilemmas of high achievement.
Roberts, B.  W., Jackson, J.  J., Fayard, J.  V., Edmonds, American Sociological Review, 70(4), 582–605.
G., & Meints, J. (2009). Conscientiousness. In M. R. Van Voorhis, F. L., Maier, M. F., Epstein, J. L., & Lloyd,
Leary & H. Hoyle (Eds.), Handbook of individual dif- C. M. (2013, October). The impact of family involve-
ferences in social behavior (pp. 369–381). New York: ment on the education of children ages 3 to 8. MDRC
Guilford Press. report.
Roscigno, V. T., & Ainsworth-Darnell, J. W. (1999). Race, Wacquant, L. (2004). Body and soul: Notebooks of an
cultural capital, and educational resources: Persistent apprentice boxer. New York: Oxford University Press.
inequalities and achievement returns. Sociology of Wacquant, L. (2005). Habitus. In J. Becket & M. Zafirovski
Education, 72, 158–178. (Eds.), International encyclopedia of economic sociol-
Rumberger, R.  W., & Palardy, G.  J. (2004). Multilevel ogy (pp. 315–319). London: Routledge.
models for school effectiveness research. In D. Kaplan Wacquant, L. (2011). Habitus as topic and tool: Reflections
(Ed.), The SAGE handbook of quantitative methodol- on becoming a prizefighter. Qualitative Research in
ogy for the social sciences (pp.  235–258). Thousand Psychology, 8, 81–92.
Oaks: SAGE Publications. Wacquant, L. (2014). Putting habitus in its place:
Ryan, W. (1971). Blaming the victim. New York: Vintage. Rejoinder to the symposium. Body and Society, 20(2),
Schultz, T.  W. (1960). Capital formation by education. 118–139.
Journal of Political Economy, 68, 571–583. Willis, P. (1977). Learning to labor. London: Gower.
Schultz, T.  W. (1981). Investing in people. Berkeley: Wilson, W. J. (1996). When work disappears. New York:
University of California Press. Knopf.
Smith, K. E., Landry, S. H., & Swank, P. R. (2006). The
role of early maternal responsiveness in support-
ing school-aged cognitive development for children

Power, Relationships, and Trust 2
in Sociological Research
on Homes, Schools,
and Communities

Erin McNamara Horvat and Karen Pezzetti

Now more than ever, the world needs research that sheds light on how social contexts
matter in learning, teaching, student achievement, and in the development of equitable
and just forms and systems of education.

Elizabeth Birr Moje (2016)

Abstract 2.1 Introduction

This chapter offers a critical perspective on Above, Moje acknowledges that the social con-
sociological research exploring the interac- texts that sociologists of education need to “shed
tions among students’ homes, schools and light on” are multiple. This multiplicity means
communities. We conceptualize each of these not only that each student lives in a unique social
spaces as a unique context that influences context, but further, that each young person
students and, as such, must be attended to grows up negotiating multiple social contexts; it
both on its own terms but also especially is the interactions and relationships among and
where each context meets, conflicts with, or between these contexts that we must explore.
exerts power over the others. We highlight
three major areas of promising research in this Children’s educational experiences are influ-
field: first, research that attends to the tensions enced by the various cultures and expectations of
inherent to the struggle for power among and their home lives, schools, and communities. It is
between these contexts; second, research important to keep in mind that while there are
that explores the foundations and practice of many differences across race, class, and culture,
creating equal, communicative relationships all families want children to do well in school.
between stakeholders from each context; and However, for some children, the specific cultures
third, research that can account for the pres- and expectations across home, school, and com-
ence, absence, or impact of trust in these munity align, working together to nurture and sup-
relationships. port the academic and social development of these
young people. The educational experiences of
E. M. Horvat (*) other children, in contrast, are characterized by
Drexel University, Philadelphia, PA, USA imbalances in power or incongruities in the reali-
e-mail: [email protected] ties across these three contexts. Too often, schools
expect racially, linguistically, and culturally
K. Pezzetti diverse families to adopt the White, ­middle-c­ lass,
Grand Valley State University, Allendale, MI, USA Eurocentric norms and values of schools, reinforc-
e-mail: [email protected]

© Springer International Publishing AG, part of Springer Nature 2018 39
B. Schneider (ed.), Handbook of the Sociology of Education in the 21st Century, Handbooks
of Sociology and Social Research, https://doi.org/10.1007/978-3-319-76694-2_2

40 E. M. Horvat and K. Pezzetti

ing a power imbalance between home and school. school relationship.” In the field of educational
The contested interactions between families, psychology, the theoretical construct parental
schools, and communities have roots in deep ten- involvement has been the focus of a considerable
sions about how various stakeholders understand body of research in the last 30 years. This litera-
the role of schools in our society. These stakehold- ture tends to focus on the activities and behaviors
ers have engaged repeatedly over questions such that parents do at home (like help with home-
as: How, when, and where should we educate our work) or at school (like attend a parent–teacher
children? For what purpose are we educating our conference) that may correlate positively with
children? What are the impacts on children when student academic achievement. Many studies
different families, schools, and communities have sought to discover what factors mediate
answer these questions in different ways? And, whether or not—or how—parents engage in
most importantly for this chapter, how do research- activities like these (i.e., Cardona et  al. 2012;
ers approach the study of the ways that interac- Davis-Kean 2005; Hoover-D­ empsey and Sandler
tions among home, school, and community 1997; Lendrum et  al. 2015; Schneider and
influence students’ experiences and achievement? Coleman 1993; Smith et  al. 1997; Spera 2005;
Wanat 2012; Widding 2012).
This chapter offers our perspective on some
current trends in sociological research, focusing Some researchers have critiqued the construct
on the interactions and relationships among three of parental involvement as limited to specific
different contexts: home, school, and community. forms of engagement dictated by schools. From
Below, we offer a brief historical and theoretical this perspective, parents who do not show up for
overview of the literature. Rather than provide parent–teacher conferences or school events risk
an exhaustive review, we explore the gains that being labeled as ineffective, uncaring, uninvolved
have been made and the areas that have been parents. These critics have proposed a different
neglected by particular perspectives. We focus on framing of the term: family engagement (Epstein
approaches that allow researchers to explore and and Sheldon 2002; Ferlazzo and Hammond
understand the complex power dynamics and ten- 2009). In contrast with parental involvement,
sions that are interwoven throughout research in which focuses on what parents do (or do not do),
this area. We conclude the chapter with a review family engagement foregrounds the responsibil-
of the most recent scholarship and policy and ity of schools to nurture trusting, two-way rela-
discuss directions for future work. tionships with all parents (Yull et al. 2014).

2.2 D efinitional Considerations The particular framing of the research term is
not just rhetoric. Whether researchers choose to
In the last 60  years, researchers, practitioners, study “parents” or “families” or “home” matters;
and policy-makers have used different and evolv- just as whether they focus on “parenting style”
ing terms to refer to the relationship between the (i.e., Darling and Steinberg 1993), “involvement”
home and school. Cutler (2000, p.  5) described or “interaction” or “engagement” or “relationship”
the home–school relationship at its best as a or “participation” (Lewis and Forman 2002). For
“marriage between distinct but reciprocal institu- instance, Mallett (2004) explored the ways that
tions,” yet parents and teachers have more fre- sociologists conceptualize “home.” She points out
quently been characterized as “natural enemies” that both the use of “home” and “family” as socio-
(Lightfoot 2004; Waller 1932). Perhaps influ- logical terms and the relationship between them
enced by underlying assumptions about the are “keenly contested” (p.  73). She argues that
parties involved, some scholars have studied researchers who use “home” and “family” inter-
parental involvement, while others have focused changeably are usually drawing on a Eurocentric,
on “family–school interactions” or “home– middle-class, heteronormative conceptualization
of a home as a particular kind of house a person
was born in, inhabited by a nuclear family.

2  Power, Relationships, and Trust in Sociological Research on Homes, Schools, and Communities 41

In this chapter, we have deliberately used the control and power shifted so far into the hands of

word home because it can encompass all individ- the professionals that some educators began to

uals who support a student in the space, including scrutinize parenting practices and eventually to

parents, grandparents, siblings, extended family, recommend “modifications in the behavior of

and non-related caregivers. This more expansive families” through parental education programs

view of the home–school relationship embedded (Cutler 2000, p. 8).

in a community context is drawn from a collec- In the twentieth century, however, schools

tive orientation towards education. As we will relinquished some of their power and control to

see, over time, schooling and the act of providing parents. In the 1960s and 1970s, for example,

for the education of children and youth have been Parents Rights Movements advocated for

at times the purview of the family, at times the increased decision-making power in public

school, and at other times the community. Each schools. In 1997, the National Parent Teacher

stakeholder has fought for the responsibility and Association adopted a set of standards or guide-

right to make decisions that impact the education lines for the home–school relationship based pri-

of children and youth. marily on the work of Joyce Epstein. The

standards highlight the importance of communi-

2.3 Historical Antecedents cation between schools and families, but make it
clear that schools should initiate that communi-

cation. In 2001, No Child Left Behind stipulated

The relationship between home and school has parental involvement as a condition for receiving

been contested for centuries. Over the past federal funding (Reynolds et al. 2015).

150 years, there have been numerous shifts in the Today, most educators and researchers

distribution of power between these two stake- acknowledge both that children’s first teachers

holders. Before the existence of widespread pub- are their families and that families should be

lic schools, White American parents had extensive involved in their children’s academic lives. Still,

control of what their children learned and how despite this more welcoming attitude toward

and when they learned it. Before the mid-1800s, family involvement in schools, issues of power

most children were primarily educated in the and control remain endemic to this relationship.

home by family members, or, for wealthier fami- (Henderson 2007; Lareau and Muñoz 2012).

lies, by tutors. Some children went to nearby Henderson delineates four different kinds of

neighbors’ homes or dame schools for lessons. power stances and practices that schools adopt

With the advent of widespread public schools in toward families: the Partnership School, the

the nineteenth century, however, control over Open-Door School, the Come-if-We-Call School

education generally shifted from the home to the and the Fortress School (p. 14). While any typol-

school (Cutler 2000). As teachers and administra- ogy can over-simplify complex relationships,

tors worked to professionalize and bureaucratize Henderson’s work ably captures the different

schooling systems, education came to be seen as approaches taken to working with students’

a scientific enterprise that was best left in the home spaces and the people in them. It also

hands of experts. As school systems grew in scale acknowledges the imbalance of power wielded

in the nineteenth century, some educators and by educators in defining these relationships.

reformers made efforts to formalize contact More recently, Lareau and Muñoz (2012) docu-

between families and schools. For example, in ment the tussles over control in middle-class

the 1840s, report cards began to replace face-to-­ schools where parents are organized, engaged,

face communication (Cutler 2000). Parents’ and want to share control with classroom teach-

groups (or PTAs) first appeared in the 1880s and ers and administration.

contributed to the institutionalization of further Historically, researchers studying parents and

aspects of the family–school relationship. In the schools tended not to adopt a critical stance.

Progressive Era and then again after World War I, What this means is that the context, power

42 E. M. Horvat and K. Pezzetti

­structures, and roles that shaped parental involve- that treated parent or family involvement as a one
ment or family involvement in schools were size fits all enterprise—miss an essential piece of
accepted without critique or question. As the puzzle in understanding how families and
Baquedano-L­ ópez et al. (2013) note, normative communities’ reciprocal relations with schools
White middle-class norms have been the default are shaped. They do not take into account the
expectations for family involvement. These social context and power dynamics that surround
expectations often translated directly into differ- these relationships. And while some studies in
ential treatment of students. There is a fair the last 20 years have begun to address power
amount of recent research that explores the ways differentials, Baquedano-López, Alexander, and
that these normative expectations for family Hernandez contend that much of this work is still
involvement shape educational experiences and rooted in a deficit narrative about racially, cultur-
outcomes (i.e., Auerbach 2012; Cardona et  al. ally, and linguistically diverse parents.
2012; Reynolds et al. 2015). Rist’s classic (1970)
study regarding teacher expectations and the Further straining the power dynamics between
way that these expectations played into aca- families and schools is the fact that each year,
demic placement as well as long-term achieve- fewer American students are taught by teachers
ment and outcomes provides an illustrative case. who share their cultural background. As the
This seminal article marked a turning point in teaching force continues to be predominantly
thinking about the impact of home influences on White and middle-class while the American pub-
academic outcomes for sociologists of educa- lic school student body diversifies, the power dif-
tion. While interpretations of this article often ferential between home and school takes on
rightly focus on the class background of the fam- added dimensions of race and class. While the
ilies and the impact of social class background politics of who should decide what and how stu-
on the teacher’s placement of students, this arti- dents should learn in school have always been
cle also illustrates the powerful role of family influenced by issues of race and class, we believe
background and context in shaping how teachers that these tensions are exacerbated in the present
and school agents interpret family involvement context in which parents and families are experi-
in education. encing tremendous pressure to advantage their
children by performing in a variety of ways
There are a few relevant points here for our dictated by White, middle- and upper-class
analysis of research on the family–school inter- policy-­makers and educators (Baquedano-López
action. Rist argues that the teacher placement of et al. 2013; Horvat and Baugh 2015; Oakes et al.
students in ability groups was based on attributes 2015).
rooted in family background. Thus, the home–
school or family–school connection extends far Although a handful of recent studies question
beyond the notion of the PTA or report card con- this assumption (i.e., Robinson and Harris 2014),
ferences. Students are in large part products of most of the literature we reviewed for this chapter
their environment, and the most formative envi- accepted as a point of departure the premise that
ronmental factor in their lives is the home. There parental involvement and a positive home–school
is power in teachers’ perceptions of students. As relationship boosts students’ academic achieve-
this classic article illustrates, these perceptions ment (i.e., Dusi 2012; Hoover-Dempsey and
are rooted in familial or home influences on stu- Sandler 1997). Epstein and Sanders (2000,
dents that are often generated in relation to a p. 287) summarize this consensus: “It is now gen-
hypothetical “ideal type” of successful student, erally agreed that school, family, and community
illustrating the pervasive presence and power of partnerships are needed in order to improve the
normative expectations for students and families children’s chances for success in school.”
(see also Rose 2016 for an extension of this argu- Generally speaking, researchers tend to study the
ment). As Baquedano-López et  al. (2013) note, relationships between parents and schools from
these early studies—as well as later formulations either the parent side of the question or the school
side. From the parent side, researchers theorize

2  Power, Relationships, and Trust in Sociological Research on Homes, Schools, and Communities 43

that parental involvement helps students in the (Cutler 2000, p. 207). As we discuss below, this
following ways: Involved parents model their has important consequences. In particular, we
value for education, which their children then fear that this trend may increase educational
adopt; involved parents better understand inequity if parents’ differential capacities to meet
schools’ expectations for their children, so they those expectations exacerbate entrenched class
can help their children meet those expectations; and race patterns of inequality.
and involved parents provide their children with
extracurricular and academic opportunities that 2.4 T heoretical Frameworks
support in-school learning outside of school
(Crosnoe 2015). Studies on the school side Many theoretical perspectives have been
include research on the efficacy of interventions employed in research and policy related to the
designed to reduce inequities in family and com- interactions between family and school. In under-
munity engagement. A strong home–school rela- standing the research and past practice and
tionship allows schools to better understand the exploring future directions for research and pol-
particular strengths, needs and goals of children icy, it is important to understand both these per-
and their families. In addition, researchers have spectives and the strengths and limitations they
found that schools favor children whose parents bring. Historically, there has been a separation
are involved (Crosnoe 2015). between home and school in both policy and
research. In other words, researchers who studied
It is also important to note that the debate schools rarely explored the influences of family,
about whether parents or teachers are to blame and, likewise, family researchers rarely explored
when children or schools perform poorly on stan- the powerful effects of school on family (Epstein
dardized tests obscures other possible responsi- and Sanders 2000). Often, explorations of the
ble parties. As the government has withdrawn wider community—including the neighborhood,
resources from public schooling, teachers have after-school issues and care and other community
borne the primary heft of responsibility (and organizations and resources such as churches,
blame) for educating (and failing to educate) recreation centers, and libraries—have been
children. In a situation in which they have chal- completely excluded in discussions of the home–
lenging jobs and limited resources, teachers look school relationship.
for someone else to shift the responsibility to—
and parents are the available suspects. This More recently, researchers have expanded
increasing tension, aided by the implementation their lenses to include a more holistic view of
of high-stakes accountability measures in an home and school that, for the most part, acknowl-
environment of decreasing resources, again edges the overlapping influences present as well
draws our attention to the contested nature of the as the important role played by the wider
home–school–community relationship. communities in which families and schools are
situated (Epstein 1987; Epstein and Sheldon
In 2016, we believe it is important to note that 2002; Epstein 2013; Epstein et  al. 2013; Smith
schools’ expectations for parents have increased et al.1997). Below, we review some of the signifi-
in the last 20 years. In order to ensure that their cant theoretical perspectives that have informed
children receive a quality education, parents must sociological research on the relationships and
do more now. Cutler summarized the current interactions between schools and families. In
state of the home–school relationship in the fol- doing so, we highlight the contributions of some
lowing way: “Today it would be unusual for par- scholars and inevitably miss others. As noted
ents to believe that they should not be active at previously, researchers operating from a psycho-
their children’s school. Educators, reformers, and logical perspective have produced a rich litera-
even politicians have made such an issue of ture on the role of parent involvement in student
parental involvement that many well-meaning achievement (see, for example, Hoover-Dempsey
mothers and fathers probably feel guilty about
not being more active than they already are”

44 E. M. Horvat and K. Pezzetti

and Sandler 1997; Hoover-Dempsey et al. 2001). from a background in psychiatry in the early
A thorough review of this body of literature is 1960s, adopted a developmental whole-child
outside the scope of this chapter (see Kim and approach. Comer and his team at the Yale Child
Sheridan 2015 for an excellent foundational Study Center were asked to work with high-­
overview of this work). In contrast, our goal in poverty low-performing schools in New Haven,
this chapter is to shine a light on some of the sem- CT.  They adopted what we might now call a
inal ideas that have informed sociological strengths-based approach that emphasized the
research in this area. role of social capital in school improvement
(Comer 1995). Comer notes, “the social capital
2.4.1 Social Capital needed for school and life success is not provided
in most public schools serving non-mainstream
Without question, one of the concepts most cen- families” (2015). Moreover, Comer acknowl-
tral to any understanding of communities and edged not only the importance of connections as
schools is social capital. Mentioned by almost all an aspect of social capital but also the trust
of the major researchers in the field, social capital embedded in these relationships. Comer’s School
refers to the value of the relationships of an indi- Development Model thus included a strong
vidual or group. James Coleman (1987) explored emphasis on the construction of trusting relation-
the social capital found within and surrounding ships across and among students, parents, teach-
families, as well as in the relationships between ers, and a wide array of actors in the surrounding
families, communities, and schools. His work community. Comer’s training was in psychiatry
with Thomas Hoffer and Sally Kilgore (1982) on and his model, therefore, logically focuses on the
social capital in Catholic schools found that the importance of attending to the psychological and
community support and shared values that individual developmental needs and safety of
inhered in these environments were critical to children as they proceed through school.
their success. Coleman’s work is foundational to However, unlike his predecessors from the field
the understanding of school–home–community of psychology, Comer emphasized the develop-
relations, as it brought significant national atten- ment of trusting relationships—social capital—
tion to the role of culture in both schools and in in his model for school improvement.
families as an important variable. Though the
findings of the Coleman Report are often misun- Like Coleman’s school improvement model,
derstood, and his work was often over-simplified Epstein’s (Epstein and Sanders 2000; Epstein
to be understood as simply finding that family et al. 2013) far more recent work on school, fam-
background matters more than money in achiev- ily, and community partnerships draws on the
ing school success, a more careful reading of concept of social capital. Epstein’s “theory of
Coleman’s work finds a groundbreaking focus on overlapping spheres of influence” highlights the
the relationships among family background, capacity of educators, parents, and community
community resources, the effects of social class, members to work together in the service of stu-
and school success. dents. Epstein’s description of “school-like”
homes (p. 36) in which a family’s expectations of
Coleman and his colleagues’ focus on the role children at home are similar to the expectations
of social capital in understanding school success of teachers in schools acknowledges the impor-
highlighted the relationship between the family tance of consistent values and expectations across
and school as a key variable in understanding these spheres.
schooling outcomes. Others in the field drew on
this foundational work. James Comer (1995, While both Comer and Epstein acknowledge
2015), who came to work in school improvement the power of social capital in their models, nei-
ther takes a particularly sociological view. What
we mean by this is that the work does not focus

2  Power, Relationships, and Trust in Sociological Research on Homes, Schools, and Communities 45

on what some see as the inherent conflict between approach (i.e., Auerbach 2012; Baquedano-­
schools and families, nor does it provide an anal- López et  al. 2013; Reay 1999; Reynolds et  al.
ysis that accounts for the differential amounts of 2015; Williams and Sanchez 2012).
power that people from different social classes
and positions in society can wield. As some Central to the critical work investigating the
scholars have noted, the work often downplays relations between home, school, and community
the role of conflict or tension between parents is a deeper and more nuanced exploration into the
and schools (Lareau and Horvat 1999; Lareau factors that promote strong relationships across
and Muñoz 2012; Lewis and Forman 2002). In these stakeholders using this concept of social
addition, we argue that this work does not suffi- capital. The work of the Consortium on Chicago
ciently account for the importance of particularly School Research (Bryk and Schneider 2002,
class but also cultural, racial, and ethnic differ- 2003; Bryk et  al. 2010) explored the important
ences in shaping home–school–community role of trust in these social relationships. We
relationships. review the practical implications of this work in
subsequent sections, however, here we note the
In our view, this theoretical difference stems theoretical sophistication of this work that
from fundamentally different theoretical formu- focused explicitly on the notion of relational trust
lations of social and cultural capital. Comer, as a key variable in promoting positive relation-
Epstein, Coleman, Putnam, and others view ships across stakeholders. This work both valued
social capital as a readily shared commodity the resources that promoted trust and school suc-
within families and communities. Bourdieu’s cess that reside in low-income communities and
conceptualization (Bourdieu 1986; Bourdieu and implicitly recognized the power of parents and
Wacquant 1992), which provides the foundation communities in advancing school reform in rela-
for Lareau (2000, 2003) and her followers’ work, tionship with school agents. With careful, detailed
takes a more critical stance. In Bourdieu’s formu- and extensive data collection, Bryk and Schneider
lation, all forms of capital (social, cultural, sym- identified the components of relational trust:
bolic) are not created equal. They are the product respect, personal regard, competence in core role
of the family social class background and are— responsibilities, and personal integrity. They
and this is the important point—differentially show that the benefits of developing trust across
valued by dominant societal institutions, includ- these domains are vast. This work illustrated that
ing schools. As Lareau (2014) notes in explaining trust is the “connective tissue that binds individu-
the central finding of her seminal 2003 work, “the als together to advance the education and welfare
key issue was not the intrinsic nature of parenting of students” (Bryk and Schneider 2003) and pro-
itself, but rather the uneven rewards dominant vided a theoretical and empirical base for further
institutions bestowed on different types of strate- development of critical research and practices to
gies.” Research like Lareau’s represents a move bridge the divides across home, school, and
away from simply examining best practices or community.
from attempting to build relationships across
overlapping spheres of influence in a child’s life These more recent theoretical developments
to include a focus on the powerful ways in which that place power at the heart of the analysis and
some displays and activities are accorded value use a more contextualized and inclusive notion of
by dominant and powerful institutions, most “family” that includes relevant actors from the
notably schools, and others are not. This acknowl- home and community provide a theoretical foun-
edgement of the differential power accorded dation for understanding collective parental and
forms of social and cultural capital by dominant community engagement in schooling. We hope
institutions lays the groundwork for a more critical that future research continues to shift away from
an “all players are equal” over-generalization and

46 E. M. Horvat and K. Pezzetti

toward a stance that recognizes the power inher- lying much of the new rhetoric remains a view of
ent in institutions and takes seriously the unequal families, particularly nondominant families, as
distribution of power across race and social class. ineffective at preparing their children for school
and life. From this perspective, poor child-­rearing
2.4.2 T he Importance of Power: practices and so-called “broken homes” are
A Critical Approach to Family– responsible for national and international
School Relations achievement gaps and the perceived decline of
American public schools.
Recent scholarship has translated these theoreti-
cal notions into a reconceptualization of the Second, Baquedano-Lόpez and her colleagues
home–school–community relationship incorpo- identify the trope Parents as First Teachers: The
rating notions of power and privilege into the literature and policy on early childhood educa-
analysis. In an excellent critical review of the lit- tion takes as a beginning point that parents are
erature on parent involvement in schools, their children’s first teachers. The creation of
Baquedano-Lόpez et  al. (2013) identify and federally-funded programs intended to close the
describe five ways that academic discourse and “school readiness gap” often begins with the
public policy have framed the relationship assumption that nondominant parents are failing
between parents and schools. Baquedano-Lόpez at this role, and therefore require training and
and her colleagues contend that although several intervention to perform the “right” (i.e., middle-­
of these tropes seem like common sense, each class, White, Eurocentric) kinds of behaviors and
also is drawn from a White middle-class American interactions with their children.
worldview and hides a deficit view of nondomi-
nant parents and families, specifically low-income A related trope is Parents as Learners.
families, families of color, and families who are Baquedano-Lόpez and her colleagues argue that
immigrants. We understand Baquedano-Lόpez, many family literacy programs sponsored by pro-
Alexander, and Hernandez’ use of the term trope grams like the Workforce Investment Act, ESEA,
as a deliberate choice meant to signal the accepted, and the Head Start Act draw on a decontextual-
common, and often overused nature of the stories ized understanding of literacy that assumes that
or narratives employed to explain the relationship some parents need support in gaining fundamen-
between parents and schools. Instead of the term tal tools and understandings so that they can
narrative, which could also signal an agreed-upon assist their children in school. This perspective
point of view or story that gives meaning to a par- ignores the home literacy practices that families
ticular set of circumstances, the authors use trope may already be engaging in and prioritizes those
to indicate that these viewpoints are widely held, practices valued by the dominant culture.
often unquestioned, and embedded into the short-
hand of the lexicon. In this context, the use of the Increasingly prominent in the legislation and
term trope implies a cynical and critical approach literature is the frame of Parents as Partners.
to the narratives used to explain family–school While the rhetoric of partnership implies equal
relationships that highlights the taken-for-granted footing, a closer look at legislation like Title I
nature of these viewpoints. Because so much of reveals that while the term “partner” is used, the
the research and practice on parent involvement in mandated parent’s role is passive and relegated to
schools takes as an underlying assumption one or surveillance activities such as “monitoring atten-
more of these tropes, we briefly review them here. dance, homework completion, and TV watching”
(Baquedano-Lόpez et al. 2013, p. 155). The lim-
Several of the tropes discussed below fall into its of these prescribed activities suggest that,
the first and largest discursive frame: Parents as from this perspective, the ideal parent’s role may
Problems. Although current programs and poli- be more like that of a “compliance officer” or
cies are eager to avoid deficit discourses, under- “watchdog” rather than a partner (Baquedano-­
Lόpez et al. 2013, p. 155).

The final trope, Parents as Choosers and
Consumers, highlights the role of parents in an

2  Power, Relationships, and Trust in Sociological Research on Homes, Schools, and Communities 47

increasingly privatized, market-based model of color in a Northeastern urban school district. In

education wherein parents are expected to make conversation with the parents, Yull and her col-

decisions like choosing which school their chil- leagues discovered that the parents saw the rac-

dren will attend. Baquedano-Lόpez and her col- ism and the cultural incompetence of the school
leagues argue that this frame is limiting in that it staff as a barrier to their effective engagement

relegates parental involvement to the act of with the school. As the study was conducted as

choosing from a limited set of options. part of a larger community-based participatory

Furthermore, the discourse of choice often hides action research approach project, the team of

underlying structural inequalities. As Baquedano-­ university-based researchers shared the parents’

Lόpez et al. contend, “the mechanisms of choice concerns with the school district administrators
create a hierarchical system of inequitable distri- and collaborated to revise the district’s strategic

bution that harms nondominant families when plan. We find research like this to be exciting for

that choice does not contest neighborhood segre- several reasons: First, it genuinely takes up the

gation, racialized tracking, or inequitable concerns of parents of color, and second, the

resource/opportunity provisions, and existing collaborative, action research design means that

systems of power harmful to nondominant peo- not only does this study contribute to the

ples” (2013, p. 156). research literature, it also seeks to immediately

Many other recent empirical studies have improve the conditions for home–school inter-

brought a critical lens to the study of home– actions in this community. Indeed, universities

community–school relationships that questions ought to consider themselves part of the com-

the assumption that families must always adapt to munities that can contribute both to individual

schools’ values and expectations. For instance, a student academic success and the creation of

recent study focused on a course that preservice positive learning environments and school cul-

teachers take that is intended to help them develop tures (McAlister 2013).

family-centered involvement practices, re-­

framing the issue of creating positive home– 2.5 New Developments
school–community relationships as at least partly in School, Home,
the responsibility of teacher education programs and Community Connection
(Amatea et al. 2012). Evans (2014) explored the Research: Escalating
ways that diverse parents made use of a Demands on Parents
community-b­ased organization, instead of the and Community Organizing
local school, in order to meet some of their chil-

dren’s educational needs, highlighting parents’

commitments to their children’s education as In recent years a growing body of research on

well as the important role of community-based school choice (Buckley and Schneider 2003;

organizations in furthering those commitments. Henig 1995; Goyette 2008, 2014; Kisida and

Jefferson (2015) studied the administrative and Wolf 2010; Ravitch 2010, 2013) has demon-

institutional barriers that prevented parents from strated the escalating demands on parents. As

fully participating in a school-turnaround pro- school choice options increase, so, too, do par-

cess, even when some of these practices and poli- ents’ responsibilities. For most of the twentieth

cies were intended to foster parent participation. century, the only real public school choice that

Jefferson’s work highlights the complexities of families had was the choice they could make

enacting policies that are, at least superficially, through moving neighborhoods. Many families

designed to support home–school relationships. who could afford to do so moved to areas with

As another example of recent critical work, schools with better reputations (Coons and

Yull et al. (2014) used Critical Race Theory as a Sugarman 1978). In the twenty-first century,

conceptual framework as they conducted focus however, with the rapid expansion of charter

group interviews with middle-class parents of schools, magnet schools, citywide admission

48 E. M. Horvat and K. Pezzetti

schools, themed schools, and others, the number dren on waitlists years before they enter a partic-

of schooling choices families must make for their ular school/grade, and becoming intensely

children has increased dramatically. While some emotionally invested in charter school lotteries.

families still live in districts where the only cost-­ Finally, many of these non-traditional public

free option is to send children to the local neigh- schools require parents to be involved in particu-

borhood school, a growing number of American lar ways that schools specify, such as volunteer-

parents—including White, middle-class subur- ing a certain number of hours per year, or

ban parents—must use their social networks and becoming organizers, fundraisers, or activists in

“do their research” (Altenhofen et  al. 2016) in the service of the school. Perhaps ironically,

order to ascertain which schools to apply to. many of the proponents of school choice pro-

Previous research has found that parents con- grams use as their most formidable argument the

sider a number of criteria when deciding which desire to increase family engagement in the edu-

school to send their children to, including the cation system, to make public education more

following factors: academics (Schneider et  al. accessible and democratic (Coons and Sugarman

1996), extracurricular activities (Harris and 1978). Research has also examined the nature of

Larsen 2014), social networks (Schneider et al. parental involvement.

1996; Cucchiara 2013a, b), safety (Stewart and Some scholars (Lareau and Muñoz 2012;

Wolf 2014), location (Goyette 2008, 2014), and Horvat et al. 2003) have noted the individualistic

the racial demographics of the school (Altenhofen nature of most research and policy related to

et al. 2016). In weighing these factors, it appears parental involvement. These scholars find that

that parents engage in a multi-step decision- most research has examined the effect of individ-

making process that involves steps such as con- ual parents on their child’s educational experi-

sulting with friends who are parents and/or ences and has largely ignored the collective nature

education professionals, researching prospective of some parental involvement in schools. Other

schools on the internet, and visiting prospective work has explored the tension between the indi-

schools (Altenhofen et  al. 2016; Harris and vidual aims of parents to advance their own child’s

Larsen 2014). This growing list of activities educational success and taking actions that bene-

engaged in by parents in selecting a school are fit children collectively (Cucchiara and Horvat

part of an ever escalating constellation of activi- 2009). In this era of increasing demands on par-

ties that are increasingly expected of parents. ents and a political climate that calls for parents to

Horvat and Baugh (2015) divide these escalat- advocate for their children, a broader approach

ing pressures related to school choice into three that includes the study of parents working together

inter-related categories. First, parents are experi- collectively to effect education reform is vitally

encing increased pressure “to secure a viable important. In addition, we have seen a rise in the

educational setting for their child.” Horvat and incidence of community organizing for educa-

Baugh explain that in previous iterations of our tional reform. This collective approach and efforts

schooling system, schools and teachers have to document and promote community organizing

been the first to blame when children are not as a strategy for reform are most effectively cap-

learning. Increasingly, however, parents are seen tured by the work of Mark Warren and Jeannie

as the responsible parties for sending their chil- Oakes and their colleagues (Oakes and Rogers

dren to “failing” schools. Second, Horvat and 2006; Warren and Mapp 2011).

Baugh describe the increased competition to Building on the early seminal work in this

secure a seat in a high-performing school. area by Dennis Shirley (1997), Warren and Mapp

Researchers have documented phenomena such (2011, p. 5) note: “Community organizing offers

as parents camping out in front of schools in a fresh approach to addressing educational f­ ailure

order to register their children, engaging in as a part of a larger effort to build power for mar-

schemes to demonstrate that they are residents in ginalized communities and tackle issues associ-

the catchments of desired schools, putting chil- ated with poverty and racism inside and outside

2  Power, Relationships, and Trust in Sociological Research on Homes, Schools, and Communities 49

of schools.” The perspective offered by commu- with a “spirit of humility and an openness to the
nity organizing builds on many of the theoretical full emotional presence” of the families. In addi-
notions discussed earlier, namely social capital— tion, leaders and teachers must adopt a Freirian
the paramount importance of power and trust in stance that positions them as “no longer the sole
relationships—as well as a contextual strengths- possessors of knowledge and power” (p.  781).
based approach to school improvement. Warren This practical advice to teachers and leaders from
and Mapp’s book provides powerful examples of a community organizing perspective clearly has
community organizing to improve schools from roots in the sociological tradition that acknowl-
around the country. The authors find that com- edges the power at work in institutions and indi-
munity organizing is a relational process that viduals that shapes educational outcomes. The
“brings a powerful bottom-up thrust to education focus on the importance of building trusting rela-
reform efforts” (p. 251). This approach not only tionships to advance educational aims draws on
focuses on schools but also on the communities the key tenets of social capital.
in which schools reside, and works to address
“educational failure as a part of a larger effort to 2.6 D irections for Future
build power for marginalized communities and Research: Relationships
tackle issues associated with poverty and racism and Context
inside and outside of schools” (p. 5).
We see potential for future work in further explor-
The community organizing paradigm brings a ing the relationships between and among schools,
strengths-based approach to school reform and homes and communities. Indeed, we must redefine
community involvement by recognizing and the way in which research is conducted and policy
valuing the assets to be found in all communities, is drafted to acknowledge the differences inherent
including low-income communities. The across geographical contexts as well as expand our
approach “takes power seriously” (p.  251), work to cross the boundaries of homes, schools,
attending to historic mistrust in the building of and communities. With federally funded programs
relationships in the community and clearly rec- such as Promise Neighborhoods, modeled on the
ognizing the differential power accorded to insti- Harlem Children’s Zone, there is wide acknowl-
tutions and individuals. Lastly, this approach is edgement that improving the educational out-
community- rather than parent-focused. comes of children and youth must be a multifaceted
Providing for the effective education of children and inclusive endeavor that cannot be confined to
and youth is a collective community endeavor, at particular spheres—home, school, or community.
times requiring professional facilitation to build Both the Harlem Children’s Zone, a groundbreak-
the capacity for collaboration. As Oakes and her ing approach begun in 1997 to end the cycle of
colleagues (2015) note, it takes the investment of poverty in New York City that provides compre-
time to build the required relationships and hensive services for an entire neighborhood, and
develop common understandings so that effective the Promise Neighborhoods that have followed in
collective action can be taken. its wake, take as gospel that the needs of commu-
nities, families, parents, children, and students
This approach has implications for leadership must be addressed in a seamless fashion to provide
and teaching. While community organizing is not every child the opportunity to thrive.
usually led by teachers, teachers and school lead-
ers can be powerful allies in this work. As In order to improve educational outcomes for
Oakes and her colleagues argue, the strategies of all students, we must find ways to promote pro-
community organizing—“building relationships, ductive relationships across homes, schools, and
forging common meanings about teaching and communities. Here, we use the word relation-
learning and taking action together” (p. 349)—are ship—as opposed to “interaction” or “involve-
key elements to creating strong ties to students’ ment”—purposefully. As Crosnoe (2015) and
homes and communities. Cooper et  al. (2011)
argue that leaders must enter these relationships

50 E. M. Horvat and K. Pezzetti

Pomerantz et  al. (2007) note, there is growing city’s mayor are multi-faceted and address the
evidence that all home–school connections and needs of children from a combined school, home,
interactions are not, in fact, positive. Greater and community perspective.
attention needs to be paid to developing an under-
standing of the important nuances that influence The capacity of Philadelphia and other urban
the effectiveness of these relationships. In addi- centers to improve the opportunity for children
tion, as Crosnoe contends, relationships and to thrive depends on increasing our capacity to
“congruence” across these contexts do not neces- work seamlessly across these spheres without
sarily need to be a function of direct interaction. becoming mired in dated debates about control
Congruence between what is done at home and while providing educators, families, and activ-
what is done at school matters. Ideally each of ists with the cultural and educational training
these spaces reinforce and build on what is done and tools to work effectively across disparate
in the other. As a goal, Crosnoe introduces the cultural contexts. We see the potential for work
concept of “mutual engagement” in which fami- in the area of educator training and develop-
lies and schools mutually reach out to one ment. As we have illustrated, educators are a
another. How and under what conditions this powerful presence in the lives of students and
relationship of mutual engagement can be built their families. Recognizing the power they
are critical research and policy questions. Such wield, we advocate for research and training for
investigations must recognize as a starting point our predominantly White and female teaching
that communities, homes, and schools vary. force that makes clear to teachers the power
Context matters. Determining how to build rela- that they hold and provides multiple pathways
tionships across these varying contexts is another for working to create trusting relationships
area worthy of the attention of researchers, poli- across the race, class, and ethnic differences.
cymakers, and practitioners. As many others (Oakes et  al. 2015; Crosnoe
2015; Kim and Sheridan 2015) have noted,
Increasingly, building these relationships intentions matter. Adopting an open, curious,
means expanding beyond the traditional bound- and respectful stance to the development of
aries of home, school, and community. Efforts in these relationships is a significant first step.
Philadelphia, currently the poorest major city in Articulating the need to work across tradition-
the nation, provides a case in point. In an effort to ally separate spheres of influence (home,
create opportunities for children to thrive in the school, community) affecting children and
city, Philadelphia local government has passed a young people and providing pathways for
beverage tax to fund quality Pre-K education seamless support across these spheres so chil-
across the city, has funded community schools dren can thrive must become the work of educa-
that provide wraparound services to students, tors, researchers, and policy advocates.
families, and communities, and has partnered
with local industry and higher education partners References
to advance career and technical education and
career access. Each of these core initiatives spans Altenhofen, S., Berends, M., & White, T.  G. (2016).
across school, home, and community. None are School choice decision making among suburban,
targeting a single sphere alone. This approach high-income parents. AERA Open, 2(1), 1–14.
acknowledges the strength in a concerted strategy
across these spheres to improve outcomes for Amatea, E.  S., Cholewa, B., & Mixon, K.  A. (2012).
children and moves beyond stand-alone efforts to Influencing preservice teachers’ attitudes about work-
move the dial on educational outcomes or career ing with low-income and/or ethnic minority families.
competence simply by “engaging parents.” Like Urban Education, 47(4), 801–834.
efforts at the national level such as Promise
Neighborhoods, these signature programs of the Auerbach, S. (Ed.). (2012). School leadership for authen-
tic family and community partnerships: Research
perspectives for transforming practice. New  York:
Routledge.

2  Power, Relationships, and Trust in Sociological Research on Homes, Schools, and Communities 51

Baquedano-López, P., Alexander, R.  A., & Hernández, Cucchiara, M.  B., & Horvat, E.  M. (2009). Perils and
S. J. (2013). Equity issues in parental and community promises: Middle-class parental involvement in urban
involvement in schools: What teacher educators need schools. American Education Research Journal,
to know. Review of Research in Education, 37(1), 46(4), 974–1004.
149–182.
Cutler, W. W. (2000). Parents and schools: The 150-year
Bourdieu, P. (1986). The forms of capital. In J. Richardson struggle for control in American education. Chicago:
(Ed.), Handbook of theory and research for the University of Chicago Press.
sociology of education (pp.  241–258). New  York:
Greenwood. Darling, N., & Steinberg, L. (1993). Parenting style as
context: An integrative model. Psychological Bulletin,
Bourdieu, P., & Wacquant, L. J. (1992). An invitation to 113(3), 487.
reflexive sociology. Chicago: University of Chicago
Press. Davis-Kean, P.  E. (2005). The influence of parent edu-
cation and family income on child achievement: The
Bryk, A., & Schneider, B. (2002). Trust in schools: A core indirect role of parental expectations and the home
resource for improvement. New  York: Russell Sage environment. Journal of Family Psychology, 19(2),
Foundation. 294.

Bryk, A.  S., & Schneider, B. (2003). Trust in schools: Dusi, P. (2012). The family–school relationships in
A core resource for school reform. Educational Europe: A research review. CEPS Journal: Center for
Leadership, 60(6), 40–45. Educational Policy Studies Journal, 2(1), 13–33.

Bryk, A.  S., Sebring, P.  B., Allensworth, E., Easton, Epstein, J.  L. (1987). Toward a theory of family–school
J.  Q., & Luppescu, S. (2010). Organizing schools connections. In K.  Hurrelmann, F.  Kaufmann, &
for improvement: Lessons from Chicago. Chicago: F. Losel (Eds.), Social intervention: Potential and con-
University of Chicago Press. straints (pp. 121–136). New York: DeGruyter.

Buckley, J., & Schneider, M. (2003). Shopping for Epstein, J. L. (2013). Ready or not? Preparing future edu-
schools: How do marginal consumers gather infor- cators for school, family, and community partnerships.
mation about schools? Policy Studies Journal, 31(2), Teaching Education, 24(2), 115–118.
121–145.
Epstein, J.  L., & Sanders, M.  G. (2000). Connecting
Cardona, B., Jain, S., & Canfield-Davis, K. (2012). home, school, and community. In M. T. Hallanan
Home–school relationships: A qualitative study with (Ed.), Handbook of the sociology of education
diverse families. The Qualitative Report, 17(35), 1. (pp. 285–306). New York: Springer US.

Coleman, J. S. (1987). Families and schools. Educational Epstein, J.  L., & Sheldon, S.  B. (2002). Present and
Researcher, 16(6), 32–38. accounted for: Improving student attendance through
family and community involvement. Journal of
Coleman, J.  S., Hoffer, T., & Kilgore, S. (1982). High Educational Research, 95, 308–318.

school achievement: Public, Catholic, and private Epstein, J.  L., Sanders, M., Simon, B., Salinas, K.,
schools compared. New York: Basic Books. Jansorn, N., & Van Voorhis, F. (2013). School, fam-
Comer, J. P. (1995). School power. New York: Free Press.
Comer, J. P. (2015). Developing social capital in schools. ily, and community partnerships: Your handbook for
Society, 52(3), 225–231. action. Thousand Oaks: Corwin.
Coons, J.  E., & Sugarman, S.  D. (1978). Education Evans, M. (2014). Natural allies or natural enemies? The
by choice: The case for family control. Berkeley: evolution of participant self-interest in community-­
University of California Press. based organizations. Journal of Family Diversity in
Copper, C., Riehl, C., & Hasan, A. (2011). Leading and Education, 1(1), 21–39.
learning with diverse families in schools: Critical Ferlazzo, L., & Hammond, L. A. (2009). Building parent
epistemology amid communities of practice. Journal engagement in schools. Santa Barbara: Linworth.
of School Leadership, 20(6), 758–788. Goyette, K. (2008). Race, social background, and school
Crosnoe, R. (2015). Continuities and consistencies across choice options. Equity & Excellence in Education,
home and school systems. In S. M. Sheridan & E. M. 41(1), 114–129.
Kim (Eds.), Processes and pathways of family–school Goyette, K. (2014). Setting the context. In A.  Lareau &
partnerships across development (pp. 61–80). Cham: K. Goyette (Eds.), Choosing homes, choosing schools
Springer. (pp. 1–24). New York: Russell Sage Foundation.
Cucchiara, M. (2013a). “Are we doing damage?” Harris, D. N., & Larsen, M. F. (2014). What schools do
Choosing an urban public school in an era of parental
anxiety. Anthropology & Education Quarterly, 44(1), families want (and why)? School demand and infor-
75–93.
Cucchiara, M. B. (2013b). Marketing schools, marketing mation before and after the New Orleans post-Katrina
school reforms. New Orleans: Education Research
cities: Who wins and who loses when schools become Alliance for New Orleans, Tulane University.
urban amenities. Chicago: University of Chicago Henderson, A. T. (2007). Beyond the bake sale: The essen-
Press. tial guide to family–school partnerships. New  York:
The New Press.

52 E. M. Horvat and K. Pezzetti

Henig, J. R. (1995). Rethinking school choice: Limits of Recommendations for Federal Education Policy, 36,
the market metaphor. Princeton: Princeton University 35–42.
Press. Moje, E. (2016). Message from division VP.
Retrieved from http://www.aera.net/Division-G/
Hoover-Dempsey, K. V., & Sandler, H. M. (1997). Why Social-Context-of-Education-G
do parents become involved in their children’s educa- Oakes, J., & Rogers, J.  (2006). Learning power:
tion? Review of Educational Research, 67(1), 3–42. Organizing for education and justice. New  York:
Teachers College Press.
Hoover-Dempsey, K.  V., Battiato, A.  C., Walker, J.  M., Oakes, J., Lipton, M., Anderson, L., & Stillman, J. (2015).
Reed, R.  P., DeJong, J.  M., & Jones, K.  P. (2001). Teaching to change the world. London: Routledge.
Parental involvement in homework. Educational Pomerantz, E.  M., Moorman, E.  A., & Litwack, S.  D.
Psychologist, 36(3), 195–209. (2007). The how, whom, and why of parents’ involve-
ment in children’s academic lives: More is not always
Horvat, E.  M., & Baugh, D.  E. (2015). Not all parents better. Review of Educational Research, 77(3),
make the grade in today’s schools. Phi Delta Kappan, 373–410.
96(7), 8–13. Ravitch, D. (2010). The life and death of the great

Horvat, E.  M., Weininger, E.  B., & Lareau, A. (2003). American school system: How testing and choice are
From social ties to social capital: Class differences undermining education. New York: Perseus.
in the relations between schools and parent networks. Ravitch, D. (2013). Reign of error: The hoax of the priva-
American Educational Research Journal, 40(2),
319–351. tization movement and the danger to America’s public
schools. New York: Vintage.
Jefferson, A. (2015). Examining barriers to equity: School Reay, D. (1999). Linguistic capital and home–school rela-
policies and practices prohibiting interaction of fami- tionships: Mothers’ interactions with their children’s
lies and schools. The Urban Review, 47(1), 67–83. primary school teachers. Acta Sociologica, 42(2),
159–168.
Kim, E.  M., & Sheridan, S.  M. (2015). Foundational Reynolds, R.  E., Howard, T.  C., & Jones, T.  K. (2015).
aspects of family–school partnership research (Vol. Is this what educators really want? Transforming the
1). Cham: Springer. discourse on Black fathers and their participation in
schools. Race Ethnicity and Education, 18(1), 89–107.
Kisida, B., & Wolf, P. J. (2010). School governance and Rist, R. (1970). Student social class and teacher expecta-
information: Does choice lead to better-informed par- tions: The self-fulfilling prophecy in ghetto education.
ents? American Politics Research, 38, 783–805. Harvard Educational Review, 40(3), 411–451.
Robinson, K., & Harris, A.  L. (2014). The broken com-
Lareau, A. (2000). Home advantage: Social class and pass: Parental involvement with children’s education.
parental intervention in elementary education. Cambridge, MA: Harvard University Press.
Lanham: Rowman & Littlefield Publishers. Rose, T. (2016). The end of average: How we succeed
in a world that values sameness. New York: Harper
Lareau, A. (2003). Unequal childhoods. Berkeley: UC One.
Press. Schneider, B., & Coleman, J. S. (1993). Parents, their
children, and schools. Boulder: Westview Press,
Lareau, A. (2014). Cultural knowledge and social inequal- Inc.
ity. American Sociological Review, 80(1), 1–27. Schneider, B., Schiller, K.  S., & Coleman, J.  S. (1996).
Public school choice: Some evidence from the National
Lareau, A., & Horvat, E.  M. (1999). Moments of social Education Longitudinal Study of 1988. Educational
inclusion and exclusion: Race, class, and cultural Evaluation and Policy Analysis, 18(1), 19–29.
capital in family–school relationships. Sociology of Shirley, D. (1997). Community organizing for urban
Education, 72(1), 37–53. school reform. Austin: University of Texas Press.
Smith, E.  P., Connell, C.  M., Wright, G., Sizer, M.,
Lareau, A., & Muñoz, V. L. (2012). “You’re not going to Norman, J.  M., Hurley, A., & Walker, S.  N. (1997).
call the shots”: Structural conflicts between the prin- An ecological model of home, school, and com-
cipal and the PTO at a suburban public elementary munity partnerships: Implications for research and
school. Sociology of Education, 85(3), 201–218. practice. Journal of Educational and Psychological
Consultation, 8(4), 339–360.
Lendrum, A., Barlow, A., & Humphrey, N. (2015). Spera, C. (2005). A review of the relationship among
Developing positive school–home relationships parenting practices, parenting styles, and adolescent
through structured conversations with parents of learn- school achievement. Educational Psychology Review,
ers with special educational needs and disabilities 17(2), 125–146.
(SEND). Journal of Research in Special Educational Stewart, T., & Wolf, P. (2014). The school choice journey:
Needs, 15(2), 87–96.
School vouchers and the empowerment of urban fami-
Lewis, A.  E., & Forman, T.  A. (2002). Contestation or lies. New York: Palgrave.
collaboration? A comparative study of home–school
relations. Anthropology & Education Quarterly, 33(1),
60–89.

Lightfoot, S. L. (2004). The essential conversation: What
parents and teachers can learn from each other.
New York: Ballantine Books.

Mallett, S. (2004). Understanding home: A critical review
of the literature. The Sociological Review, 52(1),
62–89.

McAlister, S. (2013). Why community engagement
matters in school turnaround. The Next Four Years:

2  Power, Relationships, and Trust in Sociological Research on Homes, Schools, and Communities 53

Waller, W. (1932). The sociology of teaching. New York: relationships. International Journal about Parents in
Wiley. Education, 6(1), 57–68.
Williams, T. T., & Sánchez, B. (2012). Parental involve-
Wanat, C.  L. (2012). Home–school relationships: ment (and uninvolvement) at an inner-city high school.
Networking in one district. Leadership and Policy in Urban Education, 47(3), 625–652.
Schools, 11(3), 275–295. Yull, D., Blitz, L. V., Thompson, T., & Murray, C. (2014).
Can we talk? Using community-based participatory
Warren, M.  R., & Mapp, K.  L. (2011). A match on dry action research to build family and school partnerships
with families of color. School Community Journal,
grass: Community organizing as a catalyst for school 24(2), 9.
reform. Oxford: Oxford University Press.
Widding, G. (2012). What’s gender got to do with it?:
Gender and diversity in research on home and school

Schools and Inequality: 3
Implications from Seasonal
Comparison Research

Douglas B. Downey, Aimee Yoon,
and Elizabeth Martin

Abstract 3.1 Introduction
The traditional narrative posits that differ-
ences in school quality are an important source How do schools influence inequality? This is a
of inequality in the stratification system. big question, and it is fundamental to our under-
Improving the schools attended by disadvan- standing of stratification. We consider what we
taged children, therefore, is key to reducing learn about this question by looking at the magni-
inequality. But what if this view is wrong? We tude of achievement gaps across socioeconomic
discuss the results of seasonal comparison status, race, and gender in cognitive skills at kin-
studies that analyze how achievement gaps dergarten entry, along with how those gaps
change when school is in versus out. Contrary change over the next several years of schooling.
to most education research, these studies sug- Once children are in school, we emphasize sea-
gest that the traditional narrative may be partly sonal comparison studies (observing how
wrong in some cases and entirely misplaced in achievement gaps change when school is in ver-
others. Indeed, when it comes to understand- sus out of session) because they provide an attrac-
ing socioeconomic-based gaps in math and tive way of separating school from non-school
reading skills, the evidence indicates that effects. Of course, this approach falls short of a
achievement gaps are mostly formed prior to comprehensive analysis of the relationship
formal schooling and that schools probably between schools and inequality, but we believe it
reduce the growth in gaps that we would provides important insight regarding how schools
observe in their absence. If this is correct, then influence achievement gaps in cognitive skills
the implications for battling inequality are during the first few years of school. Our review
profound. School reform efforts are likely to helps us understand whether schools tend to
have limited influence; the primary source of make achievement gaps worse, leave them largely
the problem is the level of inequality in the same, or reduce them.
broader society.
In this chapter, we discuss the traditional nar-
D. B. Downey (*) · A. Yoon · E. Martin rative about schools and inequality and then con-
Ohio State University, Columbus, OH, USA trast it with our newer perspective shaped by
e-mail: [email protected] seasonal comparison studies. We then discuss the
methodological advantages of seasonal compari-
son studies, along with their implications for
understanding the relationship between schools
and inequality. We conclude that schools, at least

© Springer International Publishing AG, part of Springer Nature 2018 55
B. Schneider (ed.), Handbook of the Sociology of Education in the 21st Century, Handbooks
of Sociology and Social Research, https://doi.org/10.1007/978-3-319-76694-2_3

56 D. B. Downey et al.

under some conditions, play a more positive role do more than just reproduce inequality; they
than previously thought and significantly reduce increase it. School funding schemes, for exam-
the kind of inequality we would observe in their ple, result in vastly different resources for chil-
absence. dren from advantaged versus disadvantaged
backgrounds (Kozol 1991). Moreover, within-­
3.2 S chools and Inequality: school processes such as ability grouping and
The Traditional Narrative tracking exacerbate skill differences because
advantaged children enjoy better learning envi-
The 1966 Coleman Report has shaped scholarly ronments than their disadvantaged counterparts
discussion of schools and inequality for the last (Condron 2008; Gamoran and Mare 1989; Oakes
half century (Coleman et al. 1966). The massive 1985).
study of over 650,000 American children
famously concluded that variations in children’s This traditional view, largely a response to the
math and reading skills were only weakly related Coleman Report, has created a dominant and
to variation in school resources (e.g., per pupil largely critical narrative about schools and
expenditures, class size). Instead, Coleman and inequality: Schools serving advantaged children
colleagues found that inequality in skills was are better equipped, safer, produce more college-­
mostly associated with inequalities in families, a going graduates, attract better teachers, and pro-
pattern echoed by Jencks (1972). This message vide more Advanced Placement classes, college
represented a serious challenge to those who test preparation courses, and extra-curricular
believed that unequal schools were key to opportunities. This well-known understanding of
inequality and so, not surprisingly, it prompted an schools in American society is why high-income
energetic response. For the last 50 years we have parents are willing to pay more for homes in
been trying to sort things out. neighborhoods with “good” schools and low-­
income parents push for more equitable funding
Critics of the Coleman Report have produced formulas and enter their children into lotteries for
a large body of scholarship outlining the ways a chance to attend a high-prestige charter school.
that schools increase inequality. Bowles and The critical narrative is the driving force behind
Gintis (1976) posited that schools provide the much of the education research aimed at identify-
capitalist economy with workers who know their ing school practices that might reduce achieve-
place and are prepared for their roles. Schools ment gaps and it continues to dominate
contribute to the reproduction of stratification, sociological research on schools. To fix inequal-
therefore, by promoting skills congruent with the ity, the story goes, America needs to improve the
students’ backgrounds. As a result, schools serv- schools serving disadvantaged children.
ing elite students prepare them for jobs as manag-
ers while schools serving poor students prepare A newer line of research consistent with the
them to be workers. Bourdieu (1977) also sees notion that schools are the problem interprets
schools as a culprit but via a different mecha- “between-school” variance as evidence of school
nism. He notes that students from elite back- effects. For example, Borman and Dowling
grounds signify their advantage by exhibiting (2010) reanalyzed Coleman’s data and concluded
“cultural capital” (styles, habits, tastes) that that “[f]ormal decomposition of the variance
allows them to affiliate with elite groups. School attributable to individual background and the
officials and teachers recognize and reward the social composition of the schools suggest that
arbitrary cultural capital of the elite, advantaging going to a high-poverty school or a highly segre-
them unfairly and reproducing inequality gated African-American school has a profound
(DiMaggio 1982). effect on a student’s achievement outcomes,
above and beyond the effect of individual poverty
Pushing the critical perspective of schools or minority status” (p. 1202). Similarly, Jennings
even further, other scholars contend that schools et al. (2015) demonstrate that, if the focus is on
college attendance rather than test scores, there is

3  Schools and Inequality: Implications from Seasonal Comparison Research 57

greater unexplained between-school variance, a years. But most importantly, even if it is possible
pattern that could be attributable to schools. to reduce some achievement gaps via school
These studies demonstrate the possibility of reform alone, it may be more efficient to support
school effects, but it is unclear whether between-­ social reform that prevents these large gaps from
school variation really does reflect differences in emerging in the first place. As we discuss later,
schools rather than the kinds of students who socioeconomic and racial achievement gaps are
happen to attend them. For example, there is sub- largely formed prior to kindergarten.
stantial between-school variance in children’s
skills at kindergarten entry, before schools have a To date, the debate about schools and inequal-
chance to matter. Between-school variance ity has largely been framed as between those who
observed at later stages of schooling may also think schools play a big role (critics of Coleman)
represent significant differences in non-school versus those who believe schools play a modest
factors that typically go unmeasured. role (supporters of Coleman). We believe that this
discussion needs to expand to include the possibil-
Others have pushed further the notion that ity that schools do not increase some achievement
schools are the key to inequality and have made gaps at all, but rather are a meaningful compensa-
the case that school reform itself is enough to tory institution. This more favorable view of
eliminate achievement gaps. For example, schools has played a minimal role in academic or
Abigail and Stephan Thernstrom made this idea policy discussions. It merits greater attention,
popular in their book, No Excuses: Closing the however, because important evidence (discussed
Gap in Learning (Thernstrom and Thernstrom in detail below) suggests that some achievement
2003). They applauded those who have: (1) gaps would be larger if not for schools.
implemented policies aimed at changing school
cultures, and (2) refused to blame family back- 3.3 S chools and Inequality:
ground disadvantage as the reason for the Black– An Alternative Perspective
White gap. And in an article testing the
effectiveness of the Harlem Children’s Zone, Our alternative perspective is motivated by a
Dobbie and Fryer (2011) concluded that school desire to understand schools’ overall role in the
reforms themselves had substantial effects on stratification system. Traditional approaches are
achievement gaps and that school effects were limited because they tend to be school-centric
not improved by the addition of broader commu- and therefore focus on variation within school
nity reforms.1 systems. This approach may reflect scholars’
beliefs that schools are mostly responsible for
Rothstein (2004) notes, however, that the evi- achievement gaps, or that even if schools are not
dence for these “high-flying” schools is substan- mostly responsible, they are the primary policy
tially weaker when examined closely. For lever available for reducing achievement gaps.
example, among schools that managed to severely Indeed, many education researchers admit that
reduce achievement gaps, the majority of them they focus on schools, in part, for political rea-
served a select group of children (e.g., children sons—they view schools as the most politically
whose parents were motivated enough to join the viable mechanism by which to influence the
program). In addition, although some schools opportunity structure.2 In contrast, we see schools
have managed impressive learning gains in a par-
ticular grade for a particular subject, there are 2 Economist Eric Hanushek (1992, p.  106) explains the
virtually no schools that produce impressive focus on schools: “While family inputs to education are
gains across many grades and subjects over many indeed extremely important, the differential impacts of
schools and teachers receive more attention when viewed
1 We are not persuaded by this conclusion because in their from a policy viewpoint. This reflects simply that the
study children in the “school-only” condition enjoyed characteristics of schools are generally more easily
many benefits typically not available to children at school, manipulated than what goes on in the family.”
such as free medical, dental, and mental health services.

58 D. B. Downey et al.

as just one institution affecting the opportunity or lever by which to shape inequality. But the
structure and so its role should be understood value of this counterfactual approach is contin-
within the broader context of other societal insti- gent on whether schools really are a primary
tutions and other social forces. For us, concen- source of inequality. If this assumption is wrong
trating on variation within school systems alone then the school-centric approach has consider-
runs the risk of distorting how schools really mat- ably less value.
ter. Our goal is to identify the kinds of social con-
ditions in general (school or non-school) that We recommend a different counterfactual—
influence inequality. “What would inequality look like if children’s
exposure to school changed?”—because it pro-
And while we are interested broadly in the vides a view of schools’ overall role in the strati-
relationship between schooling and inequality, fication system (Raudenbush and Eschmann
we limit our focus here to the formal schooling 2015). The traditional approach, focusing on
opportunities readily available to all children variation among schools, lacks the breadth neces-
(kindergarten through twelfth grade in the United sary to allow us to see the big picture. It is diffi-
States) because our primary interest is in whether cult to assess whether schools increase or
publicly provided mass education really does decrease inequality, for example, simply by doc-
serve as a “great equalizer.” Of course, there exist umenting variations among schools.
other kinds of “schooling” that are not provided
publicly (or are only partly subsidized) and there- One problem is that schools might provide
fore depend more heavily on parents’ resources, advantages to high-socioeconomic children, yet
such as preschool, shadow education, private still be an equalizing force (Downey et al. 2004),
schools, summer programs, and higher educa- as presented in Fig.  3.1. This could occur if
tion. At times, it is difficult to separate the school- unequal schools are more equal than the condi-
ing that is provided publicly from the schooling tions children experience when they are not in
that is provided privately. For example, achieve- school. In this way schools could be an equaliz-
ment gaps at kindergarten entry are probably ing force by reducing the level of inequality we
influenced to some degree by school exposure would observe in their absence. Importantly, we
(e.g., preschool), and so do not strictly represent would not be able to identify this pattern if we
“non-school” factors. But for our purposes, they focused on the traditional counterfactual.
represent the magnitude of the gap prior to the
onset of widely available publicly funded school- In addition, the traditional approach struggles
ing. What happens after that is our primary inter- to isolate school from non-school effects. The
est in this chapter. 800-pound gorilla problem education scholars
face is that children are not randomly assigned to
3.3.1 T he Seasonal Comparison schools and so differences in how children learn
Method in one school versus another could represent
either school or non-school factors. The
Traditional research frames the question as “How “measurement-­based” approach to this challenge
well would a particular student perform if they is to isolate school “effects” by identifying all
attended school A versus school B?” This fram- relevant non-school factors and statistically con-
ing promotes research aimed at determining trolling for them in a regression model. This is
whether children would have learned more had common practice but it is also insufficient
they experienced a different school or particular because scholars cannot identify and measure
school practice. Many scholars and policymakers perfectly all of the relevant factors that influence
are attracted to this counterfactual because they children’s development. In a sobering example of
assume that schools are the primary problem and/ the limitations of this method, Burkam et  al.
(2004) note that, even in models including an
impressive array of measures of the non-school
environment, they were unable to explain more

3  Schools and Inequality: Implications from Seasonal Comparison Research 59

Fig. 3.1  School as
equalizers. (Source:
Adapted from Downey
et al. (2004))

than 15% of the variation in summer learning between the seasonal comparison method and the
among children in the Early Childhood cross-­over designs employed by medical research-
Longitudinal Study—Kindergarten Cohort of ers. Medical researchers testing the effectiveness
1998.3 As a result, even in models with what of a drug may observe patients off treatment for a
seems like a comprehensive set of statistical con- period, and then observe how they change when
trols, students at two different schools may learn on treatment (von Hippel et al. 2007). The differ-
at different rates during the year because of ence between the two periods provides an esti-
“unknown differences” in their non-school envi- mate of the treatment effect. Similarly, comparing
ronments that go unmeasured. These “unknown how achievement gaps change when school (treat-
differences” in non-school environments distort ment) is in versus out provides leverage for under-
estimates of school effects in a predictable way, standing how schools matter.
making them appear larger than they really are.
While not a randomized experiment, the sea-
Seasonal comparison scholars approach the sonal design is a powerful method for separating
problem from a different angle. They leverage the the effects of the school and non-school environ-
seasonal nature of the American school ment because there are no differences between
c­alendar—9 months of school followed by a subjects receiving the treatment and those receiv-
3-month summer break—which provides a natu- ing the control—each subject is observed under
ral experiment for understanding how schools both conditions and serves as his or her own con-
matter (Gangl 2010). Scholars compare how trol. This means that there is no need to identify
achievement gaps change when school is in ver- all the various school and non-school processes
sus out, thereby gaining leverage on the schools’ at stake because the overall consequence of all
role in producing these gaps. Note the similarity mechanisms (both exacerbatory and compensa-
tory) is observable in how inequality changes
3 Burkam et al. (2004) predicted summer learning (fall first when school is in session versus out of session.
grade score minus spring kindergarten score) with socio-
economic status, race, gender, age, repeat kindergarten The advantages of seasonal comparisons over
status, family structure, home language (English or not), more traditional education scholarship are multi-
summer trips, summer literacy activities, computer for ple. First, they provide a better method for over-
educational use, and summer school attendance. They coming the formidable obstacle of isolating school
explained 0.079%, 0.136%, and 0.131% of the variation effects. Second, most traditional scholars target a
in literacy, math, and general knowledge learning respec- specific school process thought to increase
tively. Clearly, the vast majority of why some children achievement gaps (e.g., class size), which repre-
learn faster than others during the summer is not captured sents just one of the many school processes that
by the information typically available in large data sets.

60 D. B. Downey et al.

shape inequality in schools. These studies are not of our own included, relied on scales that were
without value—we can learn something about later found to fall short of this interval-level
whether a particular school practice increases requirement. More recent scales appear to
inequality—but they tell us little about how all approximate interval-level characteristics more
exacerbatory and compensatory school processes closely, but the field would still benefit from
stack up against each other. If we want to under- greater use of non-parametric methods (Ho and
stand schools’ overall effect we need to identify all Reardon 2011) that would be less dependent on
processes at stake (exacerbatory and compensa- this assumption and allow researchers to use sea-
tory) and compare their relative strength.4 Seasonal sonal methods across a broader range of depen-
comparison studies achieve that goal. Finally, tra- dent variables. Some scholars have observed
ditional education scholarship lacks the scope to changes in gaps across scales that may be inter-
assess whether schools, as a whole, do more to val level, like theta scores, and those that are
reduce or increase inequality. The problem is that clearly not, like standardized versions of theta
the school-centric approach merely looks at varia- scores. The first approach gauges whether a gap
tions in school conditions without considering the in skills changed over time and depends on inter-
bigger question, how do schools matter overall? val level assumptions. The second approach con-
The possibility that unequal schools might still be siders whether a group’s relative position in the
an equalizing force (Fig.  3.1) goes overlooked distribution changed over time (Quinn 2015;
with traditional methods. Quinn et al. 2016).

Of course, the seasonal comparison approach In addition, it is important that nothing else of
requires assumptions and these have yet to be consequence change across the summer and
scrutinized in the way that they should. Perhaps school year other than children’s exposure to
the most critical assumption is that reading and schooling. Similar to the cross-over designs in
math skills are measured on interval-level scales, medical research, we need to be confident that
and so gains at the bottom of the scale are exposure to the “treatment” is the only thing dif-
assumed to be comparable to those at the top. If ferent between treatment and non-treatment peri-
it is easier to register gains at the bottom of the ods. One can imagine ways in which this
scale than the top, then it is hard to interpret the assumption might be violated in seasonal com-
seasonal patterns.5 Early seasonal studies, some parison studies. For example, when children are
in school, we would expect parents’ time with
4 It is important to recognize that with this kind of study children to decline relative to the summer peri-
design we do not look to the treatment period alone for ods. When focusing on achievement gaps, this
our estimate of the treatment effect. We should not make could be problematic if non-school factors
the mistake, therefore, of simply observing the school- change across seasons and they do so differently
year patterns as a way of understanding how schools mat- across groups. For example, suppose high-SES
ter. If we just focus on the school year we would parents out-invest their low-SES counterparts
mistakenly conclude that high- and low-SES children during the summer and that this advantage
learn at roughly the same rate, and so schools play a increases during school periods. If this is the
mostly neutral role. But the proper way to understand how case, then seasonal comparisons underestimate
schools matter is to compare the treatment (school year) how good schools are for low-SES children
period to the control period (summer). When we make because they might misattribute, for example, an
that proper comparison, we learn that schools are compen- increase in SES-based achievement gaps
satory with respect to SES-based gaps in math and read- observed during the school year, to school rather
ing because they reduce the magnitude of the gaps we than non-school factors. Or, alternatively, if high-­
would observe in their absence (Downey et  al. 2004; SES parents out-invest low-SES parents during
Entwisle and Alexander 1992). See Downey and Condron the summer, but this pattern reverses during
(2016) for further discussion on this point. school periods, then seasonal comparison pat-
terns might underestimate the extent to which
5 We would note, however, that this issue is an awkward
explanation for seasonal patterns because it needs to be
applied selectively—the problem exists during the school
year but not the summers.

3  Schools and Inequality: Implications from Seasonal Comparison Research 61

schools advantage high-SES children. This So what do we learn about schools and

assumption is especially difficult to assess, but inequality if we employ the seasonal method?

relevant work fails to find much evidence that Below we describe patterns for achievement gaps

SES-based patterns in parental investments across socioeconomic status, race, and gender.

change systematically across seasons. For exam- We emphasize the magnitude of achievement

ple, high-SES parents are more likely than low-­ gaps at kindergarten entry, along with how school

SES parents to enroll their children in dance and exposure modifies the trajectory of those gaps

music classes during the summer, and both during 9-month school sessions versus summer

groups increase the likelihood of enrolling their periods. We start by recalling the patterns from

child in dance and music during the school year, early seasonal comparison studies before dis-

but the direction and magnitude of this advantage cussing more recent studies.

is roughly similar across seasons (Downey et al.

2017).
Finally, seasonal comparisons assume that 3.3.2 E arly Seasonal Comparison
Studies
summers represent “non-school” periods of

learning, but in reality school processes likely

contaminate most summer estimates. One Seasonal studies go back nearly a half century.

problem is that students are typically not One of the earliest seasonal studies was of over

assessed on the very first and last days of 600 children in New  York City elementary

school, and so when scholars estimate summer schools from 1965 to 1967. Researchers reported

learning between the spring of one academic that the gap in reading skills between high-­

year and the fall of the next, there are usually income White and low-income minority schools

several days of schooling on each end. Scholars grew at a faster rate during the summer than dur-

attempt to reduce the severity of this problem ing the school year (Hayes and Grether 1983).

by modeling the learning that occurs during This same pattern was replicated in New Haven

these school days and removing it from the (Murnane 1975) and in Atlanta (Heyns 1978).

estimate of summer learning, but this is an Perhaps most widely-known, however, is

imperfect approach. Entwisle and Alexander’s Beginning School

These assumptions should give scholars pause Study (BSS) of nearly 800 first graders who were

regarding seasonal results, but we posit that they followed seasonally until sixth grade and then

are significantly more palatable than the assump- into adulthood. In a series of widely-cited publi-

tions required for more traditional approaches. cations, Entwisle and Alexander demonstrated

For example, the notion that scholars can isolate that gaps in math and reading skills grew faster in

school effects by statistically controlling for the summer than the school year (Alexander et al.

observables of the family environment (e.g., 2007, 2014; Entwisle and Alexander 1992;

socioeconomic status, family structure, race) Entwisle et al. 1994). Indeed, among ninth grad-

available in surveys or by estimating school year ers, the authors found that one-third of the read-

learning gains with covariates is most certainly in ing gap between high- and low-socioeconomic

error and, as a result, produces patterns that con- children could be traced to the gap that was

sistently overestimate the negative effects of already present at the beginning of first grade,

schools. Given that our conclusions about how and two-thirds of the gap was due to the summers

schools matter for socioeconomic achievement in between the school years (Alexander et  al.

gaps change dramatically based on which 2007). The entire gap, therefore, was a product of

approach we use—schools increase inequality non-school forces. From the BSS comes the term

(traditional method) versus schools reduce “summer setback,” widely used to explain the

inequality (seasonal comparison method)—we loss of skills observed among low-income chil-

think the results from seasonal comparison dren during the summer. The studies’ patterns

research merit special attention. were popularized in a Time magazine article, and

62 D. B. Downey et al.

have motivated the proliferation of summer pro- representative of American children, were col-
grams designed to prevent setback among disad- lected on a seasonal schedule, include scales of
vantaged children (Von Drehle 2010). cognitive skills that approach interval level, and
have individual-level measures of socioeconomic
Perhaps more important than these empirical status. Of course, a limitation is that the ECLS-K
patterns, however, was the insight the authors studies only follow children seasonally until the
provided in terms of framing the question. Rather end of first grade (1998) and second grade (2010).
than merely focusing on the summer patterns as
the period when gaps grow, Alexander (1997) To estimate seasonal patterns beyond second
pointed out that the summer and school-year pat- grade researchers must revert to small-scale local
terns in combination suggest that, when it comes studies or they can employ an extract of data
to inequality, schools are “more part of the solu- from the Growth Research Database collected by
tion than the problem.” the Northwest Evaluation Association (NWEA).
The NWEA is a private non-profit organization
3.4 A Review of Recent Seasonal that partners with school districts to assess chil-
Comparison Studies dren’s math, reading, and science skills and then
provides schools with reports of children’s prog-
In this next section, we focus on what the more ress. The NWEA assesses children both at the
recent data sets reveal about the first few years of beginning and end of the school year (and some-
schooling. We are especially interested in the times winter), producing a rich seasonally-­
magnitude of achievement gaps at kindergarten collected database of over ten million American
entry because, if they are large relative to changes children from kindergarten through twelfth
in the gap, then most of the “action” generating grade. These advantages are countered, however,
inequality occurs prior to formal schooling.6 by the fact that the NWEA data are not nationally
representative and each researcher tends to ana-
Our review draws on several studies that have lyze their own unique extract of the overall data-
employed different seasonal data sets, but we end base, making it a challenge to compare results
up emphasizing patterns from the ECLS-K: 1998 from different NWEA-based studies. In addition,
and the more recent ECLS-K: 2010 for several the NWEA data lack individual-level information
reasons. Both ECLS-K data sets are nationally on children’s socioeconomic status. Finally, chil-
dren in each school are not necessarily represen-
6 Studying the kindergarten and elementary school years tative of the students in that school. Some
may offer an additional methodological advantage. Some districts, for example, may have tested all stu-
evidence suggests that children learn more rapidly during dents while others may have tested a subset.
these early years, about four times faster than during high
school (LoGerfo et al. 2006). It is hard to know if young 3.4.1 S ocioeconomic Gaps
children actually learn faster or if this pattern is merely an in Cognitive Skills
artifact of the tests—early tests focus on more basic skills
while later tests focus on the development of subject-spe- There is growing consensus that socioeconomic
cific course knowledge. Regardless of whether these pat- (SES) achievement gaps are developed predomi-
terns are real or an artifact, they have consequences for nantly prior to kindergarten entry (Duncan and
our ability to distinguish the learning “signal” from the Magnuson 2011; Reardon 2011a). Analyzing the
“noise” produced by test measurement error. This issue ECLS-K: 1998, Duncan and Magnuson (2011)
becomes especially important when we estimate chil- estimate that children from families in the top
dren’s summer learning rates that rely on test scores only SES quintile begin school, on average, 1.26 stan-
a few months apart. Given the tests that are currently dard deviation (SD) units ahead in reading and
available, high school students only demonstrate modest 1.34 standard deviation units ahead in math com-
learning gains, making it difficult to estimate learning pared to children from families in the bottom
accurately during the 9-month school year and even more
difficult to confidently estimate summer patterns. In con-
trast, young children demonstrate much faster learning
growth on currently available tests, producing a clearer
picture of schools’ role.

3  Schools and Inequality: Implications from Seasonal Comparison Research 63

SES quintile. Moreover, these gaps remain rela- 3.4.2 R acial/Ethnic Gaps
in Cognitive Skills
tively stable throughout the first few years of

school, growing only slightly larger to 1.43 SD in

reading and 1.38 SD in math by the end of fifth Seasonal comparison patterns also can shed light

grade. That means that 90% or more of the fifth on the role that schools play generating or main-

grade gaps are already in place at kindergarten taining racial/ethnic achievement gaps. The

entry. When it comes to understanding SES-­ Black–White gaps at kindergarten entry are sub-

based gaps in math and reading, the early child- stantial. For the ECLS-K: 1998 cohort, Fryer and

hood years prior to kindergarten entry are the Levitt (2004) estimate the gaps to be at 0.64 SD

dominant force. in math and 0.40 SD in reading. Using the more

The fact that the SES gaps grow little once recent 2010 cohort, Quinn (2015) estimates

school starts is the major story, but we also learn slightly smaller gaps at 0.54 SD in math and 0.32

something by observing whether the gaps grow SD in reading. And compared to the SES gaps,

faster when school is in versus out. Analyzing the the Black–White gaps increase more as children

ECLS-K: 1998, Downey et al. (2004) clarify that progress through school (Condron 2009; Fryer

the SES gaps grow faster during the summer and Levitt 2004; Quinn 2015; Reardon et  al.

months between kindergarten and first grade than 2009), although this growth is modest. Von

the school periods, suggesting that even the mod- Hippel and Hamrock (2016) find that, in the

est growth in the SES gap that occurs during the ECLS-K: 1998 data, between first and eighth

school years is driven primarily by the non-­ grades, unstandardized Black–White gaps

school environment. These findings support pre- increase by 22% in reading and 6% in math.8 The

vious seasonal research from Baltimore majority of the Black–White gap is largely

(Alexander et al. 2007; Entwisle and Alexander formed before formal schooling begins—high-

1992) and Atlanta (Heyns 1978). The more recent lighting how the early childhood environment

ECLS-K data, collected beginning in 2010, pro- plays a critical role in generating the gap.

duce a somewhat mixed picture. Schools look Once school begins, there is mixed evidence

compensatory across kindergarten, more neutral regarding whether the Black–White gap grows

during first grade, and may even play a perni- faster when school is in versus out. Some schol-

cious role during second grade, at least for read- ars find that schools play a role reducing the gap.

ing skills (Quinn et  al. 2016), raising the Heyns’ study of sixth and seventh graders in

possibility that compensatory school effects for Atlanta noted that the Black–White gap grew

socioeconomic status are strongest during faster during the summer than school year (Heyns

kindergarten.7 1978). However, studies relying on broader sam-

ples reach the opposite conclusion. Analyzing

7 One caveat to the general SES pattern is that NWEA kindergartners through eighth graders in 14 states

extracts do not always produce consistent results. In the

most extensive analysis of seasonal data sets to date, von preting the NWEA patterns, however. For example, the
Hippel and Hamrock (2016) compared patterns across the NWEA lacks an individual-level socioeconomic indica-
BSS, ECLS-K: 1998, and an NWEA extract covering 14 tor, and so von Hippel and Hamrock (2016) had to com-
states and concluded that “The preschool years are the pare school-level gaps across Title 1 and non-Title 1
period of fastest gap growth; after school starts, it is hard schools. Another challenge interpreting the NWEA pat-
to say unequivocally whether gaps grow faster during terns is that various scholars typically analyze unique sub-
school or during summer.” This impressive analysis rein- sets of the larger Research Growth Database, making
forces previous findings that most of the gap develops dur- replication difficult.
ing the early childhood years, but raises questions about

whether the SES gaps grow faster during the summers or 8 For our purposes, it would be better if this study had esti-

school periods, once schooling begins. In von Hippel and mated how the gaps increase from the beginning of kin-

Hamrock’s (2016) study, ECLS-K patterns were consis- dergarten rather than first grade, but we know from other

tent with the notion that SES gaps grow fastest when studies that the Black–White gap increases only slightly

school is out, but the patterns from the NWEA extract during kindergarten, and so these estimates would only

were at times contradictory. There are challenges inter- increase slightly.

64 D. B. Downey et al.

from the NWEA, von Hippel and Hamrock Downey et  al. (2004) clarify that the decline in
(2016) report that the Black–White gap tends to the Asian-American advantage occurs during the
grow faster during the 9-month school periods school year. They found that Asian-American
than during the summers. And using the ECLS-K: students had higher academic achievement than
1998 data set, Downey et  al. (2004) also found White students in the first two years of school,
that Black students exhibited a similar rate of but that these advantages were primarily main-
learning (relative to White students) during the tained through faster rates of learning during the
summer after kindergarten, but fell behind during summer months. Similar patterns are found in the
kindergarten and first grade, even after control- ECLS-K: 2010 for reading, but not for math
ling for SES.9 The more recent ECLS-K: 2010 (Quinn et  al. 2016). Asian-American students
also suggests that the Black–White gap grows begin kindergarten with significant advantages in
larger during the school years but either stabilizes both subjects. Nevertheless, the Asian–White gap
or narrows during the summer months (Quinn in math does not change, while the gap in reading
et al. 2016). begins to narrow. Furthermore, the seasonal pat-
terns reveal that the reading gap specifically
The evidence regarding the Black–White gap declines during the kindergarten and first grade
is mixed, but we think it leans more in one direc- school years, while Asian students learn at simi-
tion—that schools play a pernicious role. We say lar or even faster rates than White students during
this because the studies that have found that the the summer months (Quinn et al. 2016). Beyond
Black–White gap grows faster during the school second grade, Yoon and Merry (2015) analyzed
year than summer have relied on broader, more second to seventh graders in the NWEA data and
generalizable data than the studies that have noted that the decline in the Asian-American
found the opposite pattern. advantage mainly occurred during the school
year, and Asian-American students recuperated
There is also a growing group of studies focus- their loss during the summer periods. Although
ing on the Asian–White gap. These studies pro- the evidence is still accumulating, there is reason
vide provocative evidence from the ECLS-K and to worry that schools may play a role reducing
NWEA surveys that schools may undermine the the educational progress of Asian-American stu-
performance of Asian-American students dents compared to White students.
(Downey et al. 2004; Quinn et al. 2016; Yoon and
Merry 2015). In 1998, Asian-Americans began The seasonal findings for the Latino–White
kindergarten with a 0.11 SD advantage in math gap are the most limited and inconsistent.
and a 0.31 SD in reading relative to White stu- Latino/a students begin school the furthest behind
dents. Nevertheless, the Asian-American advan- White students. Once schooling begins, however,
tage begins to fade with the onset of formal it is unclear what happens to the gaps; some stud-
schooling, and completely disappears by the end ies find that gaps begin to close (Fryer and Levitt
of third grade (Fryer and Levitt 2006). In their 2004; Han 2008; Reardon and Galindo 2009),
seasonal analysis using the ECLS-K: 1998, while others find that only the gap in math shrinks
while the reading gap remains the same (von
9 Confusingly, utilizing the same ECLS-K: 1998 data set, Hippel and Hamrock 2016), and some studies
one study finds that Black students experience summer find that both gaps remain unchanged (Quinn
setbacks in math (Burkam et  al. 2004). Nevertheless, et  al. 2016) Overall, the limited seasonal com-
Quinn (2015) clarifies that these contradictory findings parison studies on Latino/a students suggest that
result from variation in modeling strategy, test metric, and schools are compensatory for math, but results
assumptions about measurement error. Burkam et  al. for reading are inconsistent. In both ECLS-K
(2004) explored conditional growth and found that Black data sets and the NWEA, scholars note that the
students who had the same spring scores as White stu- standardized Latino–White gap in math narrows
dents made slower math gains during the summer, but during the school year and grows faster in the
overall, Black and White students learn at similar rates summer months, suggesting that schools promote
during the summer and there is little evidence to show that
the summer period contributes to the growing Black–
White gaps (Quinn 2015).

3  Schools and Inequality: Implications from Seasonal Comparison Research 65

the educational progress of Latino/a students Of course, if schools play a unique role in pro-
(relative to White students) (Quinn et  al. 2016; moting gender gaps, one way or the other, we
von Hippel and Hamrock 2016). On the other would expect that gender-based gaps would grow
hand, these same studies find mixed results for faster when school is in versus out. Notably, sea-
reading, with ECLS-K: 1998 and the NWEA sonal comparison studies have tended to focus
suggesting that summers are responsible for gap their attention on SES gaps, and to a lesser extent,
growth while the ECSL-K: 2010 indicates that racial/ethnic gaps, while gender gaps have
schools are responsible. These divergent findings received little attention. Still, in some of the
may be due to the vast heterogeneity within the tables from seasonal comparison research we can
Latino/a group. For example, one study that dis- glean the necessary patterns. In Downey et  al.
aggregated the group by country of origin (2004) the authors combined the school period
reported that certain Latino/a groups (i.e., Central learning rates (kindergarten and first grade) for
American and Cubans) reach equivalent achieve- reading and math and found that, overall, gender
ment levels as White students by third grade gaps operated similarly during the school periods
despite their disadvantaged beginnings (Han and the summer (Downey et al. 2004, pp. 628–
2008). 629, Table 4). Similarly, Entwisle and Alexander
1992) analyzed the Baltimore data and reported
3.4.3 G ender Gaps in Cognitive that seasonal patterns of growth did not vary by
Skills child’s gender. These two patterns are consistent
with the view that schools are not the driving
How do schools shape gender gaps in cognitive force behind the changes in the gender gap dur-
skills? With respect to math, girls exhibit a mod- ing the first few years of schooling. We are
est advantage before kindergarten but then lose unaware of other seasonal comparison research
that advantage after school starts (Gibbs 2010). that compares how the gender gap in skills
Gibbs (2010) differentiates between types of changes when school is in versus out and so we
mathematical content to better understand the so- urge scholars to build greater empirical knowl-
called “reversal of fortunes” that girls experience edge in this area.
in terms of math achievement. Using ECLS-B
and ECLS-K: 1998 data, he found that girls 3.4.4 O verall Variation in Cognitive
excelled at less complex math skills throughout Skills
childhood, but experienced disadvantages when
the content became more complex. These pat- An additional way of considering how schools
terns direct our attention to schools as a potential matter is to ask—How does overall variation in
source of the gender gap in math skills. skills (among all children) change when school is
in versus out (Meyer 2016)? As Downey et  al.
Patterns for reading skills are different. Girls (2004) pointed out, SES, race, and gender explain
tend to begin kindergarten with better reading only a small fraction of the variation in children’s
scores than boys and their advantage increases skills—less than 10%. Of course, achievement
throughout kindergarten. Controlling for ethnic- gaps across social groups are of interest, but by
ity and poverty, boys are behind girls in reading analyzing overall variation, we may produce a
by 0.17 standard deviation units at the time of more comprehensive understanding of how
kindergarten entry and the gap increases to 0.31 schools influence inequality. Few scholars have
at the end of grade one (Chatterji 2006; Ready considered whether overall variation in skills
et al. 2005). So before turning to seasonal stud- grows faster when school is in versus out, but the
ies, these studies focusing on the growth in the exceptions are revealing. Analyzing the ECLS-K:
gaps during the first few years suggest that 1998, Downey et al. (2004) found that variation
schools may disadvantage girls with respect to in cognitive skills grew much faster during the
math and boys with respect to reading.

66 D. B. Downey et al.

summer versus school year—58% faster for read- The main message from our review is that
ing and 40% faster for math, patterns replicated achievement gaps are well-established prior to
in more recent work with the ECLS-K: 2010 kindergarten entry. This pattern highlights how
(Downey et al. 2017). early childhood experiences prepare children
unequally and send them on different learning
3.5 C onclusion trajectories. Studying achievement gaps in
schools has value, of course, but if we want to
We reviewed studies revealing the magnitude of understand why gaps emerge in the first place, we
achievement gaps at kindergarten entry, along need to focus more attention on early childhood.
with how those gaps change over the next few With respect to socioeconomic gaps in cognitive
years, when school is in versus out. The value in skills, the time prior to kindergarten explains the
this analytic approach is that it more confidently vast majority of why high-SES children outper-
separates school from non-school influences, a form low-SES children during the elementary
major stumbling block for most research designs school years. The race and gender gaps are
attempting to understand how schools matter. We smaller in magnitude than the SES-based gap,
acknowledge that this strategy falls short of a but they are also significantly formed prior to the
comprehensive analysis of the relationship onset of formal schooling. Achievement gaps are
between schools and inequality because we often observed in schools, but they are primarily
restrict our discussion to children’s cognitive formed by early childhood processes that have
skills, and then restrict our focus even further to little to do with schools (defined by formal
the first few years of schooling.10 A broader schooling available to all).
review would consider a wider range of outcomes
and extend into later stages of education.11 Another conclusion from this work is that
Nevertheless, the patterns from this exercise tell schools do not consistently advantage the already
us quite a bit about how large achievement gaps socioeconomically advantaged. There is very lit-
are prior to kindergarten, and what schools tend tle evidence that schools increase SES-based
to do those gaps over the next few years. achievement gaps; in fact, they are probably an
important compensatory force, especially during
10 It is possible that the patterns we report here, emphasiz- kindergarten. We say this because the SES gaps
ing kindergarten and the next couple years, are unique and grow when school is out, and are mostly
do not apply to later stages of the educational career. unchanged when school is in. The more children
Some scholars have questioned whether seasonal patterns are exposed to schools, the smaller the socioeco-
persist into high school, for example, where tracking nomic gaps in skills. Schools, therefore, are prob-
mechanisms may produce greater school-based inequality ably compensatory, reducing the magnitude of
(Gamoran 2016). It is worth noting, however, that prior to the SES gap we would otherwise observe in their
seasonal comparison analysis, most scholars assumed that absence. And, when we expand our focus to con-
schools increase achievement gaps, even among young sider how overall variation in children’s skills,
children. Given that seasonal analysis reversed this view, schools’ compensatory power is even clearer—
we think it is important to refrain from making a similar variation in children’s skills grows about 50%
mistake before we have seasonal analysis of high faster when out of school versus in.
schoolers.
11 It is worth noting that when seasonal comparisons are Our inferences regarding racial achievement
applied to other dependent variables we also tend to come gaps are more mixed. There are some indications
away with more favorable views of schools. For example, that schools play a pernicious role. The Black–
children’s body mass index tends to grow about twice as White gap, for example, grows faster during the
fast during the summer versus school year (von Hippel first three years of school than during the sum-
et al. 2007), and there does not seem to be any consistent mers in between, a pattern implicating schools
pattern to how SES, racial/ethnic, and gender gaps in (Downey et  al. 2017; Quinn et  al. 2016). The
social-behavioral skills change when school is in versus strongest evidence that schools undermine the
out (Downey et al. 2016b). educational achievement of Black students is the

3  Schools and Inequality: Implications from Seasonal Comparison Research 67

seasonal patterns for the standardized Black– schools than the more traditional methods. This
White reading gap using the ECLS-K: 2010. is noteworthy and should cause scholars employ-
Quinn et  al. (2016) finds that while the reading ing the more traditional methods to reconsider
gap grows during kindergarten, first, and second whether schools really exacerbate inequality in
grade, the gap significantly narrows in the two the way many have argued. This is not to say that
summers in between. In other words, Black stu- seasonal comparison methods provide the defini-
dents fall behind White students during the school tive word on how schools matter, but rather that
years, but learn at a significantly faster rate than their methodological advantages should prompt a
White students during the summer months when renewed discussion about why some studies tend
they are no longer exposed to schools. There is to describe schools as exacerbatory, while sea-
also evidence that schools reduce the Asian– sonal comparison studies produce a different
White gap, which may simply be a product of conclusion.
schools’ overall compensatory power, or it may
reflect a race-based process within schools that Of course, if schools play a more favorable
has been inadequately studied. role in the stratification system than they are gen-
erally given credit for, by what processes are they
We also applied our method to gender differ- actually reducing inequality? Sociology of edu-
ences in cognitive skills. There has been consid- cation scholars have created a wide range of plau-
erable discussion about how girls have surpassed sible mechanisms by which schools might
boys in school on a wide range of educational exacerbate inequality, but considerably less theo-
outcomes and the role that schools might play in retical effort has gone into understanding how
that process (Diprete and Buchmann 2013). schools might be compensatory (Downey and
Some have suggested that classrooms have Condron 2016). It is difficult to know what these
become a feminized environment, more condu- mechanisms might be because seasonal compari-
cive to girls’ ways of learning. If it were true, we son studies do not provide that insight, but we
would expect that girls’ advantage would grow can speculate. We suspect that schools may
faster when school is in session than during the reduce SES achievement gaps and overall vari-
summer, but we rarely observe that pattern. ance in skill because they consolidate children’s
Although schools may influence gendered out- curriculum experiences (by organizing children
comes in later grades, the seasonal patterns dur- by chronological age) more than they differenti-
ing the early grades produce no “school reason” ate curriculum via ability grouping and tracking.
for the gaps.12 In addition, despite the discriminatory processes
uncovered in some research, it may be that teach-
What does all this mean for how we under- ers generally operate in an egalitarian manner,
stand the relationship between schools and helping disadvantaged children the most. For
inequality? Seasonal comparison methods pro- example, a national survey of teachers found that,
vide a different, and we believe valuable, way of when asked who was most likely to receive one-­
understanding how schools matter. This window on-o­ ne attention, 80% of teachers said “academi-
into the relationship between schools and inequal- cally struggling students” while just 5% said
ity ends up producing a more positive view of “academically advanced” students (Duffett et al.
2008).
12 The three demographic characteristics studied here
(socioeconomic status, race/ethnicity, and gender) all pro- Finally, the seasonal results prompt us to
duced different seasonal patterns. It is worth noting that reconsider what the most effective school poli-
socioeconomic status is an indicator of diverse home and cies might be for reducing achievement gaps. We
neighborhood resources while gender is a socially con- would support increasing the amount of school-
structed status largely uncorrelated with these non-school ing available to all children because exposure to
conditions and race/ethnicity has characteristics of both. public schooling appears to reduce socioeco-
This distinction may explain why we see the clearest sea- nomic achievement gaps and the growth in over-
sonal patterns for socioeconomic status, the weakest for all variation in skills. If the U.S. expanded the
gender, and patterns somewhat in between for race/
ethnicity.